1-800 Contacts, Inc.

Text (OCR of the scan at left; may contain errors)

IN THE MATTER OF 1-800 CONTACTS, INC.

OPINION OF THE COMMISSION AND FINAL ORDER IN REGARD TO ALLEGED VIOLATIONS OF SECTION 5 OF THE FEDERAL TRADE COMMISSION ACT Docket No. 9372; File No. 141 0200 Complaint, August 8, 2016 – Decision, November 7, 2018 This case addresses 1-800 Contacts, Inc.’s settlement agreements with over a dozen rivals prohibiting both 1-800 and the other parties from bidding on keywords containing the other’s trademarks, and requiring each party to implement negative keywords to ensure that their ads do not appear in search engine results pages for searches that contain each other’s trademarks. The complaint, 164 F.T.C. 360, alleges that these agreements are in restraint of trade in violation of Section 5 of the Federal Trade Commission Act. In his Initial Decision, 164 F.T.C. 373, the Administrative Law Judge found that the challenged agreements pose significant, unjustified anticompetitive consequences in the relevant market for the sale of contact lenses online. The Administrative Law Judge ordered 1- 800 Contacts to cease and desist from enforcing or attempting to enforce any and all provisions, terms, or requirements in an existing agreement or court order that impose a condition on a Seller, which prohibits, restricts, regulates, or otherwise places any limitation on truthful, non-deceptive, and non-infringing advertising or promotion. Id. at 637. The Respondent appealed the Initial Decision. The Commission scheduled oral argument, 165 F.T.C. 1415, and subsequently rescheduled the arguments, 165 F.T.C. 1471. The Commission heard oral arguments on June 26, 2018.

Participants For the Commission: Gustav Chiarello, Kathleen Clair, Joshua Barton Gray, Stuart Hirschfeld, Nathaniel Hopkin and Charlotte Slaiman.

For the Respondent: Garth Vincent, Munger Tolles & Olson; Darryl Nirenberg, Steptoe & Johnson.

OPINION OF THE COMMISSION By Chairman Joseph J. Simons, for the Commission:

This proceeding considers Complaint Counsel’s challenge to a number of agreements among horizontal competitors—in most instances, trademark litigation settlements—that, allegedly, anticompetitively limit internet search advertising and restrict bidding in internet search auctions to the detriment of consumers. Respondent 1-800 Contacts sued rival contactlens sellers for trademark infringement when sellers’ online advertising appeared in response to consumers’ internet searches for “1-800 Contacts.” In nearly all cases, the litigation settled before trial. The resulting settlement agreements require the parties, when bidding at search engine advertising auctions, to take steps to ensure their ads do not appear in response to searches for the other party’s trademark terms.

At first glance, this proceeding may appear to contemplate little more than a few terms embedded in a document that purports to resolve a trademark dispute among internet sellers of 251 1-800 CONTACTS, INC. Opinion of the Commission contact lenses. But, in reality, this case grapples with issues of enormous import. We consider here consumer marketplaces that embody the very basic institutions of 21st century commerce. Increasingly, consumers no longer shop for goods by walking down Main Street and looking at the price tags on window displays or by wandering through the aisles of retail establishments comparing prices on shelves and product characteristics written on packages. Rather, consumers now frequently—and with increasing frequency—open their web browsers, enter desired product names or qualities into a search engine, and wait for Main Street or supermarket aisles to be digitally transported to them. This phenomenon is comparatively recent, but e-commerce already comprises a significant and growing share of our economy’s retail sales. Indeed, the Census Bureau estimated that e-commerce retail sales in the United States totaled $127.3 billion in the second quarter of 2018, which comprised approximately 9.6 percent of total retail sales.1 We consider here the manner in which and conditions under which prices for contact lenses are advertised throughout the internet economy. Our decision will affect not only the price that consumers pay for some contact lenses but also the very manner in which substantial parts of price competition will occur throughout consumer markets today and tomorrow. As this agency has explained time and again, robust, accurate, and intelligible price competition among those who compete for consumers’ dollars is one of the cornerstones of our vibrant market economy. When information is withheld from consumers, it frustrates their ability to compare the prices and offerings of competitors. This is as true today, when consumers search for goods online, as it was when people shopped open-air markets for vegetables every evening. In that important respect, nothing has changed.

Chief Administrative Law Judge (“ALJ”) D. Michael Chappell held a 19-day administrative hearing involving the testimony of 43 witnesses, either live or by deposition, and more than 1250 exhibits. Judge Chappell issued an Initial Decision that held that the advertising restraints at issue harm consumers and competition in the market for the sale of contact lenses online. The ALJ held that the Supreme Court’s decision in FTC v. Actavis, 570 U.S. 136 (2013) did not establish antitrust immunity for the trademark settlements. The ALJ also determined that the agreements do not have countervailing procompetitive benefits that outweigh or justify the demonstrated anticompetitive effects. He therefore concluded that the agreements unreasonably restrain trade in violation of the Federal Trade Commission Act (“FTC Act”). Respondent has appealed, and Complaint Counsel oppose that appeal.

Respondents in this appeal ask us to permit them to eradicate an important form of price competition as a means to protect the intellectual capital embedded in their trademarks. Of course, their claims deserve and receive full and respectful consideration. At the same time, we must be mindful that what is at stake is not only the proper antitrust response to certain lawsuit settlements, but also the very means by which and conditions under which retail price competition takes place in the 21st century internet economy. These are matters vital to the interests of consumers and producers in our evolving marketplace economy. 1 U.S. Census Bureau, U.S. Census Bureau News (Aug. 17, 2018), https://www.census.gov/retail/mrts/www/data/ pdf/ec_current.pdf (estimating adjusted retail e-commerce sales for the second quarter of 2018). VOLUME 166 Opinion of the Commission We affirm the ALJ’s conclusions that Actavis does not confer antitrust immunity. If anything, Actavis follows a long line of cases that holds that patent-related settlements can sometimes violate the antitrust laws. Moreover, Actavis made clear that the effect of intellectual property on the application of antitrust laws in such settlements should be assessed through consideration of traditional antitrust factors. We therefore hold that the challenged agreements unreasonably restrain trade in violation of Section 5 of the FTC Act, although our analysis differs from that of the Initial Decision in some respects. We find that the agreements harm consumers and competition for the online sale of contact lenses. We also find that Respondent has not demonstrated valid offsetting procompetitive justifications for the advertising restraints, and that the restraints were not reasonably necessary to achieve the claimed procompetitive benefits. Consequently, we enter a cease-and-desist order that prohibits 1-800 Contacts from enforcing the unlawful provisions in the challenged agreements and prevents 1-800 Contacts from entering into similar agreements in the future. We also find that challenged agreements harm competition in bidding for search engine key words, artificially reducing the prices that Respondent paid and the quality of the search engine results delivered to consumers—without offsetting efficiencies. I. BACKGROUND A. Respondent, 1-800 Contacts 1-800 Contacts sells contact lenses to consumers throughout the United States. It started as a mail-order contact lens business in a college dorm in 1992. IDF 30-33.2 The business changed its name in 1995 when it obtained the 1-800 Contacts telephone number. IDF 36. The company launched its website in 1996, and beginning in 2004, its internet sales exceeded its telephone sales. IDF 37, 67. In 2015, 1-800 Contacts’ revenues were approximately $460 million. IDF 68. Its annual volume of contact lenses sold via the Internet to U.S. consumers currently exceeds the online sales of contact lenses to U.S. consumers by any other company. IDF 69.

B. The Contact Lens Industry Contact lenses are a billion dollar industry in the United States. IDF 4. Contact lenses are medical devices that can be sold only pursuant to a prescription written by an optometrist or ophthalmologist, also called eye care practitioners (“ECPs”). IDF 8-12. A consumer interested in wearing contact lenses must first visit an ECP for a lens fitting and prescription. IDF 10. A consumer’s prescription specifies the brand as well as the power and other characteristics of the 2 We use the following abbreviations in this opinion:

Compl.: Complaint Answer: Respondent 1-800 Contacts, Inc.’s Answer and Defenses to Administrative Complaint ID: Initial Decision IDF: Initial Decision Finding of Fact Stip.: Joint Stipulation Regarding Search Engine Mechanics and Glossary of Terms RAB: Respondent’s Brief on Appeal CCB: Complaint Counsel’s Answering Brief to Respondent’s Appeal Brief RRB: Respondent’s Reply Brief on Appeal 253 1-800 CONTACTS, INC. Opinion of the Commission contact lenses. A consumer’s prescription expires in one year in most states, and two years in others; consequently, the consumer must regularly return to an ECP. IDF 18-19, 23. In 2003, Congress enacted the Fairness to Contact Lens Consumers Act, which, along with the FTC’s implementing regulations, gives patients an automatic right to their contact lens prescriptions upon completion of a fitting. IDF 17; see also 15 U.S.C. § 7601 (2003). This facilitates their ability to fill contact lens prescriptions through any retail channel they choose. ID at 111. Because prescriptions identify the power, base curve, and specific lens brand, the lenses a consumer receives are identical in every way, irrespective of the choice of retailer. IDF 23-27; Coon, Tr. 2667, 2687; Bethers, Tr. 3612; CX9017 (Blackwood Dep.) at 304. Consumers often buy contact lenses from ECPs when they return every 1-2 years for a new prescription. Other contact lens retailers compete mostly for consumers’ “refill” sales. IDF 76, 403-05. There are four types of retailers in the industry. IDF 73. First, many ECPs operate independent practices (“independent ECPs”) and sell both their services (eye exams) and the products they prescribe (contact lenses). Independent ECPs sell contact lenses directly to consumers and, in 2015, accounted for 40 percent of all contact lens sales in the United States. IDF 75-76, 79, 491. Second, national and regional optical retail chains, such as LensCrafters and Visionworks, accounted for 20 percent of contact lens sales in 2015. IDF 84-85, 491. Third, mass merchants, such as Walmart and Target, and club stores, such as BJ’s and Sam’s Club, sell contact lenses and accounted for 23 percent of contact lens sales in 2015. IDF 90, 93, 491. Fourth, so-called “pure-play” online retailers, such as 1-800 Contacts, sell only online and do not have brick-and-mortar locations; pure-play online sellers accounted for 17 percent of contact lens sales in 2015. IDF 77, 98-99, 491. In 2015, 1-800 Contacts accounted for approximately {54} percent of online sales, which is more than four times the sales of the second-largest online retailer. IDF 495; see also IDF 494 (citing CX9001 (Bethers, IHT) at 159-60 (1-800 Contacts’ CEO testifying that 1-800 Contacts’ sales constituted more than 60 percent of the online contact lens market)).

The price for contact lenses varies significantly based on the retail channel. Among brick-and-mortar retailers, independent ECPs typically have the highest prices for contact lenses, followed by optical retail chains, which, on average, sell contact lenses priced just below those sold by ECPs. IDF 431-32. Mass merchants offer lower contact lens prices than independent ECPs and optical retail chains. IDF 441. Membership club stores have the lowest contact lens prices. IDF 448; Bethers, Tr. 3545.

Online contact-lens retailers—other than 1-800 Contacts—generally offer prices well below those of independent ECPs, optical retail chains, and mass merchants. IDF 442 (citing Bethers, Tr. 3536-37, 3544-45, Clarkson, Tr. 189-90 (“in most cases online pricing is significantly lower than for any of the brick-and-mortar channels, with the exception of the clubs”)), 444 (online retailer AC Lens explaining ECPs’ “prices are typically so much higher”); 446 (online seller Memorial Eye charged significantly lower prices online than it did in its physical stores).

But 1-800 Contacts’ prices are higher than those of other online retailers. IDF 691. It sets its prices below ECPs’ and optical retail chains’ prices, but above prices offered by mass VOLUME 166 Opinion of the Commission merchants and club stores. IDF 433-34, 441; Coon, Tr. 2695, 2708-10; Bethers, Tr. 3543-46. Importantly, 1-800 Contacts’ prices are approximately {15 to 20} percent higher than other online retailer prices. IDF 692 (citing CX8007 (Athey Expert Report at 013-14, 045-51 ¶¶ 31- 32, Exhibit D-1 to D-7)); see also CX0295 at 064, in camera (in January 2014, 1-800 Contacts prices were higher than those of other online retailers by {22} percent per box, {15} percent for a six month supply, and {14} percent for a 12 month supply); RX1228 at 036, in camera (2015 analysis showing that 1-800 Contacts’ prices were higher than those of other online retailers; the net prices of Coastal Contacts, LensDirect, AC Lens, Vision Direct, and Lens.com were {2 percent, 4 percent, 9 percent, 15 percent, and 24 percent} lower than 1-800 Contacts’ net prices).

C. Paid Search Advertising Online retailers use online advertising to attract new customers. Internet search engines, such as Google, Bing, and Yahoo!, allow internet users to search and retrieve content on the World Wide Web. In response to a user’s search query, the search engine employs algorithms to match the text of the query with portions of the Web that may contain relevant content. Stip. at 2, 5. Links to webpages deemed potentially responsive to the user’s search are ranked and presented to the user on a search engine results page (“SERP”). Id. A typical SERP displays two sorts of search results: “organic” links and “sponsored” links, which are advertisements. Stip. at 2; ID at 36. Organic search results are links to websites that the search engine has identified as relevant to the user’s query. No one can pay the search engine to have an organic result appear or to change a result’s rank on the SERP. Rather, the appearance and rank of organic links are based on relevance to the user’s search, with the most relevant results at the top of the SERP. Stip. at 5.

Sponsored links typically are displayed above, below, or to the side of the organic results, and often appear in a different colored box labeled with the word “Ad.” Stip. at 5-6; ID at 36. Google and Bing display up to four search advertisements at the top of the page, above the organic search results. Stip. at 5; IDF 212, 234. As the name suggests, advertisers pay to have sponsored links appear on a SERP. To determine which ads appear, and in what order, search engines use an auction to sell advertising positions. Advertisers bid on “keywords,” which are words or phrases that trigger the display of ads when they are determined to “match” a user’s search. Stip. at 2; IDF 163. But the auction bids alone do not determine whether a particular ad appears. Search engines evaluate other factors, such as an ad’s quality and its relevance to a user’s search query, in determining the ad’s location on a SERP and whether it displays at all. Stip. at 9-12; ID at 26-28. Thus, even a high auction bid will not result in an ad appearing if the search engine does not find the ad relevant to the user’s search. Search engines have an incentive to show relevant ads because search engines are paid for displaying an ad only if the user clicks on the ad.3 Juda, Tr. 1072.

3 The price that an advertiser pays to the search engine each time its advertisement is clicked is the cost-per-click (“CPC”). IDF 155. The CPC for each advertiser is based on the outcome of a generalized second-price auction. 255 1-800 CONTACTS, INC. Opinion of the Commission Google, the leading search engine in the United States, receives more than eight out of every ten dollars spent on paid search advertising. IDF 137; Stip. at 5. Google’s paid search platform is called “AdWords.” When bidding on keywords in AdWords, advertisers may designate a keyword as “broad match,” “phrase match,” “exact match,” or “negative match.” Stip. at 6-9. When an advertiser designates a keyword as “broad match,” its ad may appear when a Google search contains the specific keyword, any of its plural forms, synonyms, or phrases similar to the word.4 When designated as “phrase match,” the ad may appear when a search contains the keyword with additional words before or after.5 And when designated as “exact match,” the ad may appear when a search contains the exact keyword and nothing more.6 In contrast, an advertiser may use “negative keywords” to ensure its ad does not appear when a user performs a search for a selected word or phrase.7 Similar to other keywords, negative keywords can be designated as broad match, phrase match, or exact match. See Stip. at 8-9. Generally, search engines do not currently restrict keywords available for bidding in advertising auctions.8 IDF 290, 298. In fact, it is common for companies to pay search engines to present their ads in response to a user’s search query of another company’s brand name. IDF 651-53. Before the agreements at issue in this case were in place, Google displayed ads for many of 1-800 Contacts’ retail competitors when those retailers bid on, and Google determined the ads were relevant to searches for 1-800 Contacts’ trademarks. IDF 653, 656. RX 0733 (Ghose Report) ¶51. Advertisers are not charged the amount they bid. Instead, the CPC is the bid amount needed to beat the rank of the advertiser in the next lower position. CX9019 (Juda Dep.) at 60, 137-38. 4 Broad match seeks to match within the meaning of the user’s search, rather than focusing on the text of any particular keyword. For example, if an advertiser purchases the keyword “low-carb diet plan” and selects broad match, Google may select that advertiser’s ad in response to searches for “carb-free foods” or “Mediterranean diets” even though the advertiser did not bid on those particular keywords. Stip. at 7. 5 For example, for the phrase match keyword “tennis shoes,” ads may be shown on searches for “red leather tennis shoes” or “buy tennis shoes on sale.” Stip. at 7-8.

6 For example, the exact match keyword “tennis shoes” may be matched to searches for “tennis shoes” but not for “red tennis shoes.” Stip. at 8.

7 For instance, “a retailer that sells eyeglasses may add the negative keyword ‘wine glasses’ to prevent its ads from showing in response to searches for that term.” ID at 25.

8 Prior to 2004, Google permitted a trademark owner to restrict use of its trademark by third parties as keywords and in the text of advertisements. In April 2004, Google changed its trademark policy to allow third parties to bid on trademarks as keywords, but still prohibited advertisers from using others’ trademarks in the text of their ads without authorization. When Google changed its policy, Google stated on its website that “Google is not in a position to arbitrate trademark disputes between advertisers and trademark owners.” Google encouraged “trademark owners to resolve their disputes directly with the advertisers.” IDF 293. Google further revised its trademark policy in 2009 and now allows advertisers to include another company’s trademark in the text of ads unless the trademark holder complains to Google. IDF 287, 290-91, 294; see also CX9022 (Charlston Dep.) at 16-17 (describing Google’s policy as “reactive”).

At the time Microsoft launched Bing in 2009, Microsoft did not permit advertisers to bid on keywords consisting of a trademark owned by a third-party. IDF 296. In 2011, Bing changed its policy and began permitting advertisers to bid on competitors’ trademarked keywords. IDF 298.

VOLUME 166 Opinion of the Commission Paid search advertising is an important method for marketing contact lenses online to obtain new customers and increase brand awareness. The paid search ad is presented to the consumer at a time when the consumer is more likely looking to buy. IDF 497-98, 500-03 (importance of paid search advertising to AC Lens), 523 (importance to LensDirect), 528-29 (importance to Lens Discounters), 532-33 (importance to Lenses for Less), 535, 538 (importance for Memorial Eye), 542-43 (importance for Vision Direct), 547-50 (importance for Walgreens), 553-54 (importance for Walmart), 557-58 (importance for Web Eye Care). In fact, many online retailers devote most of their advertising expenditures to search advertising. IDF 499 (search advertising accounts for 60-70 percent of AC Lens’ advertising expenditures), 521 (most of Visionworks contact lens marketing budget is spent on keyword search advertising), 522 (paid search advertising accounted for 85-90 percent of LensDirect’s marketing expenditures in 2016), 527 (online paid search advertising is the “main form of advertising” for Lens Discounters), 531 (Lenses for Less uses no forms of internet advertising other than search advertising), 534 (the “vast, vast, vast majority” of advertising spending for Memorial Eye was for online search advertising), 540-41 (Vision Direct spent more for search advertising than for any other type of advertising), 546 (most of Walgreens’ contact lens advertising budget was spent on paid search advertising), 552 (search advertising is the only type of online advertising for contact lenses used by Walmart), 555 (Web Eye Care only engages in online advertising); CX9014 (Batushansky Dep.) at 110 (approximately {75} percent of Web Eye Care’s online advertising expenditures are for search advertising).

In contrast to other online contact lens retailers, 1-800 Contacts also advertises heavily offline, including printed matter, radio, television, and other means. IDF 61-62. According to Respondent, the company has “made enormous investments” in building its brand and convincing consumers to buy contact lenses online rather than from brick-and-mortar retailers. RAB at 6; IDF 50-66. Between 2002 and 2014, 1-800 Contacts spent a total of {$186 million} on television advertising. IDF 64. Yet online advertising is still important to 1-800 Contacts. Between 2002 and 2014, it spent a total of {$86 million} on online advertising. IDF 65. In 2014, {37} percent of 1-800 Contacts’ advertising budget was spent on internet advertising, and between {29 and 53} percent of 1-800 Contacts’ internet advertising budget was spent on paid search advertising each year from 2004 through 2014. IDF 66. When 1-800 Contacts bids on its trademark keywords, it bids high enough to ensure that 1-800 Contacts’ sponsored ad is the first advertisement displayed in response to searches for its own trademark. IDF 575; CX9028 (Roundy Dep.) at 86; CX9031 (Schmidt Dep.) at 125-26.

D. 1-800 Contacts’ Conduct, Litigation, and the Settlement Agreements In 2002, 1-800 Contacts filed a complaint against Vision Direct alleging, inter alia, trademark infringement, claiming Vision Direct caused pop-up ads to appear when internet users visited the 1-800 Contacts website. The complaint did not include allegations regarding the use of 1-800 Contacts’ trademarks as keywords to trigger search engine advertisements.9 IDF 301. 1-800 Contacts filed a similar action challenging pop-up ads against Coastal Contacts in March 9 This lawsuit predated the change in Google’s trademark policy that allowed advertisers to bid on other companies’ trademarks as keywords. IDF 304.

257 1-800 CONTACTS, INC. Opinion of the Commission 2004. CX1615 (including trademark dilution claims). 1-800 Contacts resolved its disputes with Vision Direct and Coastal Contacts by executing settlement agreements that included terms related to pop-up advertising and the use of trademark keywords. IDF 306, 307 (Vision Direct settlement agreement, CX0311, included as prohibited acts “causing a Party’s website or Internet advertisement to appear in response to any Internet search for the other Party’s brand name, trademarks, or URLs”), 314, 315 (Coastal Contacts settlement agreement, CX0310, included as prohibited acts “causing a Party’s website or Internet advertisement to appear in response to any Internet search for the other Party’s brand name, trademarks or URLs but not through a search employing Generic or Descriptive terms”).

In addition to addressing pop-up ads, 1-800 Contacts monitored whether sponsored ads of its online competitors appeared on SERPs for queries involving the 1-800 Contacts trademarks. IDF 319-20, 322-23. Between 2005 and 2010, the company sent cease-and-desist letters to many of the online contact lens retailers whose ads appeared in the monitoring. IDF 325. The company later filed suit against several of these online retailers alleging federal trademark infringement and unfair competition under the Lanham Act §§ 32 and 43(a),10 trademark dilution, state and common law unfair competition, and unjust enrichment based on the retailers’ ads appearing on SERPs in response to searches for 1-800 Contacts’ trademark terms. IDF 328- 31. 1-800 Contacts filed suit against AC Lens (CX1623, Feb. 18, 2010), Contact Lens King (CX0461, Mar. 8, 2010), Empire Vision (CX0808, Feb. 25, 2010), EZ Contacts USA (CX1617, Dec. 6, 2007), Lensfast (CX1618, Dec. 23, 2008), Lenses for Less (CX0452, Jan. 20, 2010), Lens.com (CX1125, Aug. 13, 2007), LensWorld (CX1622, Jan. 8, 2008), Memorial Eye (RX0072, Dec. 23, 2008), Standard Optical (CX0965, July 13, 2010), Tram Data (CX0638, May 6, 2010), Walgreens (CX1620, June 8, 2010), and Web Eye Care (CX1621, Aug. 10, 2010). 1-800 Contacts settled most of the cases.11 In the suit against Lens.com, however, the case went to a judge. In December 2010, the U.S. District Court for the District of Utah issued an opinion granting summary judgment in favor of Lens.com on 1-800 Contacts’ trademark litigation claims. See 1-800 Contacts, Inc. v. Lens.com, Inc., 755 F. Supp. 2d 1151 (D. Utah 2010). The court found “insufficient evidence for a jury to conclude that Defendant infringed on Plaintiff’s mark for all advertisements that did not use Plaintiff’s mark in them.” Id. at 1181.12 10 Under federal trademark law, to succeed on a trademark infringement claim, a plaintiff must prove (1) that it has a protectable mark, and (2) that the defendant used the mark without the plaintiff’s consent in a manner that is likely to cause consumer confusion. See RX0734-0018 (Hogan Expert Report). “The most traditional form of trademark confusion is generally known as ‘source confusion,’ which is confusion as to the source of a good or service. . . . [C]ourts also recognize confusion as to affiliation, connection, or sponsorship. . . . because the Lanham Act . . . prohibits activity likely to cause ‘[t]he public’s belief that the mark’s owner sponsored or otherwise approved the use of the trademark[.]’” Id. at 0025. Courts have recognized that confusion is possible even if it does not occur at the point of sale. “Initial interest confusion . . . refers to the use of another’s trademark in a manner calculated to capture initial consumer attention.” Id. at 0027.

11 In litigation against LensWorld, the court entered a default judgment and an order that prohibited LensWorld from purchasing 1-800 Contacts’ federally registered trademarks as keywords for search engine advertising and required LensWorld to implement certain negative keywords where possible. IDF 337. 12 The court, in dictum, went on to discuss the propriety of a claimed oral agreement that assertedly required Lens.com to employ negative keywords to prevent its ads from appearing in response to searches for 1-800 VOLUME 166 Opinion of the Commission In July 2013, the Tenth Circuit upheld this portion of the district court judgment. See 1-800 Contacts, Inc. v. Lens.com, Inc., 722 F.3d 1229 (10th Cir. 2013). The appellate court, however, did not resolve the question of whether use of challenged trademark keywords, divorced from the text of the resulting ads, could result in a likelihood of confusion, because it found that 1-800 Contacts’ infringement claim “fail[ed] for lack of adequate evidence” of confusion. Id. at 1242- 43.

Between 2004 and 2013, 1-800 Contacts entered into thirteen settlement agreements (including the agreements with Vision Direct and Coastal Contacts) to resolve its trademark disputes. IDF 343; see CX0313 (2008 settlement agreement with EZ Contacts USA); CX0314; CX0316 (2009 settlement agreement with Vision Direct entered as a permanent injunction by the federal court); CX0315 (2009 settlement agreement with Lensfast); RX0028 (2010 settlement agreement with AC Lens); CX0319 (2010 settlement agreement with Empire Vision); CX0320 (2010 settlement agreement with Lenses for Less); CX0321 (2010 settlement agreement with Tram Data); CX0322 (2010 settlement agreement with Walgreens); CX0323 (2010 settlement agreement with Contact Lens King); CX0324 (2010 settlement agreement with Web Eye Care); RX0408 (2011 settlement agreement with Standard Optical); CX0326 (2013 settlement agreement with Memorial Eye). The settlement agreements include recitals that describe the litigation between the parties and state “the Parties have determined that, in order to avoid the expense, inconvenience, and disruption” of litigation, “it is desirable and in their respective best interests to terminate” the litigation and “settle any claims related thereto.” IDF 359. The settlement agreements release the parties of “any and all liability” arising from the claims and require dismissal of the litigation. IDF 360.

Although the language of the agreements varies, each includes provisions that prohibit the parties from using the other party’s trademarks, URLs, and variations of marks as search advertising keywords. IDF 361, 363. The settlement agreements also require the parties to employ “negative” keywords to prevent their ads from displaying whenever a search includes (or, as stated in some of the agreements, “contains”) the other party’s trademarks---even in situations when the advertiser did not bid on the other party’s actual trademark and the ad appears due to the search engine’s determination that the ad is relevant and useful to the consumer. IDF 364, 368. The agreements, however, do not specify whether negative keywords must be implemented under broad-match, phrase-match, or exact-match protocols. IDF 365.13 The settlement agreements do not prohibit parties from bidding on generic keywords such as “contacts” or “contacts lens,” so long as they employ negative keywords as required. IDF 366- 67.

Ten of the thirteen settlement agreements state that they do not prohibit the “use of the other Party’s trademarks on the Internet in a manner that would not constitute an infringing use Contacts’ trademark. “Were this actually an agreement entered into by the parties, the court questions whether it would survive an antitrust challenge. . . . A trademark right does not grant its owner the right to stamp out every competitor advertisement.” Id. at 1188.

13 Absent a negative keyword, in a broad or phrase match, a party that bids on the keyword “contacts” might find its ad displayed in response to a search for 1-800 Contacts.

259 1-800 CONTACTS, INC. Opinion of the Commission in a[] non-Internet context” (e.g., comparative advertising, parodies, and other non-infringing uses). See IDF 369.

1-800 Contacts enforced the settlements to prevent advertisements prohibited by the agreements from appearing. For instance, in April 2010, legal counsel for 1-800 Contacts wrote to AC Lens claiming AC Lens had breached the settlement agreement. IDF 372. Again, in 2014, 1-800 Contacts’ legal counsel notified AC Lens of another claimed breach. IDF 373. Legal counsel and 1-800 Contacts employees similarly contacted other online retailers to notify them that they were breaching the settlement agreements. See IDF 374-79 (communications with Coastal Contacts in 2006, 2011, and 2014), 380-82 (communications with Vision Direct in 2009, 2010, and 2013), 383-86 (communications with Walgreens in 2010 and 2014), 387 (communications with EZ Contacts in 2008), 388 (“Notice of Breach” sent by 1-800 Contacts’ counsel to Lensfast in 2014), 389-90 (letters from legal counsel sent to Contact Lens King in 2010 and 2014), 391 (legal counsel contacted Empire Vision in 2010), 392 (legal counsel sent letter to Lenses for Less in 2010).

In 2013, 1-800 Contacts entered into a sourcing and services agreement with Luxottica, a company that sells and distributes contact lenses through affiliates. IDF 86, 393; see CX0331 (Luxottica Sourcing and Services Agreement). Under that agreement, 1-800 Contacts provides fulfillment services by shipping contacts to Luxottica’s retail chain stores (e.g., LensCrafters, Pearle Vision, Sears Optical, and Target Optical). The sourcing and services agreement contains reciprocal advertising restrictions similar to those in the thirteen settlement agreements; it prohibits use of trademark keywords and requires exact-match negative keywords. IDF 396. E. Procedural History 1. The FTC’s Complaint In August 2016, the FTC issued an administrative Complaint against 1-800 Contacts, alleging that the thirteen settlement agreements and the sourcing agreement (collectively, the “Challenged Agreements”) and subsequent policing of the agreements unreasonably restrain both price competition in search advertising auctions and the availability of truthful, non-misleading advertising in violation of Section 5 of the FTC Act. Compl. ¶¶ 3, 25, 33. The Complaint alleges that the Challenged Agreements prevented the parties from disseminating ads that would have informed consumers that identical products were available at different prices, which reduced price competition among online contact lens retailers and made it costlier for consumers to search prices offered by the retailers. Compl. ¶ 31. As a result, the Complaint alleges, at least some consumers paid higher prices for contact lenses. Id.

The Complaint also alleges that Respondent’s conduct undermined the efficiency of search advertising auctions, distorted the prices in those auctions by eliminating bidders, and degraded the quality of service offered by search engines, including the quality of the SERP displayed to users. Id.

VOLUME 166 Opinion of the Commission 2. Complaint Counsel’s Motion for Partial Summary Decision Prior to the hearing before the ALJ, Complaint Counsel filed a motion for partial summary decision to dismiss the second and third defenses asserted in Respondent’s Answer. Respondent’s second defense contended that the Complaint is barred because the trademark lawsuits underlying the settlement agreements had not been alleged and shown to be objectively and subjectively unreasonable. The third defense argued that 1-800 Contacts’ conduct is protected under the Noerr-Pennington doctrine and the First Amendment. On February 1, 2017, the Commission granted the motion for partial summary decision. See 1-800 Contacts, Inc., Docket No. 9372, Commission Opinion and Order Granting Motion for Partial Summary Decision (Feb. 1, 2017). The Commission found that Noerr is not a defense because the Complaint challenges only private agreements. The Commission also found that the objective or subjective reasonableness of the trademark disputes is not an affirmative defense, even if the nature of the disputes may inform the antitrust analysis. 3. The Initial Decision Chief Administrative Law Judge Chappell issued a 214-page Initial Decision and Order on October 20, 2017, finding the Challenged Agreements violated Section 5 of the FTC Act. ID at 7, 138, 166, 190, 200. At the outset, the ALJ rejected 1-800 Contacts’ assertion that, under FTC v. Actavis, the trademark settlement agreements should not be subject to antitrust scrutiny. The ALJ found that trademark settlements are not antitrust immune. Id. at 7, 120-22. When considering liability, the ALJ applied Sherman Act Section 1 principles. To begin, he found there was “no dispute in this case that there was a contract, combination, or conspiracy” because 1-800 Contacts entered into fourteen agreements with online competitors. Id. at 118. Applying the rule of reason, the ALJ found that the relevant market in which to analyze the agreements’ effects was “the online sale of contact lenses in the United States,” id. at 138, 200, and that Complaint Counsel had met its burden of showing anticompetitive effects in that market. Id. at 7, 190, 200.

Specifically, the ALJ ruled that Complaint Counsel had established actual anticompetitive effects with harm to consumers and competition. Id. He explained that the advertising restrictions imposed by the Challenged Agreements harmed consumers by reducing the availability of information, which made it costlier for consumers to find and compare options for buying contact lenses online. He concluded that the reduced advertising “more likely than not resulted in consumers purchasing from 1-800 Contacts at higher prices than they would have paid to lower-priced competitors.” Id. at 155-56.

The ALJ stated that, because Complaint Counsel had proven that the challenged agreements resulted in harm to consumers and competition, his Initial Decision need not, and did not, determine whether 1-800 Contacts’ motives were anticompetitive. Id. at 139. In addition, although the Complaint alleged that “[a]s horizontal agreements that restrain price competition and restrain truthful non-misleading advertising, the Bidding Agreements are inherently suspect,” Compl. ¶ 32, the ALJ did not address this allegation. ID at 138-39. The ALJ also 261 1-800 CONTACTS, INC. Opinion of the Commission concluded that, having found liability under one theory (harm to consumers), he did not need to consider the other theory of alleged harm, based on injury to search engines. Id. at 166. After finding anticompetitive effects, the ALJ considered and rejected Respondent’s asserted procompetitive justifications. He concluded that, even if the settlement agreements reduced litigation costs and were favored by public policy, Respondent failed to proffer any consumer benefits flowing from the reduced litigation costs. Id. at 167-69. The ALJ also rejected Respondent’s justifications related to trademark law. Id. at 169-84. According to the ALJ, even if protecting trademarks and thereby encouraging investment in a brand name is a procompetitive goal, Respondent improperly assumes the fact of infringement. The ALJ found that Respondent failed to provide legal support for its assertion that merely displaying an ad in response to a search query for a trademark term is “likely to confuse” consumers about source or affiliation, regardless of the text of the ad. Id. at 170. The ALJ also found that Respondent failed to support the conclusion that the appearance of an online ad in response to a trademark search due to broad matching an advertiser’s bids on generic keywords (i.e., the failure to identify trademark terms as negative keywords) is a trademark “use.” Id. Having found liability, the ALJ issued an order that bars 1-800 Contacts from agreeing with any marketer or seller of contact-lens products to prohibit or limit participation in search advertising auctions (including prohibiting or restricting the use of keywords or requiring the use of negative keywords) or to prohibit or limit search advertising. ID at 203. The ALJ’s order contains a carve-out clause regarding future litigation; the carve-out establishes that the order does not prohibit Respondent from initiating or prosecuting a lawsuit or implementing or enforcing the order entered by any court of law, including an order approving a litigation settlement. Id. The ALJ’s order also requires Respondent to cease enforcing existing agreements that are inconsistent with the terms of the order’s prohibitions. Id. at 204. II. STANDARD OF REVIEW The Commission reviews the ALJ’s findings of fact and conclusions of law de novo, considering “such parts of the record as are cited or as may be necessary to resolve the issues presented.” 16 C.F.R. § 3.54(a). The Commission may “exercise all the powers which it could have exercised if it had made the initial decision.” Id. The de novo standard of review applies to both findings of fact and inferences drawn from those facts. See Realcomp II, Ltd., 2007 WL 6936319, at *16 n.11 (F.T.C. Oct. 30, 2009), aff’d, 635 F.3d 815 (6th Cir. 2011). We adopt the ALJ’s findings of fact to the extent that they are not inconsistent with this opinion. III. JURISDICTION Respondent does not dispute that the Commission has jurisdiction over it and over the conduct challenged in the Complaint. Section 5 of the Federal Trade Commission Act grants the Commission authority to prevent “unfair methods of competition in or affecting commerce” by “persons, partnerships, or corporations,” 15 U.S.C. § 45(a)(1)-(2). 1-800 Contacts is a corporation as “corporation” is defined in Section 4 of the FTC Act, 15 U.S.C. § 44, over which the Commission has jurisdiction. See Joint Stipulations of Jurisdiction, Law, and Fact, JX0001 ¶ 2. The acts and practices of 1-800 Contacts at issue, including the agreements being challenged, VOLUME 166 Opinion of the Commission are in commerce or affect commerce as “commerce” is defined in Section 4 of the FTC Act. IDF 3; Joint Stipulations of Jurisdiction, Law, and Fact, JX0001 ¶ 3; Answer ¶ 6. IV. 1-800 CONTACTS’ SETTLEMENTS ARE NOT IMMUNE FROM ANTITRUST SCRUTINY A. Actavis Does Not Immunize Commonplace Settlement Agreements or Settlements within the Scope of Potential Judicial Relief Respondent contends the settlement agreements between 1-800 Contacts and thirteen rival online sellers of contact lenses are not subject to antitrust scrutiny.14 Respondent asserts that Actavis stands for the proposition that there can be no antitrust challenge to a settlement agreement that is commonplace in form. Here, Respondent claims its settlements of trademark litigation took the form of common, non-use agreements. According to Respondent, Actavis exempted commonplace forms of settlement from antitrust scrutiny and held that “a party challenging a settlement must show that the settlement’s form is unusual.” RAB at 10 (internal quotation marks omitted). Respondent, however, reads Actavis much too broadly; the Court created no such shield from antitrust review.

As support for its argument, Respondent quotes the following sentence fragment in Actavis: “commonplace forms have not been thought for that reason alone subject to antitrust liability.” RAB at 3 (quoting Actavis, 133 S. Ct. at 2233). The Court’s wording is much more limited than Respondent suggests. The Supreme Court presented two examples of settlements: (1) where “Company A sues Company B for patent infringement and demands, say $100 million in damages” and receives “some amount less than the full demand as part of the settlement – $40 million, for example”; and (2) where “B has a counterclaim for damages against A” and “the original infringement plaintiff, A … end[s] up paying B to settle B’s counterclaim.” Actavis, 570 U.S. at 151-52. The Court then explained: “Insofar as the dissent urges that settlements taking these commonplace forms have not been thought for that reason alone subject to antitrust liability, we agree, and do not intend to alter that understanding.” Id. at 152 (emphasis added). The Court did not state a general rule that removes settlement agreements from antitrust scrutiny, but rather characterized two specific types of settlements as commonplace, and made it clear that the form of the settlement alone is not what subjects an agreement to antitrust scrutiny. Other portions of Actavis confirm this conclusion. Specifically, Actavis favorably cites three precedents that found antitrust liability for patent-related settlement agreements. First the Court relied on United States v. Singer Mfg. Co., 374 U.S. 174 (1963). Actavis characterized the Singer Court as “emphasizing that the Sherman Act ‘imposes strict limitations on the concerted activities in which patent owners may lawfully engage’ . . . it held that the agreements, although settling patent disputes, violated the antitrust laws.” Actavis, 570 U.S. at 149 (quoting and citing Singer, 374 U.S. at 195, 197). Actavis also discussed United States v. New Wrinkle, Inc., 342 U.S. 371 (1952), and its holding that the settlement agreement between two patentees did not confer antitrust immunity: “Far from it, the agreement was found to violate the Sherman Act.” 14 Respondent’s arguments about immunity for its settlement agreements, of course, offer no shelter for its Source and Services Agreement with Luxottica. That Agreement is not a settlement agreement. 263 1-800 CONTACTS, INC. Opinion of the Commission Actavis, 570 U.S. at 150 (citing New Wrinkle, 342 U.S. at 380). Finally, the Actavis Court noted that Standard Oil Co. v. United States, 283 U.S. 163 (1931) warned that the settlement agreements among patentees would have violated the Sherman Act “had the patent holders … ‘dominate[d]’ the industry and ‘curtail[ed] the manufacture and supply of an unpatented product.’” Actavis, 570 U.S. at 150-51 (quoting Standard Oil, 283 U.S. at 174). The Actavis Court stated these three cases sought “to accommodate patent and antitrust policies, finding challenged terms and conditions unlawful unless patent law policy offsets the antitrust law policy strongly favoring competition.” Id. at 151 (emphasis added). There is no hint that the Court was departing from precedent and implementing a new standard limiting antitrust liability to “commonplace” forms of settlement agreements.

In any case, the challenged settlements are in fact unusual. Respondent directs us to consider the “form” of the settlements, not their substance. Thus, Respondent describes each settlement as “a standard, non-use agreement whereby a party agreed not to use another’s trademark,” a form that practicing lawyers allegedly recognize as regularly used to settle trademark litigation. RAB at 11, 13. Antitrust law, however, “has consistently prioritized substance over form.” See In re Loestrin 24 Fe Antitrust Litig., 814 F.3d 538, 550 (1st Cir. 2016) (citing, inter alia, American Needle, Inc. v. Natl Football League, 560 U.S. 183, 191–92 (2010); Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 760 (1984) (“[The Sherman Act] is aimed at substance rather than form.”)).

When we consider the substance of these settlement agreements, we find they are unusual. Trademark litigation typically seeks to bar the use on the infringer’s labels, ads, or other promotional materials of the plaintiff’s trademark or a similar mark in a way likely to confuse consumers. Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50 (2d Cir. 1997), cited repeatedly by Respondent, provides a classic example where Clorox’s PINE-SOL products allegedly confused consumers of Sterling Winthrop’s LYSOL products. The settlement agreement upheld by the court restricted Clorox’s ability to market products as disinfectants or as special purpose cleansers under the PINE-SOL mark. Id. at 54. There the agreement did “no more than regulate how the name PINE-SOL may be used” in direct competition with LYSOL and did not restrict Clorox or other firms15 from selling products that compete with LYSOL under a brand name other than PINE-SOL. Id. at 57. It therefore raised none of the competition concerns attached to agreements that divide markets. Id. at 55. Given this limited restraint upon one competitor among many, the court concluded that Clorox had not shown that the agreement significantly restricted Clorox, or restricted at all any of the other large potential entrants, from competing. Id. at 59.

Here, as discussed below, the settlement agreements effectively shut off an entire—and very important—channel of advertising triggered by an alleged use of the trademark in the generation of search advertising. Stated differently, each settlement reaches farther than a cure based on rewording a label or an ad—effectively eliminating an entire channel of competitive advertising at the key moment when the consumer is considering a purchase. Furthermore, 1- 800 Contacts systematically applied similar restrictions to rival after rival that sought to 15 The court emphasized that Clorox had “presented no evidence” that other firms could not enter. Id. at 58. VOLUME 166 Opinion of the Commission challenge its position. And, contrary to Clorox’s premise, the agreements did achieve a market division through their reciprocal prohibitions on bidding in specific search auctions. Thus, from the perspective of substance, the settlement agreements between 1-800 Contacts and its thirteen rivals were indeed unusual.

Respondent, however, argues that under Actavis, settlement agreements that provide the same relief a court could have ordered are commonplace vis-à-vis the asserted trademark rights and immune from antitrust scrutiny. Respondent asserts that the challenged settlement agreements merely provide relief a court could have ordered if 1-800 Contacts had prevailed. RAB at 12-13. Respondent identifies no statement in Actavis of their asserted rule and no court opinion supporting the assertion or explaining why the scope of plenary powers of courts should determine the allowable extent of private agreements. In none of the cases addressed above did the Court, while engaging in its antitrust analysis of intellectual property-related settlement agreements, ask whether the agreement provided relief that a court could have ordered. See Actavis; Singer Mfg.; New Wrinkle; Standard Oil. A court’s plenary authority is irrelevant to whether private parties may agree to restrict competition, and private parties cannot rely on a court’s remedial authority to shield their agreements from antitrust scrutiny. See infra Section V.A.3.a.i.

Respondent appears to argue that because a prohibition on use of a trademark is within the exclusionary potential of the trademark (and therefore is a remedy that a court could have ordered), a settlement requiring non-use is immune from antitrust condemnation. See RRB at 4. But the crux of the Actavis decision was that there could be antitrust liability for a settlement of non-sham litigation with anticompetitive effects within the scope of the patent’s exclusionary potential. The Actavis majority could not have been clearer: Solvay’s patent, if valid and infringed, might have permitted it to charge drug prices sufficient to recoup the reverse settlement payments it agreed to make to its potential generic competitors. And we are willing to take this fact as evidence that the agreement’s anticompetitive effects fall within the scope of the exclusionary potential of the patent. But we do not agree that that fact, or characterization, can immunize the agreement from antitrust attack. Actavis, 570 U.S. at 147 (internal quotation and citation omitted). Here, even assuming arguendo that the settlement agreements’ effects were within the scope of Respondent’s enforceable trademark rights16—and hence within the scope of relief that a court might have ordered, Actavis stands for the possibility of antitrust liability, not for the foreclosure of antitrust review. As Actavis explains, we need to consider both antitrust and intellectual property policies. See id. at 148 (“it would be incongruous to determine antitrust legality by measuring the settlement’s anticompetitive effects solely against patent law policy, rather than by measuring them against procompetitive antitrust policies as well”). Respondent’s rule looks only to half of the equation, i.e., trademark policies, and does not withstand a thorough understanding of Actavis.

16 But cf. infra Section V.B.1.b (discussing Respondent’s assertions regarding trademark rights). 265 1-800 CONTACTS, INC. Opinion of the Commission B. The “Actavis Considerations”

Respondent argues that even if the Commission finds the challenged settlements were unusual, dismissal still would be appropriate because Complaint Counsel did not prove any of the five “Actavis considerations” that, taken together, could outweigh the desirability of settlements, to favor antitrust scrutiny. The Actavis Court identified five factors that convinced it to give the FTC an opportunity to prove its antitrust claim: (1) the specific restraint’s potential for genuine adverse effects on competition; (2) the potential that the anticompetitive consequences will sometimes prove unjustified; (3) the likelihood that the patentee possesses the power to bring about unjustified competitive harm in practice; (4) the administrative feasibility of an antitrust action; and (5) the risk that finding antitrust liability for a particular form of settlement would prevent litigants from settling (i.e., the litigants’ ability to settle in other ways that do not harm competition). 570 U.S. at 153-58. Respondent treats these factors as threshold requirements for conducting antitrust review and argues that the ALJ erred by ignoring these considerations. RAB at 16. We disagree that Actavis requires this five-factor test to be applied to antitrust review of all settlements of intellectual property litigation. Moreover, even if the Court had created such a requirement, the litigation in this case would pass. But the Court did not characterize these considerations as prerequisites for antitrust review of all intellectual property-related settlement or as defining the content of their analysis under the rule of reason. Rather, the Court described the factors as considerations relevant to the particular antitrust claim before it:

We recognize the value of settlements and the patent litigation problem. But we nonetheless conclude that this patent-related factor should not determine the result here. Rather, five sets of considerations lead us to conclude that the FTC should have been given the opportunity to prove its antitrust claim. Actavis, 570 U.S. at 153. Interpreting these considerations as requirements applicable to all settlements (or even all settlements of intellectual property disputes) risks straight-jacketing the analysis within bounds that were intended to address a particular case. Cf. Loestrin 24, 814 F.3d at 551 n.12 (“the five [Actavis] considerations should not overhaul the rule of reason, nor should they create a new five-part framework in antitrust cases”). Regardless of whether the Court intended to create a new litmus test, many of the same considerations are present in this case, and they favor proceeding with the antitrust inquiry. As Respondent suggests, RAB at 16, the first three factors all relate to whether a challenged settlement poses a significant risk of unjustified anticompetitive harm. Sections V.A.1, V.A.3.b, and V.C below explain at length that the restraints at issue bear considerable potential for unjustified competitive harm by limiting truthful advertising, increasing prices paid for contact lenses, and impeding search auction bidding. The remaining two Actavis considerations also support antitrust review. Our analysis below demonstrates the administrative feasibility of this inquiry: antitrust liability can be found without the need to relitigate trademark infringement issues in situations such as this, where the challenged restraints are not reasonably necessary to achieve procompetitive benefits. For all the reasons stated above, we conclude that 1-800 Contacts’ settlements are not immune from antitrust scrutiny. VOLUME 166 Opinion of the Commission V. ANTITRUST ANALYSIS OF THE CHALLENGED AGREEMENTS The Complaint alleges that the series of agreements between 1-800 Contacts and numerous online sellers of contact lenses are agreements to restrain competition in violation of Section 5 of the FTC Act and constitute unfair methods of competition in or affecting commerce in violation of Section 5 of the FTC Act. Compl. ¶¶ 33-34. To assess whether the Challenged Agreements violate Section 5 of the FTC Act, we are guided by case law concerning Section 1 of the Sherman Act.17 Under Section 1 of the Sherman Act,18 except for a small group of restraints that are per se unlawful because they “always or almost always tend to restrict competition,” restraints are evaluated under the rule of reason. See, e.g., Ohio v. American Express Co., 138 S. Ct. 2274, 2284 (2018) (quoting Bus. Elecs. Corp. v. Sharp Elecs. Corp., 485 U.S. 717, 723 (1988)). When applying the rule of reason, courts rely on a burden-shifting framework. Under this framework, the plaintiff has the burden to prove that the challenged restraint has, or is likely to have, a substantial anticompetitive effect that harms consumers. If the plaintiff meets its initial burden, the burden shifts to the defendant to show a procompetitive rationale for the restraint. If the defendant makes this showing, then the plaintiff must show that the procompetitive justification could be reasonably achieved through less anticompetitive means or that the anticompetitive harms outweigh the procompetitive benefits. See, e.g., id. at 2284; Geneva Pharm. Tech. Corp. v. Barr Labs., Inc., 386 F.3d 485, 507 (2d Cir. 2004); accord Polygram Holding, Inc., 136 F.T.C. 310, 349-50 (2003), aff’d, 416 F.3d 29 (D.C. Cir. 2005). When operationalizing this framework, the sequence for evaluating particular evidence may vary under a particular structured analysis, but the ultimate burdens remain unchanged.

In Polygram, the Commission traced the Supreme Court’s development of the rule of reason. 136 F.T.C. at 325-44. One feature of the Court’s jurisprudence is that the rule of reason calls for “an enquiry meet for the case, looking to the circumstances, details, and logic of a restraint,” with a goal to reach “a confident conclusion about the principal tendency of a restriction.” Realcomp, 2007 WL 6936319, at *18 (quoting California Dental, 526 U.S. at 781). Cases such as “BMI, NCAA, and IFD indicate[] that the evaluation of horizontal restraints takes place along an analytical continuum in which a challenged practice is examined in the detail necessary to understand its competitive effect.” Polygram, 136 F.T.C. at 336 (citing Broadcast Music, Inc. v. Columbia Broadcasting Sys., Inc., 441 U.S. 1 (1979); Natl Collegiate Athletic Assn v. Bd. of Regents of the Univ. of Oklahoma, 468 U.S. 85 (1984); FTC v. Indiana Fed’n of Dentists, 476 U.S. 447 (1986) (“IFD”)).

17 The Commission’s authority under Section 5 of the FTC Act extends to conduct that violates the Sherman Act. See, e.g., Actavis, 570 U.S. at 145; California Dental Assn v. FTC, 526 U.S. 756, 762 & n.3 (1999); FTC v. Motion Picture Adver. Serv. Co., 344 U.S. 392, 394-95 (1953), Fashion Originators’ Guild of Am., Inc. v. FTC, 312 U.S. 457, 463-64 & n.4 (1941). In the present case, our analysis under Section 5 is the same as it would be under Section 1 of the Sherman Act.

18 Violations of Section 1 of the Sherman Act require (1) a contract, combination, or conspiracy, that (2) unreasonably restrains trade. See, e.g., Realcomp II, Ltd. v. FTC, 635 F.3d 815, 824 (6th Cir. 2011). Here, the trademark litigation settlement agreements and the Luxottica Sourcing and Services Agreement easily satisfy the first element. ID at 117-18. Consequently, our analysis focuses on the second element. 267 1-800 CONTACTS, INC. Opinion of the Commission The Court has defined three separate but not entirely distinct ways for a plaintiff to show that a challenged restraint resulted in anticompetitive effects under a rule of reason analysis. First, the IFD Court observed that the particular horizontal restraint at issue, by its very nature established that anticompetitive effects were likely; it did not require “elaborate industry analysis . . . to demonstrate the anticompetitive character of such an agreement.” IFD, 476 U.S. at 459. In Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988), and Polygram, we labeled such restraints “inherently suspect.” Second, the Court in IFD held that, even if the restriction in question was “not sufficiently naked” to be considered inherently suspect based on the nature of the restraint, the plaintiff’s prima facie case was established, even without a detailed market analysis, because the record contained direct evidence of anticompetitive effects. IFD, 476 U.S. at 460. Third, the Court’s discussion made clear that the traditional mode of analysis—inquiring into market definition and market power to determine whether an arrangement has the potential for genuine adverse effects on competition—was also available. Id. Any one of these modes of analysis is sufficient to establish a prima facie case. In this case, we use two of these modes of analysis to assess whether 1-800 Contacts’ agreements resulted in anticompetitive effects: (1) we consider whether the Challenged Agreements are inherently suspect; and (2) we examine whether there is direct evidence of anticompetitive effects. Each mode of analysis provides an independent basis for finding that the Challenged Agreements have substantial anticompetitive effects and leads us to find liability. We explain the structure of the analysis based on the case law for these modes in the sections devoted to each. We also examine Complaint Counsel’s allegation that the Challenged Agreements have substantial anticompetitive effects on competition with respect to bidding on search terms, which again leads us to find a violation of Section 5 of the FTC Act. Although we discuss particular evidence that leads us to conclude that the restraints in the Challenged Agreements have substantial anticompetitive effects under different modes of analysis, our review of the evidence is not rigidly compartmentalized. For instance, evidence regarding the significance of search advertising generally and searches for 1-800 Contacts’ trademarks in particular, which is discussed as part of the inherently suspect analysis, informs our understanding of the direct evidence of anticompetitive effects. Although the two modes of analysis provide different structures, they reach the same conclusion. The restraints on advertising and bidding at advertising auctions imposed by 1-800 Contacts’ agreements have substantial anticompetitive effects and, unless reasonably necessary to achieve a valid procompetitive rationale, violate Section 5 of the FTC Act. A. Analysis of the Challenged Agreements for Effects on Consumers Under Polygram’s Inherently Suspect Framework In Polygram, we held that in a limited but significant category of cases, “the conduct at issue is inherently suspect owing to its likely tendency to suppress competition.” Polygram, 136 F.T.C. at 344. In these cases, “scrutiny of the restraint itself . . . without consideration of market power” is sufficient to condemn the restraint, unless the defendant can “articulate a legitimate justification” for that restraint. Id. at 344-45; see also California Dental Assn v. FTC, 526 U.S. 756, 770 (1999) (describing a “quick-look analysis” applicable when “an observer with even a VOLUME 166 Opinion of the Commission rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect on customers and markets”); IFD, 476 U.S. at 459 (finding “no elaborate industry analysis” was required to demonstrate the anticompetitive character of a “horizontal agreement among participating dentists to withhold from their customers a particular service that they desire”).

Drawing from the Supreme Court’s analysis in California Dental, 526 U.S. at 779, Polygram spelled out the structure of the “inherently suspect” analysis for the plaintiff’s demonstration that a restraint has anticompetitive effects. A plaintiff must demonstrate[] that the conduct at issue is inherently suspect owing to its likely tendency to suppress competition. . . . [T]he defendant can avoid summary condemnation only by advancing a legitimate justification for those practices. . . . When the defendant advances such cognizable and plausible justifications, the plaintiff must make a more detailed showing that the restraints at issue are indeed likely, in the particular context, to harm competition. Such a showing still need not prove actual anticompetitive effects or entail “the fullest market analysis.” Depending upon the circumstances of the cases and the degree to which antitrust tribunals have experience with restraints in particular markets, such a showing may or may not require evidence about the particular market at issue, but at a minimum must entail the identification of the theoretical basis for the alleged anticompetitive effects and a showing that the effects are indeed likely to be anticompetitive. Such a showing may, for example, be based on a more detailed analysis of economic learning about the likely competitive effects of a particular restraint, in markets with characteristics comparable to the one at issue. The plaintiff may also show that the proffered procompetitive effects could be achieved through means less restrictive of competition.

Polygram, 136 F.T.C. at 344-49 (quoting California Dental, 526 U.S. at 779) (citations omitted). On review, then Chief Judge Douglas Ginsburg, writing for the D.C. Circuit, “accept[ed] the Commission’s analytical framework.” Polygram, 416 F.3d at 36; see also North Texas Specialty Physicians v. FTC, 528 F.3d 346, 361 (5th Cir. 2008) (“the Commission’s articulation of the shifting burdens employed in its [inherently suspect] analysis appears, at least facially, to comport with the framework provided by the Supreme Court’s precedent”). 1. The Anticompetitive Nature of the Restraints Inherently suspect conduct “ordinarily encompasses behavior that past judicial experience and current economic learning have shown to warrant summary condemnation.” Polygram, 136 F.T.C. at 344-45. Consequently, our analysis considers whether there is a “close family resemblance between the suspect practice and another practice that already stands convicted in the court of consumer welfare.” Polygram, 416 F.3d at 37. The determination is based on the conduct’s “likely tendency to suppress competition.” Polygram, 136 F.T.C. at 344. “At this stage, the focus of the inquiry is on the nature of the restraint rather than on the market effects in a particular case.” North Texas Specialty Physicians, 140 F.T.C. 715, 733 (2005) (“NTSP”).

269 1-800 CONTACTS, INC. Opinion of the Commission We previously recognized that an inherently suspect analysis and a per se analysis are “close neighbors.” NTSP, 140 F.T.C. at 719. Consequently, the Commission previously condemned conduct as inherently suspect that approximates conduct that had otherwise been characterized as per se violations of the antitrust laws. For instance, in Polygram, we condemned as inherently suspect an agreement between record album distributors to suspend temporarily advertising and discounting of particular performers’ earlier concert albums. The D.C. Circuit agreed, explaining that “[a]n agreement between joint venturers to restrain price cutting and advertising with respect to products not part of the joint venture looks suspiciously like a naked price fixing agreement between competitors, which would ordinarily be condemned as per se unlawful. The Supreme Court has recognized time and again that agreements restraining autonomy in pricing and advertising impede the ‘ordinary give and take of the market place.’” Polygram, 416 F.3d at 37 (quoting IFD, 476 U.S. at 459). Similarly, in NTSP, we condemned as inherently suspect certain contracting practices of a physician trade association, while also recognizing that “NTSP’s activities could be characterized as per se illegal because they are closely analogous to conduct condemned per se in this and other industries . . . .” NTSP, 140 F.T.C. at 731.

In the present case, the agreements between 1-800 Contacts and its rivals prohibit each party from causing or allowing advertisements to appear in response to an internet search for the other party’s trademarks or URLs, or variations of the trademarks or URLs. Those agreement terms and 1-800 Contacts’ subsequent enforcement of them prevent the agreeing parties from offering advertising in response to an internet search for “1-800 Contacts” or similar queries. IDF 371. Thus, the Challenged Agreements are, in essence, agreements between horizontal competitors to restrict the information provided by advertising to consumers when they search for 1-800 Contacts’ trademark terms and URLs; consumers could have used that withheld information to compare and evaluate the prices and other features of competing online sellers. Ultimately, the effect of the advertising restrictions is to make information enabling consumer comparisons more difficult and costly to obtain.

Online search is one of the key methods by which consumers discover vendors and compare products and services. IDF 564. It is an important method by which lower-priced rivals compete with 1-800 Contacts. IDF 565. Rival online sellers generally offer lower prices than 1-800 Contacts, IDF 693, and much of the advertising for those retailers emphasizes those lower prices. See IDF 587, 591, 603, 611, 646, 703, 724; Holbrook, Tr. 1904. This is particularly important because the advertising is presented to a consumer at a time when the consumer is more likely to be looking to buy. IDF 498.

Economic theory indicates that restrictions on this type of advertising are likely to harm competition. A flow of information between buyers and sellers is an essential part of the market system. Buyers have to find out who they can buy from and on what terms, and sellers must let consumers know how to find them, what they have to offer, and on what terms. IDF 681 (citing CX8006 (Evans Expert Report) at 080 ¶ 178). Restrictions on advertising interfere with that flow of information and raise the cost to consumers of finding the most suitable offering of a product or service. CX8006 (Evans Expert Report) at 080-084; IDF 683. Faced with these higher search costs, consumers must either spend more time and money looking for a lower- VOLUME 166 Opinion of the Commission priced supplier or end their search because the cost of continued search exceeds the likelihood of finding a lower price. Ultimately, as a result of the reduced information flow, some consumers will pay higher prices for the particular good or service while others stop their search before they find a price that induces them to buy, which reduces the quantity sold. In addition, advertising restrictions “reduce[] sellers’ incentives to lower prices. One reason a restriction on advertising may reduce a seller’s incentives to lower prices is that, absent an ability to advertise, lower perunit prices may not be sufficiently offset by higher volume.” Polygram, 136 F.T.C. at 355 (citations omitted).

Empirical studies confirm the anticompetitive effects of advertising restrictions. Complaint Counsel’s expert, Dr. Evans explains, Economists have conducted more than 21 studies that assess the effect of advertising restrictions on prices and other aspects of competition. . . . Almost all of these studies19 find that advertising restrictions result in higher prices. Many of them show that the consumers are not getting higher quality products or services at those higher prices. At least one of the studies finds that the advertising restrictions tend to suppress entry.

CX8006 (Evans Expert Report) at 081-082. As Dr. Evans concludes: “There is a consensus in the economics literature that restrictions on advertising among rivals impair competition and harm consumers.” Id. at 081. Dr. Evans also confirmed that greater availability of pricing information affects the prices that consumers pay for products sold online. Id. at 084. Dr. Evans noted that prior empirical work found that consumers paid significantly less for life insurance plans and cars because online price comparison sites made price shopping much easier. Id. Dr. Evans also cited a study finding that dissemination of price information online made demand curves for online sellers much more elastic. Id. The Commission has prior experience with this literature; we cited many of these same empirical studies when we considered the economics of advertising restrictions at issue in Polygram. See Polygram, 136 F.T.C. at 355 n.52. Consistent with the economic literature, over the past 40 years, the Commission has repeatedly found that advertising restrictions harm competition and consumers. See Am. Med. Assn, 94 F.T.C. 701, 1010 (1979) (condemning an agreement among physicians not to advertise), aff’d, 638 F.2d 443 (2d Cir. 1980), aff’d by an equally divided Court, 455 U.S. 676 (1982) (per curiam); Mass. Board, 110 F.T.C. at 598 (condemning a licensing board’s ban on advertising). Among the more recent cases, in Polygram, we concluded that an agreement between music companies not to advertise two recordings for a short time period was inherently suspect. See Polygram, 136 F.T.C. at 353-58. Courts have similarly recognized the role of advertising in fostering competition and have condemned advertising restrictions. The Supreme Court explained that advertising “serves to inform the public of the availability, nature, and prices of products and services, and thus performs an indispensable role in the allocation of resources in a free enterprise system.” Bates v. State Bar of Ariz., 433 U.S. 350, 364 (1977). The Supreme Court further explained, “[I]t is clear as an economic matter that . . . restrictions on 19 Among the cited studies, there is only one, involving advertising for professional services, that shows lower prices when there is no advertising. See CX8006 (Evans Expert Report) at 081 n.186. 271 1-800 CONTACTS, INC. Opinion of the Commission fare advertising have the forbidden significant effect upon fares. . . . Restrictions on advertising ‘serve to increase the difficulty of discovering the lowest cost seller . . . and [reduce] the incentive to price competitively.’” Morales v. Trans World Airlines, Inc., 504 U.S. 374, 388 (1992) (quoting Bates, 433 U.S. at 377).

More recently, in California Dental, the Court found “unexceptionable” the Ninth Circuit’s “statements that ‘price advertising is fundamental to price competition’ and that ‘restrictions on the ability to advertise prices normally make it more difficult for consumers to find a lower price and for [sellers] to compete on the basis of price.’” California Dental, 526 U.S. at 773 (quoting California Dental Assn v. FTC, 128 F.3d 720, 727 (9th Cir. 1997)). The Court, however, found that the professional services market at issue permitted “the possibility that the particular restrictions on professional advertising could have different effects from those ‘normally’ found in the commercial world.” Id. Thus, even when the Court did not affirm liability in California Dental, it recognized that in ordinary commercial markets, bans on truthful advertising normally are likely to cause competitive harm. Courts have long condemned advertising restrictions. The D.C. Circuit affirmed our analysis in Polygram. Polygram Holding Inc. v. FTC, 416 F.3d 29, 37 (D.C. Cir. 2005) (“we have no difficulty with the Commission’s conclusion . . . An agreement between joint venturers to restrain price cutting and advertising with respect to products not part of the joint venture looks suspiciously like a naked price fixing agreement between competitors”). Other advertising restrictions have similarly been condemned. See, e.g., Blackburn v. Sweeney, 53 F.3d 825 (7th Cir. 1995) (finding advertising restraint that prohibited attorneys from advertising in particular geographical regions per se unlawful); United States v. Gasoline Retailers Assn, Inc., 285 F.2d 688 (7th Cir. 1961) (agreement between trade association and gasoline station operators that stations would not advertise—including by posting signs at the stations showing prices—or give premiums was per se Sherman Act violation).

Our conclusion that the particular advertising restrictions imposed by the Challenged Agreements are inherently suspect is a limited finding. We do not contend that all advertising restrictions are necessarily inherently suspect. The restrictions in this particular case prohibit the display of ads that would enable consumers to learn about alternative sellers of contact lenses and give them the opportunity to make price comparisons at the time they are likely to make a purchase. Importantly, the restrictions at issue here are not limitations on the content of an advertisement a consumer would otherwise see; they are restrictions on a consumer’s opportunity to see a competitor’s ad in the first place. Moreover, the record shows that the suppressed ads often emphasize lower prices. In this context, we find the advertising restrictions are inherently suspect. Because the Challenged Agreements restrict the ability of lower cost online sellers to show their ads to consumers, it is easy to see how “an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect on customers and markets.” California Dental, 526 U.S. at 770. VOLUME 166 Opinion of the Commission 2. Preliminary Analysis of Respondent’s Justifications That conclusion is not the end of the analysis. As we explained in Polygram, If the challenged restrictions are . . . inherently suspect, then the defendant can . . . advanc[e] a legitimate justification for those practices. . . . At this early stage of the analysis, the defendant need only articulate a legitimate justification. . . . [T]he proffered justifications must be both cognizable under the antitrust laws and at least facially plausible. . . . When the defendant advances such cognizable and plausible justifications, the plaintiff must make a more detailed showing that the restraints at issue are indeed likely, in the particular context, to harm competition. Polygram, 136 F.T.C. at 345-48. Moreover, Respondent bears the burden of “articulat[ing] the specific link between the challenged restraint and the purported justification.” Polygram, 136 F.T.C. at 347. In this case, Respondent must articulate the specific link between restraints on its competitors’ use of search advertising and the protection of its own trademark rights. “[C]ognizability allows the deciding tribunal to reject proffered justifications that, as a matter of law, are incompatible with the goal of antitrust law to further competition. Cognizable justifications ordinarily explain how specific restrictions enable the defendants to increase output or improve product quality, service, or innovation.” Id. at 345-46. “A justification is plausible if it cannot be rejected without extensive factual inquiry. The defendant . . . must articulate the specific link between the challenged restraint and the purported justification to merit a more searching inquiry into whether the restraint may advance procompetitive goals . . . .” Id. at 347.20 Here, Respondent has articulated two legitimate justifications that are cognizable and, at least, facially plausible: avoidance of litigation costs through settlement and trademark protection.21 Settling costly litigation is a cognizable and facially plausible justification for the settlement agreements. As the Supreme Court explained in Actavis and we recognized in Schering-Plough, 136 F.T.C. 956, 1003 (2003), there is a “general legal policy favoring settlement of disputes.” Actavis, 570 U.S. at 153; see also In re Tamoxifen Citrate Antitrust Litig., 466 F.3d 187, 202 (2d Cir. 2006) (noting public’s “strong interest in settlement” of complex and expensive cases). While this public policy favoring settlements does not create antitrust immunity, see supra Section IV, it is, nonetheless, a legitimate justification that we do not ignore.

Settling lawsuits is generally economically efficient. IDF 355 (citing RX0739 (Murphy Expert Report) at 0053; CX9042 (Evans Dep.) at 196). Avoiding unnecessary expenses is consistent with competition principles. The record shows that these concerns motivated 1-800 Contacts’ rivals to settle. See IDF 349 (knowing that Lens.com had spent $2 million and its 20 Respondent apparently concedes that it bears the burden of showing that its justification is cognizable. RAB at 39 (stating “even if 1-800 Contacts had the burden to do more than prove that its claims were cognizable . . . .”). 21 A third purported benefit—avoidance of consumer confusion—is subsumed among the benefits of trademark protection.

273 1-800 CONTACTS, INC. Opinion of the Commission litigation was not yet concluded, Memorial Eye settled its case because of the cost of the litigation and the legal uncertainty), 352 (AC Lens made a business decision to settle in light of potential costs and protracted nature of the litigation), 353 (Web Eye Care settled because the costs of litigation were “way more than we wanted to spend” and “not worth it” and because of the risks of losing the litigation), 354 (Empire Vision settled to avoid the litigation expense). But cf. infra Section V.A.5.a (noting the absence of evidence linking litigation cost savings in this case to benefits to consumers).

Similarly, at this stage of the analysis, we consider protecting trademark rights to be a legitimate procompetitive justification. As Respondent’s experts, Drs. Landes and Murphy, explained, trademarks provide informational benefits to consumers about product and quality attributes that reduce consumers’ search costs. Trademark protection preserves those quality signals for consumers and encourages firms to invest in both product quality and the trademark. See RX0737 (Landes Expert Report) at 0005-0014, RX0739 (Murphy Expert Report) at 0032- 0035. Also, at least facially, Respondent’s contention that the settlement agreements advance this procompetitive goal is plausible; “it cannot be rejected without extensive factual inquiry.” Polygram, 136 F.T.C. at 347. The trademark litigation underlying the settlement agreements was not sham. Lens.com, Inc. v. 1-800 Contacts, Inc., 2014 WL 12596493 (D. Utah Mar. 3, 2014); IDF 340 (District court dismissed Memorial Eye’s counterclaim alleging the suit filed by 1-800 Contacts was sham litigation). Also, the record shows that 1-800 Contacts had a brand identity that it wished to preserve. It had a marketing strategy to create brand awareness and during the period 2002 through 2014 had spent {$186 million} on television advertising and {$86 million} on internet advertising to build that brand. IDF 60, 64-65. It is important to note that our determination that two of 1-800 Contacts’ procompetitive justifications are legitimate at this stage of the analysis is not the end of our evaluation. We return to Respondent’s procompetitive justifications with an “extensive factual [and legal] inquiry” when we move farther into the rule of reason analysis. In Sections V.A.3.a and V.A.5, we consider Complaint Counsel’s contention that the procompetitive benefits could be reasonably achieved through less anticompetitive means and examine whether Respondent’s procompetitive rationales are supported by the facts.22 3. Complaint Counsel’s More Detailed Showing Because Respondents have advanced legitimate procompetitive justifications, we do not summarily condemn the Challenged Agreements based only on an initial review of the nature of the restraints. Instead, to satisfy their burden under the rule of reason, Complaint Counsel must make a further showing. As we explained in Polygram, 22 We recognize the current limited inquiry regarding 1-800 Contacts’ procompetitive justifications and the later steps in the rule of reason burden-shifting analysis “could be combined, [but] we think it analytically superior and consistent with the relevant case law to first screen the purported justification for legitimacy before engaging in a more extensive, and therefore longer and more resource-intensive, inquiry whether detailed analysis supports or refutes the justification. Antitrust courts have long held that preliminary analysis of purported justifications is appropriate.” Polygram, 136 F.T.C. at 348 n.43.

VOLUME 166 Opinion of the Commission When the defendant advances such cognizable and plausible justifications, the plaintiff must make a more detailed showing that the restraints at issue are indeed likely, in the particular context, to harm competition. Such a showing still need not prove actual anticompetitive effects or entail “the fullest market analysis.” Depending upon the circumstances of the cases and the degree to which antitrust tribunals have experience with restraints in particular markets, such a showing may or may not require evidence about the particular market at issue, but at a minimum must entail the identification of the theoretical basis for the alleged anticompetitive effects and a showing that the effects are indeed likely to be anticompetitive. . . . The plaintiff may also show that the proffered procompetitive effects could be achieved through means less restrictive of competition. Polygram, 136 F.T.C. at 348-49 (footnote and citations omitted); see also Actavis, 570 U.S. at 159 (explaining that the showing does not require that “the Commission . . . litigate the patent’s validity, . . . present every possible supporting fact or refute every possible pro-defense theory. . . . ‘[t]here is always something of a sliding scale in appraising reasonableness,’ and as such ‘the quality of proof required should vary with the circumstances.’”) (quoting California Dental, 526 U.S. at 780 and 7 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law ¶ 1507 (1986)). In short, Complaint Counsel may meet their burden to show that the restraints at issue are likely to harm competition by either (i) identifying the theoretical basis for the alleged anticompetitive effects and showing that these effects are likely in this particular setting or (ii) explaining how Respondent could have minimized the anticompetitive effects of its conduct or accomplished its procompetitive justifications through less restrictive alternatives. Here, Complaint Counsel show both that, in the context of online sales of contact lenses, the proffered procompetitive effects of the advertising restraints in the Challenged Agreements could be achieved through means less restrictive of competition, and that restraints “are indeed likely . . . to harm competition,” Polygram, 136 F.T.C. at 348. We address each of these approaches separately.

a. Respondent’s Proffered Procompetitive Justifications Could Be Achieved Through Less Anticompetitive Means First, Complaint Counsel can rebut Respondent’s showing that litigation cost savings and trademark protection are cognizable and plausible procompetitive justifications by establishing that “the proffered procompetitive effects could be achieved through means less restrictive of competition.” Polygram, 136 F.T.C. at 349; see, e.g., American Express, 138 S. Ct. at 2284 (if defendant successfully shows a procompetitive justification, “then the burden shifts back to the plaintiff to demonstrate that the procompetitive efficiencies could be reasonably achieved through less anticompetitive means”); Law v. National Collegiate Athletic Assn, 134 F.3d 1010, 1019 (10th Cir. 1998) (plaintiff may demonstrate that the challenged conduct is not reasonably necessary or could be achieved by less restrictive means); 7 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law ¶ 1505 (4th ed. 2017). The challenged conduct is not reasonably necessary if the parties could have achieved similar efficiencies by practical, significantly less restrictive alternatives. See United States v. Brown Univ., 5 F.3d 658, 678–79 (3d Cir. 1993); 275 1-800 CONTACTS, INC. Opinion of the Commission United States v. Visa U.S.A., Inc., 344 F.3d 229, 238 (2d Cir. 2003); 7 Areeda & Hovenkamp, supra, ¶ 1505; FTC & U.S. Dept of Justice, Antitrust Guidelines for Collaborations among Competitors § 3.36(b).

Complaint Counsel argue that the Challenged Agreements are not reasonably necessary to protect 1-800 Contacts’ trademarks, but rather are unreasonably overbroad, prohibiting a wide range of truthful, non-confusing advertising; according to Complaint Counsel, the asserted procompetitive benefits could be achieved through less restrictive means. CCB at 4. Respondent disagrees. It contends that the Challenged Agreements are not overbroad and maintains that none of Complaint Counsel’s alternatives is workable. i. Overbreadth When an agreement limits truthful price advertising on the basis of trademark protection, it must be narrowly tailored to protecting the asserted trademark right. The agreements here are not—they restrict advertising regardless of whether the ads are likely to be confusing and, apparently, regardless of whether competitors actually use the trademark term (requiring negative keywords).

Respondent and the Dissent argue that the Challenged Agreements cannot be overbroad because their restrictions are similar to what a court could have ordered. RRB at 4. The fact that a court has authority to enter an order of non-use, however, does not support a finding that it is always a permissible restraint when implemented by private parties. As we have already discussed, see supra Section IV.A, a court’s plenary authority to issue relief is irrelevant to the question of whether private parties may, consistent with the antitrust laws, agree to restrict their competition. Courts have broad injunctive authority, and Respondent has failed to explain why the scope of judicial powers should define the scope of lawful private activity. Indeed, courts can order “fencing-in” relief, which restricts even legal conduct in order to help prevent future violations; this does not mean that private parties can agree among themselves to bar the same lawful, competitive activities.23 Private parties cannot agree to limit non-infringing conduct with the effect of restraining competition, even if a court could do so. Moreover, in fashioning relief in trademark cases, courts are guided by equitable principles, which require closely tailoring injunctions to the harm that they address and giving due consideration to the public interest and the potential effect on competition between the parties. SunAmerica Corp. v. Sun Life Assur. Co. of Canada, 77 F.3d 1325, 1336 (11th Cir. 23 In some particularly egregious cases, for example, courts have banned defendants from practicing in certain industries altogether. See, e.g., FTC v. Think Achievement Corp., 144 F. Supp. 2d 1013, 1018 (N.D. Ind. 2000), aff’d, 312 F.3d 259 (7th Cir. 2002) (banning defendants from engaging or assisting others in the businesses of telemarketing and marketing career advisory goods or services); FTC v. Gill, 265 F.3d 944, 957-58 (9th Cir. 2001) (affirming district court order banning defendant from engaging in the credit repair business); FTC v. E.M.A. Nationwide, Inc., 2013 WL 4545143, at *8 (N.D. Ohio Aug. 27, 2013), aff'd, 767 F.3d 611 (6th Cir. 2014) (enjoining defendants from working in the debt relief and mortgage assistance industries). That does not mean that private parties who are competitors can enter into an agreement preventing one of them from practicing in a particular industry.

VOLUME 166 Opinion of the Commission 1996) (quoting 4 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition, § 30.03 [1] (4th ed. 1995)). Private parties need not consider public interest factors in their settlement agreements, and they cannot rely on the remedial authority of courts to shield their private agreements from antitrust scrutiny.

Similarly unavailing is Respondent’s and the Dissent’s suggestion that, just because a court issues a non-use injunction or approves non-use settlements in other trademark cases, it means that the remedy is appropriate here. Respondent asserts that non-use injunctions are “common” and are “the order of the day” in trademark infringement actions. RAB at 11-12. But, in the vast majority of trademark infringement cases, non-use is a perfectly reasonable remedy because it resolves the trademark issue without affecting competition; it simply requires a company with a name confusingly similar to a rival’s to refrain from identifying itself or its products by such a name. Consider, for example, the Clorox case cited above. Clorox v. Sterling Winthrop, Inc., 117 F.3d 50 (2d Cir. 1997). In Clorox, the court found that the “PINE-SOL” trademark and the “LYSOL” trademark were confusingly similar. Indeed, the U.S. PTO initially refused to grant the PINE-SOL trademark for that very reason. Id. at 53. In that context, the non-use agreement prevented a competitor from using a confusingly similarly name for its products.

Here, no company names are alleged to be the cause of any confusion, and as noted above, non-use restrictions cut off an important channel of truthful price advertising. See supra Section IV.A. Whereas a typical trademark non-use remedy affects how a product may be labeled or what language may be used in the text of an ad, the non-use restriction here limits the number of times competitor ads are shown and insulates some 1-800 Contacts’ consumers from becoming aware of its rivals.

Respondent also points to cases in which courts have issued orders that would prohibit use of trademarks in internet advertising, but most of those cases are either consent judgments or default judgments and involve infringing conduct beyond mere keyword bidding.24 In any event, decisions about the appropriate remedy are inherently case-specific, and the fact that a court in some other context, with no or little consideration of the effects on competition, granted a broad injunction does not constitute an endorsement of the private agreements here or render them procompetitive.

24 The two litigated cases Respondent’s expert cites are distinguishable in material respects. See RX0734 (Hogan Expert Report) at 0099-0100 ¶149. In Skydive Ariz., Inc. v. Quattrocchi, 2010 WL 1743189, at *8 (D. Ariz. Apr. 29, 2010), among other things, defendant used plaintiff’s trademark on its website and falsely told its customers that the skydiving certificates it sold would be redeemable at plaintiff’s facilities. The other litigated case is PODS Enterprises, LLC v. U-Haul International, Inc., 126 F. Supp. 3d 1263, 1292 (M.D. Fla. 2015). In that case, PODS Enterprises sued U-Haul for referring to its product as a “U–Box pod” and using the terms “pod” or “pods” thousands of times on its website. The court broadly prohibited use of the term except in comparative advertisement, but it did not specifically discuss banning use of the term in keyword bidding or the effect on competition.

277 1-800 CONTACTS, INC. Opinion of the Commission ii. Less Anticompetitive Alternatives Complaint Counsel identify three alternatives to the restrictions in the Challenged Agreements. They suggest that Respondent could (1) bar the rival from using specific text alleged by 1-800 Contacts to cause confusion, including prohibiting the rival from using a name confusingly similar to its own; (2) require clear disclosure in each search advertisement of the identity of the rival seller; or (3) require the rival to refrain from using confusing or deceptive language in its search ads.

These options present alternative ways for avoiding litigation costs and achieving the procompetitive benefits that flow from trademark protection. The first and third proposed alternatives would adequately address consumer confusion stemming from the content of the ad. Respondent, however, claims that its trademark was infringed by the mere appearance of competitor ads in response to a trademark search, not from any confusing ad content. See RAB at 41. Assuming, arguendo, and contrary to our findings below, that protection against such infringement has been established as a valid procompetitive benefit here, alternatives one and three would not adequately address it. But the second proposed alternative—requiring clear disclosure of the identity of the rival seller—is a workable option that would achieve both litigation cost savings and protection of trademark rights, including prevention of the consumer confusion associated with infringement, in a significantly less anticompetitive manner. Respondent’s arguments to the contrary are unpersuasive.

Respondent contends that a disclosure requirement would be unworkable because “clear and conspicuous disclosure” is an amorphous standard that would likely generate future litigation. RAB at 40-41.25 We, however, do not find a requirement to clearly disclose the seller’s identity to be “amorphous.” The Commission has ordered parties to implement clear and conspicuous disclosures in numerous cases involving misleading advertising and did not find the requirements too amorphous or otherwise problematic to serve its remedial goals.26 Moreover, nothing prevents the parties, as part of their settlement, from agreeing on the specific language of the disclosure that would need to be included in the ads to dispel any purported consumer confusion.

25 Courts have been inconsistent in their burden allocation in assessing less restrictive alternatives, “[b]ut the difference in assignment of this proof burden is more apparent than real.” 11 Herbert Hovenkamp, Antitrust Law ¶ 1914c (3rd ed. 2011). As a leading antitrust treatise explains: The most workable allocation gives the plaintiff the burden of suggesting, or proffering, a particular alternative claimed to achieve the same benefits but less restrictive of competition. The defendant then has the burden of showing that the proffered alternative is either unworkable or not less restrictive.

Id.

26 See, e.g., Paypal, Inc., 2018 WL 2716645 (F.T.C. May 23, 2018); Lenovo (United States), Inc., 2017 WL 6885837 (F.T.C. Dec. 20, 2017); Warner Bros. Home Entertainment, Inc., 2016 WL 6892613 (F.T.C. Nov. 17, 2016); Machinima, Inc., 2016 WL 1130011 (F.T.C. Mar. 16, 2016). VOLUME 166 Opinion of the Commission Respondent also asserts that Complaint Counsel failed to introduce evidence that a clear disclosure would reduce consumer confusion.27 But Complaint Counsel introduced evidence that there was only a de minimis likelihood of confusion in the first place,28 and Respondent’s attempt to demonstrate confusion from keyword bidding was severely flawed, see infra note 39. At the same time, courts have held that a “minimal or moderate amount of potential confusion found could be cured effectively by use of a disclaimer.” See Soltex Polymer Corp. v. Fortex Indus., Inc., 832 F.2d 1325, 1330 (2d Cir. 1987). Indeed, the Tenth Circuit’s ruling in 1-800 Contacts’ own litigation supports the adequacy of disclosure. See 1-800 Contacts, Inc., 722 F.3d at 1245 (consumer confusion would be unlikely “when the entry is clearly labeled as an advertisement and clearly identifies the source, which has a name quite different from the business being searched for”); see also Multi Time Machine, Inc. v. Amazon.com, Inc., 804 F.3d 930, 937 (9th Cir. 2015) (“clear labeling can eliminate the likelihood of initial interest confusion in cases involving Internet search terms”); Network Automation, Inc. v. Advanced Sys. Concepts, Inc., 638 F.3d 1137, 1154 (9th Cir. 2011) (clear labeling might eliminate the likelihood of initial interest confusion in internet advertising); Playboy Enters. Inc. v. Netscape Commc’ns Corp., 354 F.3d 1020, 1025 & n.16 (9th Cir. 2004) (“if a banner advertisement clearly identified its source or, even better, overtly compared PEI products to the sponsor’s own, no confusion would occur under PEI’s theory” (that appearance of a banner advertisement immediately after users type in PEI’s marks caused initial interest confusion)). Any potential for confusion lingering after clear disclosure of the rival seller’s identity could be removed by a further disclosure disclaiming affiliation with 1-800 Contacts.29 Respondent additionally argues that this proposed alternative is “merely theoretical” because the record does not contain any real-world trademark settlements embodying such terms. RAB at 40. But insistence on identifying examples of other settlements that incorporate Complaint Counsel’s specific proposal is unrealistic given the relatively new context of searchbased keyword advertising, particularly in light of the large number of cases dismissing claims based on keyword bidding altogether.30 Moreover, settlement agreements are often subject to confidentiality provisions and consequently unavailable. Cf. RX0734 (Hogan Expert Report) at 0107 (“many of the agreements I have knowledge of are subject to confidentiality provisions”). In any event, an absence of such settlement examples in the record does not determine whether 27 The Dissent argues that we “[d]ismiss 1-800 Contacts’ trademark infringement claims based on [our] evaluation of consumer confusion . . . .” Dissenting Statement at 21. That is incorrect. We merely evaluate Respondent’s argument regarding the evidence produced showing consumer confusion. As Section V.A.3.a.i. and this section show, our Opinion does not hinge on the merits of the trademark claim. We find that challenged restraints are overbroad, which is a question wholly suited for a rule-of-reason inquiry. 28 See CX8008 (Jacoby Expert Report) at 008-010; Jacoby, Tr. 2130. 29 During oral argument, Respondent’s counsel indicated that Google’s policy prohibits a party from including a competitor’s trademark in the text of the ad, even if the trademark is mentioned in order to disclaim any affiliation. Stone, Oral Arg. Tr. at 90. Google’s trademark counsel, however, testified that Google does not prohibit use of another’s trademark in an ad unless the trademark owner notifies Google that it does not want its trademark to be used in the ads by that advertiser. See CX9022 (Charlston Dep.) at 16-17, in camera. 30 See infra note 38 and accompanying text.

279 1-800 CONTACTS, INC. Opinion of the Commission the proposed alternative is workable. The idea that disclaimers can be used to eliminate consumer confusion is not new, and courts have ordered disclaimers as a remedy in internetbased trademark infringement cases. See, e.g., Nissan Motor Co., Ltd. v. Nissan Computer Corp., 89 F. Supp. 2d 1154 (C.D. Cal. 2000), aff’d, 246 F.3d 675 (9th Cir. 2000) (preliminary injunction requiring defendant Nissan Computer Corporation, owner of the websites nissan.com and nissan.net, to clearly identify itself on the website, disclaim affiliation with, and identify the correct website for, Nissan Motor Co., and not to display any automobile-related information or web links.); Tempur-Pedic N. Am., LLC v. Mattress Firm, Inc., 2017 WL 2957912, at *11 (S.D. Tex. July 11, 2017) (permitting defendant to continue to use plaintiff’s trademark in Google AdWords, but limiting number of times defendant could use the mark on its webpage and requiring disclaimer of affiliation); Simone v. VSL Pharm., Inc., 2016 WL 3466033, at *27 (D. Md. June 20, 2016) (allowing competitor to post AdWords ads containing trademark term if such ads also include adequate disclaimer of affiliation, to be pre-approved by the court). The FTC, too, in its decades of experience preventing and remedying false advertising claims and consumer deception, has ordered respondents to provide disclosures to avoid consumer confusion.31 In fact, in 2013, the Commission published guidelines to assist businesses in providing clear, effective disclosures in space-constrained internet ads.32 We thus have successfully employed remedial mechanisms similar to those urged by Complaint Counsel. We see no reason why a brief statement identifying the ad sponsor and/or disclaiming affiliation with 1-800 Contacts would be ineffective or unworkable.33 31 See supra note 24.

32 .Com Disclosures: How to Make Effective Disclosures in Digital Advertising, available at https://www.ftc.gov/sites/default/files/attachments/press-releases/ftc-staff-revises-online-advertising-disclosureguidelines/130312dotcomdisclosures.pdf.

33 In addition to asserting federal trademark infringement claims under the Lanham Act §§ 32 and 43(a), Respondent’s lawsuits alleged state and common law unfair competition (Utah Code Ann. § 13-5-1 et seq.), federal trademark dilution (15 U.S.C. § 1125(c)), and unjust enrichment, which they claim provide additional justifications for the challenged settlements. See RAB at 2. The state and common law unfair competition claims as well as the unjust enrichment claims are co-extensive with the federal trademark infringement claims, so they do not justify the restrictive settlement terms for the reasons discussed in the text. See Amoco Oil Co. v. Rainbow Snow, 748 F.2d 556, 558 (10th Cir. 1984) (“This ‘likelihood of confusion’ test is also applicable to Amoco’s . . . state claims of infringement, Utah Code Ann. § 70-3-13 (1953), and its common law claims of unfair competition and deceptive trade practices.”); Primary Children's Med. Ctr. Found. v. Scentsy, Inc., 2012 WL 2357729, at *9 n.4 (D. Utah, June 20, 2012), as amended, (July 6, 2012) (“Because [plaintiff's] state law claims [including unfair competition under Utah Code Ann. § 13-5a-101 et seq. and common law unfair competition] all require a finding of trademark infringement or a likelihood of confusion, the court does not find it necessary to analyze these claims separately [from the federal claims].”); 1-800 Contacts, Inc. v. Lens.com, Inc., 755 F. Supp. 2d 1151, 1190 (D. Utah 2010) (granting summary judgment on Respondent’s unjust enrichment claim because it did not adequately show trademark infringement; “if Plaintiff were able to obtain payment under unjust enrichment, common law would effectively expand the scope of Plaintiff's statutory protection”). As to Respondent’s federal trademark dilution claim, this purported justification pertains at most to two agreements. Of the thirteen complaints filed in connection with the challenged settlement agreements, only two asserted federal trademark dilution, and these complaints challenged pop-up ads appearing on 1-800 Contacts’ website, not the display of ads on SERPs in response to searches for trademark terms. Indeed, Respondent stopped asserting trademark dilution claims after 2004, which suggests that even Respondent believed these claims to be either weak or at most peripheral to its case. VOLUME 166 Opinion of the Commission Given the inherently suspect nature of Respondent’s advertising restraints and our finding that the procompetitive benefits asserted to justify those restraints could be achieved by significantly less anticompetitive means, we can conclude that Respondent has engaged in unfair methods of competition in violation of Section 5 of the FTC Act. Although Complaint Counsel have met their burden by demonstrating that Respondent could have chosen a less restrictive alternative to achieve the same procompetitive benefit, we also consider whether Complaint Counsel has satisfied its further showing by focusing alternatively on the particular restraints and context presented here.

b. Complaint Counsel’s Showing that the Restraints Are Likely, in the Particular Context, to Harm Competition Our review of the record shows that the restraints “are indeed likely, in the particular context, to harm competition.” Polygram, 136 F.T.C. at 348. Online search advertising is a key method for consumers to discover, compare, and reach online contact lens vendors and for lower-priced retailers to compete. IDF 564-65. It enables online sellers to increase brand awareness and to obtain new customers. IDF 497. It is displayed at the key moment when the consumer is more likely to be looking to buy. IDF 498.

For 1-800 Contacts, search advertising is important. From 2004 to 2014, between {29 and 53} percent of 1-800 Contacts’ internet advertising budget was spent on paid search advertising each year. IDF 66. 1-800 Contacts earns approximately {20 percent} of its sales from paid search advertising. IDF 580.

Search advertising is similarly important for 1-800 Contacts’ online competitors. The record shows that online retailers have found search advertising much more effective in reaching potential buyers than other types of advertising. For example, AC Lens has found that, compared to other marketing channels, search advertising generates the most new customer orders and the most revenue, at a cost consistent with AC Lens’ financial goals. IDF 500-01. Thus, for AC Lens, search advertising is the most effective and important marketing channel to grow its business. IDF 502.

Other online competitors reported similar reliance on search advertising. Vision Direct advertised almost exclusively online. IDF 540. Search advertising “was a major driver” in building its business, including driving traffic to Vision Direct’s website and generating new and repeat sales. IDF 542-43. Web Eye Care predominantly relies on paid search advertising, because it has determined that search advertising “drives the most traffic” and orders at an acceptable cost. IDF 556-58. For LensDirect, paid search advertising constitutes its most important marketing channel and has been effective in generating growth. IDF 523. Lens Discounters found that paid search advertising is “essential” to its ability to attract new customers because it reaches customers who are seeking to purchase contact lenses online. IDF Respondent’s briefing on appeal does not provide any explanation as to why dilution claims justify the settlements. Nor does it provide any reasons why its dilution concerns would not be adequately addressed by the less restrictive alternatives that Complaint Counsel have proposed. Accordingly, trademark dilution does not justify the Challenged Agreements either.

281 1-800 CONTACTS, INC. Opinion of the Commission 528. For Memorial Eye, search advertising was the “most efficient,” form of advertising, which was “critical” to the company’s growth. IDF 535, 537. Similarly, search advertising was “[e]specially important” for Walgreens when it began selling contact lenses online because it helped let people know that Walgreens sold contact lenses; it was “an essential form” of advertising for Walgreens to remain competitive with other online retailers of contact lenses. IDF 549-50.

Not only is search advertising in general important, the record shows that search advertising generated by searches for 1-800 Contacts’ trademark terms is important. Trademark search is a significant source of 1-800 Contacts’ business. IDF 566. It accounts for the substantial majority of 1-800 Contacts’ new customer orders attributable to paid search advertising. IDF 570. In 2006, 2007, and 2008, trademark search generated far more orders than non-trademark searches. IDF 572. In 2015, between 20 and 31 percent of 1-800 Contacts’ initial web orders came from users searching for 1-800 Contacts’ trademark terms. IDF 571. 1­800 Contacts’ trademark terms have higher conversion rates34 than non-branded search terms. IDF 573.

Similarly, for 1-800 Contacts’ online rivals, advertising displayed for searches on 1-800 Contacts’ trademark terms is important. During the period from 2002 through 2016, Google displayed advertisements for nine of the 14 contact lens retailers that are parties to the Challenged Agreements, as a result of their direct bidding on 1-800 Contacts’ trademark terms prior to entering into the agreements. IDF 653. These nine firms found such keyword bidding to be worth the cost, and Google determined their advertisements were sufficiently relevant to warrant display. Id. In addition, parties to the Challenged Agreements consistently testified that, absent the agreements, they would bid, or test bidding, on 1-800 Contacts’ trademark terms and/or remove negative keywords from their advertising accounts. IDF 590 (AC Lens), 595 (Empire Vision), 616 (Lenses for Less), 630 (Vision Direct), 634-35 (Walgreens), 650 (Web Eye Care).

Respondent argues that the Challenged Agreements prohibit ads for only a small number of searches. RAB 17. That argument is contradicted by the evidence. The volume of searches for 1-800 Contacts’ trademark terms is significant. Based on the comScore dataset of searches by users for the period July 2013, through July 2016 (the “comScore dataset”35) analyzed by Complaint Counsel’s expert witness, Dr. Susan Athey, 17 percent of search queries for contact lenses were for 1-800 Contacts’ trademark terms. IDF 657. The volume of searches for 1-800 Contacts’ trademark terms in the comScore dataset was similar in size to the collective volume of searches for the top three generic terms (“contact,” “contact lenses,” and “contacts”). IDF 34 A “conversion” refers to a sale made over the internet. The conversion rate is the number of times a conversion occurs divided by the total number of ad clicks. IDF 156.

35 ComScore is a company that collects data from a panel of internet users by installing software on consumers’ devices to track their behavior, including collecting information on the screens that users see when they perform searches. IDF 700. Dr. Athey received from comScore detailed online search information from 377,002 internet users in the United States from July 11, 2013, through August 14, 2016, covering all the search queries those users performed on all major search engines and reported at a query-by-query level. IDF 701. VOLUME 166 Opinion of the Commission 658-59. The 1-800 Contacts search term is the largest, single brand-name search term, according to the comScore data analyzed by Dr. Athey. IDF 660.

The reason that 1-800 Contacts’ rivals’ ads are so important at this key moment is that the rival online sellers offer lower prices, IDF 661, 693, and advertising for those retailers often emphasizes those lower prices. See IDF 587, 591, 603, 611, 646, 703, 724; Holbrook, Tr. 1904. That information is valuable: online shoppers for contact lenses are primarily concerned with low prices, IDF 705-08. Yet, in a 2012 consumer survey of 1-800 Contacts’ customers, more than one-third of respondents explained that they initially purchased from 1-800 Contacts because “It Was the Only Online Contacts Site of Which I Was Aware,” IDF 695, 697, and a 2015 AEA Investors Fund analysis based on another survey found that actual price variances were “much more” than consumers thought them to be. RX1228 at 36. Consumers respond to competitors’ ads displayed in response to searches for 1-800 Contacts’ trademark terms. Based on data analyzed by Complaint Counsel’s expert, Dr. Athey, firms that are currently bidding on “1-800 Contacts,” have a higher conversion rate for those searches than for other search terms. IDF 661. 1-800 Contacts observed that an increase in competitor ads appearing in response to a search for 1-800 Contacts’ trademark terms tends to decrease sales for 1-800 Contacts. IDF 711. For example, in a report covering the week ending September 22, 2007, 1-800 Contacts noted a 6 percent week-over-week drop in trademark paid search orders, in part because of competition from Vision Direct, which had been “advertising in the 2nd position on many of [the SERPs for searches for 1-800 Contacts’] branded terms in Google.” IDF 717. See also IDF 726 (report for the week ending March 12, 2010: 1-800 Contacts experienced a lower click-through-rate than in prior weeks, which is “likely the result of additional competitor’s ads . . . showing up on our best terms such as 1800contacts and 1800 contacts”), 727 (report for the week ending June 11, 2010: 1-800 Contacts’ trademark paid search orders through Google and click-through rates for trademark ads “were slightly softer than [the preceding week] because of increased competition on [1-800 Contacts’] best branded terms”).

Similarly, 1-800 Contacts found that reducing the competitor ads that appear in response to searches for 1-800 Contacts’ trademark terms increased sales. IDF 710. For example, in a report covering the week of June 20, 2008, 1-800 Contacts attributed an increase in orders as being helped in part by “LensWorld finally removing all their ads from all of [1-800 Contacts’] trademark keywords.” IDF 719. See also IDF 725 (report for the week ending January 8, 2010: 1-800 Contacts achieved “an all-time record high” for orders obtained through searches for its trademark keywords, due in part to fewer advertisers appearing on searches for 1-800 Contacts’ trademark terms that week, “which always helps improve performance”), 730 (reporting that in late August 2010, orders from new customers coming through search ads on searches for 1-800 Contacts’ trademarks “jumped to the highest level of the year,” due in part to the appearance of “fewer competitors on [1-800 Contacts’] best TM words such as 1800contacts 1 800 contacts and 1800 contacts.”), 723 (report for the week of March 6, 2009: “[t]here are substantially less competitors showing up on our list of monitored TM words . . . in Google[,] which is likely helping improve our TM [conversion rate] and TM order volume.”). 283 1-800 CONTACTS, INC. Opinion of the Commission In addition, it is worth highlighting that Challenged Agreements covered 14 different online contact-lens retailers that account for 79 percent of online contact lenses in the United States. IDF 496. That is in stark contrast to the Clorox case. There, the court saw a jilted competitor who wanted to use an antitrust claim to negotiate a better trademark settlement. The court recognized that “the antitrust laws do not exist to protect competitors from agreements that in retrospect turn out to be unfavorable to the complaining party.” Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50, 57 (2d Cir. 1997). The court further expressed its skepticism where the challenged agreement restricted only one brand, PINE-SOL, and left “established large competitors” that are “some of the largest corporations in the country” free to compete, “as these companies have repeatedly done.” Id. at 58. Predictably, Clorox was unable to muster much evidence of consumer harm. The court found implausible the assertion by Clorox, a “megabrand with substantial brand equity,” that restrictions on the ability to use the PINE-SOL brand somehow erected an insurmountable barrier to competition. Id. The court dismissed as unpersuasive Clorox’s anecdotes of failed products because the trademark allegedly did not fit the new market. The court also rejected Clorox’s theory that only Clorox could provide competition to the LYSOL brand. Id. In short, Clorox bears very little resemblance to the present case. Unlike the one competitor that was worse off in Clorox, the challenged agreements here cover the landscape of online contact-lens retailers resulting in harm to competition overall. And unlike Clorox’s thin evidence of consumer harm, the record here is replete with direct evidence of anticompetitive effects.

The Dissent claims that the disposition of Clorox should be the same here. In particular, the Dissent claims that the court in Clorox thought “the form and scope of trademark settlement agreements deserves ‘substantial weight’ because the settling parties ‘are in the best position to determine what protections are needed’ and ‘it is usually unwise for courts to second-guess such decisions.’” Dissenting Statement at 19 (quoting Clorox, 117 F.3d at 57). But looking at the form and scope of an agreement is at the very heart of a Section 1 analysis. For instance, we neither ignore nor defer to the parties in assessing the form and scope of an agreement in reverse payment cases because, indeed, the form and scope of the agreement lie at the very core of how parties make a reverse payment. By asking us to simply defer to the parties to a settlement, we fear that the Dissent essentially advocates for application to cases at the intersection of antitrust and trademarks a version of the “scope of the patent” test that was rejected in Actavis. We decline to follow that suggestion.

This examination of the context of the particular advertising restraints in the Challenged Agreements demonstrates that anticompetitive effects are likely. Economics and prior cases counsel that the challenged advertising restrictions prevent consumers from obtaining information that would permit price and service comparisons. The record evidence showing the significance of search advertising and searches for 1-800 Contacts trademark terms in particular; the price competition offered by 1-800 Contacts’ rivals; and the consumer responses to online competitors’ ads generated by searches for 1-800 Contacts trademarks confirm that the Challenged Agreements are “indeed likely, in [this] particular context, to harm competition.” Polygram, 136 F.T.C. at 348.

VOLUME 166 Opinion of the Commission 4. 1-800 Contacts’ Response to Complaint Counsel’s Showing of Anticompetitive Effects 1-800 Contacts responds to Complaint Counsel’s showing that the restraints are inherently suspect and likely to have substantial anticompetitive effects by challenging the factual, economic, and legal support for the demonstration of those anticompetitive effects. These challenges are not persuasive.

Respondent and the Dissent argue that the Challenged Agreements affected advertising by only some companies, in only one medium, in response to only a portion of internet searches related to contact lenses. RAB at 17. Of course, the fact that some advertising remained unrestrained does not excuse a restraint affecting a competitively significant subset of ads. While the Challenged Agreements do not prevent all advertising for the online sale of contact lenses, they affect a particularly significant type of advertising for online sales at the crucial moment when sales are about to be made. See IDF 661 (finding a higher conversion rate for bids on “1-800 Contacts” than for other search terms). The suppressed ads would have enabled consumers to learn about alternative, lower-priced sellers of contact lenses and to make price comparisons. Prohibiting this particular type of advertising is likely to have substantial anticompetitive effects.

Respondent argues that the ads banned by the settlements could have different effects from advertising in other markets. See RRB at 23. Although Respondent points to some attributes of search advertising—some consumers may be conducting navigational searches and expect to see the most relevant results appearing first,36 and some consumers may be unable to distinguish between paid search ads and organic results—Respondent does not identify any record evidence demonstrating that consumers’ purchasing behavior in response to search ads generated by 1-800 Contacts’ trademark terms differs from their response to other advertising. Nor does Respondent identify any other market effects that differ from other contexts. As previously discussed, consumers respond to the presence of rivals’ contact lens ads by clicking on the ads and converting those clicks to sales, even if some consumers are performing navigational searches. IDF 710-19, 723-31. Thus, when consumers are presented with information that informs them of alternative online sellers offering lower prices, they respond to advertising in this market the same way that they do elsewhere. Respondent similarly argues that the economic literature has not looked specifically at paid search advertising, which involves “complexities” in the algorithms employed by search engines. RRB at 22. Although the algorithms underlying the search auctions are complex, the behavior of consumers and advertiser-sellers in response to this type of advertising is the same as for other types of advertising. Respondent identifies no differences in the responses of market participants, so the fact that economic studies did not specifically examine search advertising 36 Respondent suggests that absent the advertising restrictions rivals’ ads would appear first. Yet “1-800 Contacts’ strategy to search advertising was to spend as much as necessary when bidding on its trademark keywords to meet its goal of ensuring that 1-800 Contacts’ advertisement was the first advertisement displayed in response to searches for its trademark.” IDF 575; CX0935. Viewed in light of this strategy, the advertising restrictions may enable 1- 800 Contacts to reduce its bids and pay lower prices, but they do not better satisfy consumer expectations. 285 1-800 CONTACTS, INC. Opinion of the Commission does not affect their relevance. As the D.C. Circuit has explained, condemnation of a particular horizontal restraint as inherently suspect looks only for “the close family resemblance between the suspect practice and another practice that already stands convicted in the court of consumer welfare.” Polygram, 416 F.3d at 37.

Finally, Respondent argues that a finding that the settlement agreements are inherently suspect is inconsistent with the Supreme Court’s opinion in Actavis. See RRB at 22. We disagree. Actavis does not stand for the proposition that no restriction in a settlement agreement– even an intellectual property settlement agreement—can be inherently suspect. Indeed, the Supreme Court has often concluded that restraints embedded in settlement agreements are unlawful without resorting to a full rule-of-reason analysis. See Singer Mfg.; New Wrinkle; U.S. v. Masonite Corp., 316 U.S. 265 (1942). Rather, Actavis describes how to analyze the reverse payment settlements there at issue. It says that Hatch-Waxman reverse-payment, patent settlements, without more, are not inherently suspect because reverse payments are a special breed of settlement. In particular, the Court recognized that reverse payment settlements are complex cases where the likelihood of “anticompetitive effects depends upon [the payment’s] size, its scale in relation to the payor’s anticipated future litigation costs, its independence from other services for which it might represent payment, and the lack of any other convincing justification.” Actavis, 570 U.S. at 159. Here, Respondent’s agreements take the form of a quintessential advertising restraint that is targeted to interfere with price-setting mechanism among online contact-lens sellers. The restraint involves none of the complexities identified by the Court in conjunction with reverse payments, and there is no reason its competitive harms cannot be established through Polygram’s inherently suspect framework. Putting aside whether the restraints at issue here are properly classified as inherently suspect, Complaint Counsel’s more detailed showing described above and lack of offsetting efficiencies meet the requirements of the rule of reason to support liability. The Actavis Court noted that the Commission need not “litigate the patent’s validity, empirically demonstrate the virtues or vices of the patent system, present every possible supporting fact or refute every possible pro-defense theory” to show that a reverse payment is anticompetitive. Id. at 159. Rather, “‘[t]here is always something of a sliding scale in appraising reasonableness,’” and as such “‘the quality of proof required should vary with the circumstances.’” Id. (quoting California Dental, 526 U.S. at 780) (internal citations omitted). The Court stressed: “As in other areas of law, trial courts can structure antitrust litigation so as to avoid, on the one hand, the use of antitrust theories too abbreviated to permit proper analysis, and, on the other, consideration of every possible fact or theory irrespective of the minimal light it may shed on the basic question—that of the presence of significant unjustified anticompetitive consequences.” Id. Here, we simply follow Polygram’s framework for structuring the rule-of-reason inquiry in the context of an advertising restriction case and analyze the intellectual property issues in that framework.

5. Validity of the Asserted Procompetitive Justifications As discussed in Section V.A.2, the avoidance of litigation costs through settlement and the protection of trademark rights are cognizable and facially plausible procompetitive VOLUME 166 Opinion of the Commission justifications because, under the right circumstances, both cost savings and trademark protection can result in enhanced competition and innovation. But Complaint Counsel have shown that the purported procompetitive benefits could have been accomplished through means less restrictive of competition. See supra Section V.A.3.a. Even if no less restrictive alternative were available, however, Respondent’s case would falter because it has not shown that its purported justifications have a basis in fact, i.e., that they are valid as well as plausible and cognizable. As we noted in Polygram, the respondent “has the burden of producing factual evidence in support of its contentions.” 136 F.T.C. at 350; see also Mass. Board, 110 F.T.C. at 604 (“if the efficiency justification is plausible, further inquiry . . . is needed to determine whether the justification is really valid”).

At this point, then, we look closer at Respondent’s asserted justifications and require sufficiently detailed evidence to establish that the justifications are not merely plausible, but actually valid. See Polygram, 416 F.3d at 36 (explaining that respondent’s burden at this stage is to show the restraint “in fact” does not harm consumers or has procompetitive virtues). We find that Respondent has not met this burden.37 a. Avoidance of Litigation Costs through Settlement Although Respondent has identified litigation cost savings, it has not demonstrated that these cost savings would have procompetitive effects. Respondent must provide “some explanation connecting [its] practice[s] to consumers’ benefits.” Chicago Prof’l Sports, L.P. v. Natl Basketball Assn, 961 F.2d 667, 674 (7th Cir. 1992); see also Polygram, 136 F.T.C. at 345 (describing legitimate justifications as “reasons why the practices are likely to have beneficial effects for consumers”). But Respondent provides no basis for finding that the litigation cost savings would be passed through to consumers or would otherwise benefit competition in a way that could offset the anticompetitive effects. Capital savings are not cognizable efficiencies in and of themselves, though they may be cognizable if defendant demonstrates that avoidance of capital expenditures provides a tangible, verifiable benefit to consumers by lowering prices or improving service quality. FTC v. Penn State Hershey Med. Ctr., 838 F.3d 327, 350 (3d Cir. 2016). “While increasing output, creating operating efficiencies, making a new product available, enhancing product or service quality, and widening consumer choice have been accepted by courts as justifications for otherwise anticompetitive agreements, mere profitability 37 In addition, the ALJ found that two other justifications suggested by Respondent—increased online sales of contact lenses and minimization of search costs—were unsupported, but Respondent has not challenged those portions of the ALJ’s decision on appeal. Accordingly, Respondent waived its arguments with respect to those justifications. See 16 C.F.R. § 3.51(b) (“[a]ny objection to a ruling by the Administrative Law Judge, or to a finding, conclusion, or a provision of the order in the initial decision, which is not made a part of an appeal to the Commission shall be deemed to have been waived”); United States v. Jernigan, 341 F.3d 1273, 1283 n.8 (11th Cir. 2003) (“a party seeking to raise a claim or issue on appeal must plainly and prominently so indicate”; otherwise, the issue “will be considered abandoned”). In any event, we agree with the ALJ that the increased sales justification lacks evidentiary support (see ID at 188-89), and the asserted benefit from minimization of consumer search costs is neither factually nor legally valid (see id. at 184-87 & n.45). Cf. Polygram, 136 F.T.C. at 356 (noting expert testimony that an agreement among competitors not to advertise is likely to harm consumers and competition by raising consumers’ search costs).

287 1-800 CONTACTS, INC. Opinion of the Commission or cost savings have not qualified as a defense under the antitrust laws.” Law, 134 F.3d at 1023. Respondent has not demonstrated that the litigation cost savings provide benefits to consumers that could or would offset the competitive harms attributable to its conduct. Moreover, the litigation settlement justification is at most partial. It has no bearing whatsoever on the Luxottica Sourcing and Services Agreement. That agreement involved no litigation, no settlement, and no litigation cost savings.38 The Dissent argues that our decision errs by failing to account for saved litigation costs that do not result in cost savings to consumers. The Dissent claims that our analysis contradicts Actavis, which it believes accommodates any saved litigation costs—irrespective of whether the savings passed down to consumers or not. Though not openly stated, the Dissent asks us to take up the total welfare standard for evaluating efficiencies, which does not require a showing of how the proffered efficiency benefits consumers.39 We, however, believe the sounder approach—and the approach that is most consistent with long-standing antitrust practice—would be to ensure that if consumers are harmed by the challenged restraints, Respondent should be required to explain and detail how its restraints actually benefit consumers. The Dissent advocates skipping that step; we decline.

b. Trademark Protections Respondent and the Dissent argue that 1-800 Contacts’ agreements facilitate trademark protection, which allows retailers to market products in a way that reduces the likelihood of consumer confusion and incentivizes brand-building. Both maintain that brand-building, in turn, assures consumers of consistent quality and reduces consumer costs of making purchasing decisions. Respondent’s Post-Trial Brief at 36-38, 45; Dissenting Statement at 24-26. Although trademark protection can be a legitimate justification, it does not justify the restraints challenged in this case.

To overcome Complaint Counsel’s showing of anticompetitive effects, Respondent must show that trademark protection is more than a procompetitive justification in theory and is, in fact, a valid justification for the restraints challenged here. See Polygram, 136 F.T.C. at 349 (explaining that if the respondent fails to refute the plaintiff’s detailed showing of competitive harm, the respondent has the burden of showing that “detailed evidence supports its proffered justification”); Mass. Board, 110 F.T.C. at 604 (requiring a showing that “the justification is really valid”). We find that Respondent has not carried that burden. 38 Section III.B of the Dissent offers considerations that might justify the challenged restraints in the Luxottica Sourcing and Services Agreement. But Respondent did not assert these potential efficiencies as procompetitive benefits and consequently did not attempt to carry its burden of establishing them. Nor has Respondent argued or submitted evidence that the challenged restraint is an ancillary restraint saved by the Luxottica Sourcing and Services Agreement.

39 See Dennis W. Carlton, Does Antitrust Need to be Modernized?, 21 J. ECON. PERSP. 155, 157 (2007) (“The proper objective of antitrust should be total surplus, not consumer surplus.”). VOLUME 166 Opinion of the Commission To establish a federal trademark infringement claim under either Lanham Act § 32 (15 U.S.C. § 1114) or § 43 (15 U.S.C. § 1125(a)), a plaintiff must show that use of its mark is likely to cause consumer confusion as to the source, affiliation, or sponsorship of a company’s products or services. Scott Fetzer Co. v. House of Vacuums, Inc., 381 F.3d 477, 483 (5th Cir. 2004) (citing 15 U.S.C.A. § 1114(1); id. § 1125(a)); A & H Sportswear, Inc. v. Victoria's Secret Stores, Inc., 237 F.3d 198, 210 (3d Cir. 2000). Confusion must be probable, not merely possible, id., and use of the mark must be likely to confound “an appreciable number of reasonably prudent purchasers exercising ordinary care.” Boston Duck Tours, LP v. Super Duck Tours, LLC, 531 F.3d 1, 12 (1st Cir. 2008) (quoting Intl Assn of Machinists and Aerospace Workers, AFL–CIO v. Winship Green Nursing Ctr., 103 F.3d 196, 200 (1st Cir.1996)) (internal quotation marks omitted); see also Entrepreneur Media, Inc. v. Smith, 279 F.3d 1135, 1151 (9th Cir. 2002); Savin Corp. v. Savin Grp., 391 F.3d 439, 456 (2d Cir. 2004).

Although claims based on keyword bidding have sometimes withstood dispositive motions,40 apart from a single district court summary judgment decision from over ten years ago,41 no court has found bidding on trademark keywords to constitute trademark infringement, absent some additional factor, such as a misleading use of the trademark in the ad text that confuses consumers as to the advertisement’s source, sponsorship, or affiliation.42 Rather, “[c]ourts have consistently rejected the notion that buying or creating internet search terms, alone, is enough to raise a claim of trademark infringement.” Tempur-Pedic N. Am., 2017 WL 2957912, at *7 (holding, on motion for preliminary injunction, that “[b]ecause the court has concluded that the purchase of AdWords alone, without directing consumers to a potentially confusing website, is unlikely to cause customer confusion, the AdWords will not be included in the injunction”); see Acad. of Motion Picture Arts & Sciences v. GoDaddy.com, Inc., 2015 WL 5311085, *50 (C.D. Cal. Sept. 10, 2015) (“There is a growing consensus in the case authorities that keyword advertising does not violate the Lanham Act.”).43 Indeed, Respondent lost the one 40 See e.g., Hearts on Fire Co., LLC v. Blue Nile, Inc., 603 F. Supp. 2d 274, 288 (D. Mass. 2009); Fair Isaac Corp. v. Experian Info. Sols. Inc., 645 F. Supp. 2d 734, 760–61 (D. Minn. 2009). 41 See Soilworks, LLC v. Midwest Indus. Supply, Inc., 575 F. Supp. 2d 1118 (D. Ariz. 2008). 42 See CX8014 at 021 (¶ 43) (Tushnet Rebuttal Report) (noting that the “preeminent expert on internet advertising law . . . has been unable to identify any case in which a defendant lost a trial on likely confusion based on purchases of a plaintiff’s trademark as a search engine keyword – despite the filing of over a hundred such cases”); Hogan, Tr. 3459-61 (1-800 Contacts’ trademark law expert acknowledging that he was not aware of any court that had found liability based on keyword bidding alone); USA Nutraceuticals Grp., Inc. v. BPI Sports, LLC, 165 F. Supp. 3d 1256, 1266 (S.D. Fla. 2016) (“[Plaintiff] points to no case indicating that the simple purchase of advertising keywords, without more, may constitute initial interest confusion. . . .”). 43 See also, e.g., USA Nutraceuticals, 165 F. Supp. 3d at 1274 (denying motion for preliminary injunction; “consumers viewing the advertisements are unlikely to be confused as to what, if any, relationship or affiliation exists” between plaintiff and defendant, as the advertisement “makes clear [who] is the proponent of the particular product”); Novation Ventures, LLC v. J.G. Wentworth Co., LLC, 2015 WL 12765467, at *8 (C.D. Cal. Sept. 21, 2015) (granting motion to dismiss; “[i]f a consumer conducts an Internet search for the term ‘Novation’ and Defendants’ advertisements appear in the search results – again, labeled with the word ‘Ad’ – it would not confuse consumers.”); Infogroup, Inc. v. DatabaseLLC, 95 F. Supp. 3d 1170, 1190-91 (D. Neb. 2015) (denying motion for preliminary injunction; no likelihood of success on claim based on keyword bidding where ads “do not use [plaintiff’s] marks in the advertisement itself, and each is either separated from the search results or plainly labeled 289 1-800 CONTACTS, INC. Opinion of the Commission infringement case that it pursued to judgment. See 1-800 Contacts, Inc. v. Lens.com, Inc., 722 F.3d 1229 (10th Cir. 2013) (affirming, in relevant part, summary judgment in favor of defendant). As the appellate court explained:

Perhaps in the abstract, one who searches for a particular business with a strong mark and sees an entry on the results page will naturally infer that the entry is for that business. But that inference is an unnatural one when the entry is clearly labeled as an advertisement and clearly identifies the source, which has a name quite different from the business being searched for.

as a sponsored advertisement.”); Goldline, LLC v. Regal Assets, LLC, 2015 WL 1809301, at *3 (C.D. Cal. Apr. 21, 2015) (granting motion to dismiss claims based on keyword advertising; “there is simply nothing stated, that if deemed true, constitute[s] commercial use that would likely cause confusion as to the origin or affiliation”); Infostream Grp., Inc. v. Avid Life Media Inc., 2013 WL 6018030, at *5 (C.D. Cal. Nov. 12, 2013) (granting motion to dismiss; “[plaintiff] cannot plausibly claim that [defendant’s] mere use of keywords caused any consumer confusion”); Allied Interstate LLC v. Kimmel & Silverman, P.C., 2013 WL 4245987, *6 (S.D.N.Y. Aug. 12, 2013) (granting defendant’s motion for judgment on the pleadings; display of ads “is an indicator of the relevance, not of the source of Defendants’ advertising,” and Google’s labeling of the ads “in no way suggests that it is advertising for or by Plaintiff”); Gen. Steel Domestic Sales, LLC v. Chumley, 2013 WL 1900562, *10 (D. Colo. May 7, 2013), judgment aff’d, 627 Fed. Appx. 682, 2015-2 Trade Cas. (CCH) ¶ 79255 (10th Cir. 2015) (after trial, finding no likelihood of confusion due to trademark keyword bidding; “[a]advertisements on Google appear in a list as distinct and independent entries that internet users can browse and select at will …. [T]he connection between the search term entered and the appearance of an advertisement is too attenuated to suggest an actual affiliation.”); CollegeSource, Inc. v. AcademyOne, Inc., 2012 WL 5269213, at *20 (E.D. Pa. Oct. 25, 2012), aff'd, 597 F. App'x 116 (3d Cir. 2015) (granting summary judgment for defendant; no likelihood of confusion found where the surrounding ad context, including separation of sponsored ad links and labeling of sponsored links, decreased any potential likelihood of confusion); Jurin v. Google Inc., 695 F. Supp. 2d 1117, 1122 (E.D. Cal. 2010) (granting motion to dismiss; “it is hardly likely that with several different sponsored links appearing on a page that a consumer might believe each one is the true ‘producer’ or ‘origin’ of the [plaintiff’s] product”); Boston Duck Tours, LP v. Super Duck Tours, LLC, 527 F. Supp. 2d 205 (D. Mass. 2007) (finding that trademark keyword purchase did not violate preliminary injunction; because triggered advertisement clearly identified the defendant as the source of the ad, trademark use did not result in a likelihood of confusion but constituted “fair, albeit aggressive, competition not prohibited by the Lanham Act”); J.G. Wentworth, S.S.C. Ltd. P’ship v. Settlement Funding LLC, 2007 WL 30115, at *6 (E.D. Pa. Jan. 4, 2007) (granting motion to dismiss; “Even accepting plaintiff’s allegations as true – i.e., assuming that defendant did in fact use plaintiff’s marks through Google’s AdWords program or in the keyword meta tags for its web site – as a matter of law defendant’s actions do not result in any actionable likelihood of confusion under the Lanham Act.”); Govt Employees Ins. Co. v. Google, Inc., 2005 WL 1903128, at *7 (E.D. Va. Aug. 8, 2005) (on motion for judgment as a matter of law, finding that “plaintiff has failed to establish a likelihood of confusion stemming from Google's use of GEICO’s trademark as a keyword and has not produced sufficient evidence to proceed on the question of whether the Sponsored Links that do not reference GEICO’s marks in their headings or text create a sufficient likelihood of confusion); cf. 3form, Inc. v. Lumicor, Inc., 2012 U.S. Dist. LEXIS 27504, at *26 (D. Utah Mar. 1, 2012) (granting motion for summary judgment based on use of trademark metatags because “the fact that a competitor’s search results appear as one of many options when conducting a web search will not confuse consumers, as they will have different appearances”). VOLUME 166 Opinion of the Commission Id. at 1245. Despite the accumulating evidence regarding the weakness of trademark infringement claims, 1-800 Contacts continued to police and enforce the Challenged Agreements and, consequently, continued to extend their anticompetitive effects.44 A leading trademark treatise agrees that displays of non-deceptive advertising links arising from competitors’ purchases of trademark keywords are not confusing. See 5 McCarthy on Trademarks and Unfair Competition § 25A:8 (5th ed. Supp. 2018 update). The author explains that while “a web user may be ‘distracted’ or ‘diverted’ by the search engine displaying ads for other sources . . . distraction or diversion is not the same as ‘confusion’ by the web shopper.” Rather, initial interest confusion can occur “only if the web user mistakenly thought she was going to a web site about TOYOTA cars when she clicked on the keyword link for VOLKSWAGEN. That would depend on how clearly labeled was the advertising link for VOLKSWAGEN.” Id.

We are neither deciding matters of trademark law nor suggesting that to determine whether the Challenged Agreements unreasonably restrain competition, we need to conduct a mini-trial on the merits of the underlying trademark litigations. Respondent’s justifications, however, must meet at least a minimum threshold of validity—more than merely surviving challenges as shams. In this case, the agreements restrict a type of competitive advertising that has never been found to violate the trademark laws, and the weight of authority overwhelmingly points to non-infringement. We are not convinced that trademark protection in this case is a valid procompetitive benefit that merits suppressing truthful advertising. The justification for including negative keywords in the agreements is even weaker. Not only is there a lack of support for a finding of confusion, discussed above, but no court has ever found that bidding on a generic keyword (like “contacts”), which may be broad or phrase matched by the search engine to a trademark search, is even a “use.” On the contrary, in the 2010 decision rejecting Respondent’s case against Lens.com, the district court stated: It is beyond dispute that a competitor cannot be held liable for purchasing a generic keyword to trigger an advertisement that does not incorporate a holder’s mark in any way, even if that competitor’s advertisement appeared when a consumer entered a trademarked search term. 44 Respondent’s effort to supply new information in support of its contention that the appearance of competitor ads in response to searches for “1-800 Contacts” generated consumer confusion—by introducing a consumer survey conducted by its expert, Dr. Van Liere—is unpersuasive. We find that survey deeply flawed and skewed in a way that overstates the difference between the percentage of confused consumers in the test group and the percentage of confused consumers in the control group. Among the problems with the survey are the removal of 1-800 Contacts’ own ad from the test SERP, even though that ad would generally appear at the top of the sponsored results in the real world (IDF 767, 772-73); the failure to test whether the purported confusion was caused by use of the trademark keyword, rather than by other factors, such as the existence of sponsored links (see CX8011 (Jacoby Rebuttal Report) at ¶¶ 7-9, 11-19); and providing more total links in the test SERP than in the control SERP (id. at ¶¶ 30, 33(b)). For the reasons described in the Initial Decision, we also find that the expert opinions of Dr. Goodstein and Dr. Ghose, as well as the Memorial Eye customer service records, do not establish that consumers are likely to be confused about source, sponsorship, or affiliation of the sponsored ads. See ID at 172-75, 181-84. 291 1-800 CONTACTS, INC. Opinion of the Commission Lens.com, 755 F. Supp. 2d at 1174 (emphasis in original).45 Because there is no support for a trademark infringement claim based on a failure to designate negative keywords, Respondent has failed to establish that protecting trademark rights justifies negative keyword agreements between competitors.

Given the inherently suspect nature of Respondent’s advertising restraints and Complaint Counsel’s more detailed showing of likely competitive harm to consumers in the particular context at hand, Respondent’s failure to establish a basis in fact for its asserted procompetitive justifications—a showing that they are valid as well as plausible and cognizable—provides a further basis for condemning its conduct. Even if there were no less restrictive alternatives, Respondent has not established that its anticompetitive restraints in fact have procompetitive virtues. We conclude that Respondent has engaged in unfair methods of competition in violation of Section 5 of the FTC Act.

The Dissent criticizes this Opinion for classifying the challenged restraints as inherently suspect. The Dissent asserts that we have not analyzed the challenged agreements under the rule of reason and therefore risk suppressing procompetitive conduct. These criticisms are misplaced. We rely on the Polygram framework because the challenged restraints are of a type that have been routinely condemned as inherently suspect, and Polygram furnishes a well-crafted framework for analyzing such restraints. But we also recognize that there may be plausible, cognizable justifications for trademark settlements. In fact, we consider Respondent’s specific evidence in support of those procompetitive justifications and ultimately find the evidence wanting. We also find that Respondent has less restrictive ways of accomplishing those procompetitive justifications and evaluate an extensive record of direct evidence showing anticompetitive effects. In short, though we find these restraints inherently suspect, we ultimately perform the “sedulous” analysis required under the rule of reason. See Cal. Dental, 526 U.S. at 781.

B. Analysis of the Challenged Agreements for Effects on Consumers Using Direct Evidence of Anticompetitive Effects Even if we did not rely on the inherently suspect nature of the restraints in the Challenged Agreements to conclude that there is liability, a second way independently to establish plaintiff’s initial burden to show that a particular horizontal restraint has anticompetitive effects is to consider direct evidence of those effects. When there is direct evidence of anticompetitive effects, detailed market analysis and proof of market power is unnecessary. “[S]ince the purpose of the inquiries into market definition and market power is to determine whether an arrangement has the potential for genuine adverse effects on competition, proof of actual detrimental effects, such as a reduction of output can obviate the need for an inquiry into market power, which is but 45 Respondent asserts that, on appeal, the “Tenth Circuit expressly did ‘not resolve [this particular] matter’” concerning whether bidding on generic keywords could be trademark infringement. RRB at 7 (quoting 1-800 Contacts, Inc., 722 F.3d at 1243) (brackets in RRB). This misconstrues the Tenth Circuit’s decision. What the appellate court expressly chose not to resolve was whether use of the challenged keywords alone could result in a likelihood of confusion. 1-800 Contacts, Inc., 722 F.3d at 1242-43. VOLUME 166 Opinion of the Commission a surrogate for detrimental effects.” IFD, 476 U.S. at 460-61 (internal quotation marks omitted); see also Ohio v. American Express Co., 138 S. Ct. at 2285 n.7 (explaining that market definition is unnecessary for the analysis of horizontal restraints when actual anticompetitive effects have been demonstrated).

When plaintiff satisfies the initial burden to show anticompetitive effects using direct evidence, the burden then shifts to the defendant. The defendant can challenge plaintiff’s support underlying the initial showing. In addition, defendant can seek to establish procompetitive justifications for its conduct. Realcomp II, Ltd. v. FTC, 635 F.3d 815, 825 (6th Cir. 2011). Ultimately, the fact-finder must consider whether the anticompetitive harms outweigh any procompetitive benefits.

1. Direct Evidence of Anticompetitive Effects for Consumers Like Judge Chappell, who considered the evidence only under this mode of analysis, we conclude that Complaint Counsel successfully established their prima facie case through direct evidence of two anticompetitive effects: the restriction of truthful advertising and an increase in contact lens prices sold online.

a. Restriction of Truthful Advertising Respondent and the Dissent argue that a restriction on truthful advertising does not qualify as an anticompetitive effect; according to Respondent, only reduced output or higher prices for the underlying product is sufficient. RAB at 22-23. Respondent and the Dissent rely on California Dental’s statement that “the relevant output for antitrust purposes here is presumably not information or advertising, but dental services themselves” so that “the question is not whether the universe of possible advertisements has been limited (as assuredly it has), but whether the limitation on advertisements obviously tends to limit the total delivery of [the product being advertised].” RAB at 22 (quoting California Dental, 526 U.S. at 776). But the Court’s concern in California Dental was that the normal linkage between advertising restrictions and price/output effects in the underlying product market was attenuated in the context of professional services because consumers may not be able to make valid assessments regarding advertising claims about the quality, comfort, or other non-price aspects of dentists’ services. See, e.g., 526 U.S. at 778 (citing the “plausibility of competing claims about the effects of the professional advertising restrictions” as the basis for concluding that “[t]he obvious anticompetitive effect that triggers abbreviated analysis has not been shown”). We find Respondent’s and the Dissent’s reliance on California Dental misplaced because there is no similar concern that consumers may be unable to assess the information contained in advertising for the sale of contact lenses. The record shows a focus on price advertising by many of 1-800 Contacts’ online rivals. See IDF 587, 591, 603, 611, 646; Holbrook, Tr. 1904. When consumers have a prescription and are shopping for contact lenses, the lenses they purchase are identical—by prescription, brand name, and even type (e.g., daily or biweekly)—regardless of the retailer. IDF 24-25. For such commodity products, consumers can comparison shop. In fact, the Fairness to Contact Lens Consumers Act, which requires prescribers to provide a patient 293 1-800 CONTACTS, INC. Opinion of the Commission with a portable copy of his or her prescription, “promotes competition in retail sales of contact lenses by facilitating consumers’ ability to comparison shop for contact lenses.” FTC Contact Lens Rule, 81 Fed. Reg. 88526 (Dec. 7, 2016) (review of Rule). Congress apparently had no concern that consumers would be unable to assess competing offers and prices for contact lenses. Restricting the availability of truthful information that guides consumer decisions in the marketplace is a competitive harm. As the Supreme Court explained in IFD, “a concerted and effective effort to withhold (or make more costly) information desired by consumers for the purpose of determining whether a particular purchase is cost justified is likely enough to disrupt the proper functioning of the price setting mechanism of the market that it may be condemned even absent proof that it resulted in higher prices or . . . the purchase of higher priced services than would occur in its absence.” IFD, 476 U.S. at 461-62. We similarly found direct evidence of competitive harm from a showing that there were “significantly fewer discount [residential real estate] listings” available to consumers after an association of real estate brokers adopted rules that limited consumers’ access to information about the availability of lower-priced real estate services. See Realcomp, 2007 WL 6936319.

The record demonstrates that the settlement agreements were effective in restricting advertisements from 1-800 Contacts’ rivals. As already described, parties to the agreements consistently testified they would bid on 1-800 Contacts’ trademark terms or remove the negative keywords if the agreements were not in place. See IDF 590, 595, 616, 630, 634-35, 650. Yet, data provided by Google covering the period January 2002 to September 2016 show that the competitors who had been bidding on 1-800 Contacts’ trademark terms ceased doing so almost entirely after entering the Challenged Agreements. IDF 687, 689 (citing CX8006 (Evans Expert Report) at 061-062). Similarly, the use of negative keywords by 1-800 Contacts’ competitors required by the Challenged Agreements prevented ads from being shown to consumers even when the competitor did not bid on 1-800 Contacts’ trademark terms; a second Google data set covering the period from January 1, 2010 to November 2016 showed a substantial decline in advertisements displayed in response to a search that includes a 1-800 Contacts’ trademark term through phrase match to a generic term such as “contacts.” IDF 655, 688, 690 (citing CX8006 (Evans Expert Report) at 056-057). Here, the negative keyword requirement forces 1-800 Contacts’ rivals to override the search engines’ determination that the rivals’ ads are relevant and valuable to consumers. See {Juda, Tr. 1153}.

Two models presented by Complaint Counsel’s experts predicted the but-for world without the advertising restrictions. Similar to the Google data, they show that the advertising restrictions here substantially reduce truthful advertising provided to consumers. Professor Susan Athey constructed a two-stage model of the but-for world. In the first stage, based on data from the current, actual world, a multinomial logistic regression model predicts consumer click behavior when a consumer conducts a Google search related to contact lenses. The model considers variables for the consumer appeal of the advertised brand, the position of the ad on the SERP, whether the ad is for the seller searched for by the consumer, whether the ad is for 1-800 Contacts, and the propensity of the consumer to click on any ad. Athey, Tr. 766-72. In the second stage, Dr. Athey constructed the ad layout that a consumer would be likely to see in response to a search for 1-800 Contacts if rivals were free to bid on such search terms. That ad VOLUME 166 Opinion of the Commission layout assumes that, without the advertising restraints in the Challenged Agreements, the SERP triggered by a search for 1-800 Contacts would be similar to the SERP triggered by queries such as “contact lenses” or “contacts.” Dr. Athey then applied the model of consumer click behavior from the first stage to the ad layout in stage 2. Dr. Athey’s model predicted that, absent the Challenged Agreements, the number of competitors’ ads appearing on a SERP would increase from 0.54 to 1.85 per search, IDF 749, and consumer clicks on those ads would increase by 3.5 clicks per 100 searches. IDF 750.46 Complaint Counsel’s expert, Professor David Evans constructed a model using a different data set and different methodology that produced results consistent with Dr. Athey’s findings. One of Dr. Evans’ empirical studies relies on the bidding experience of Memorial Eye, an online retailer that offered prices significantly lower than those of 1-800 Contacts. See IDF 693. Its advertisements to consumers heavily promoted its low pricing. Holbrook, Tr. 1904. Unlike most online competitors, Memorial Eye continued to advertise against 1-800 Contacts for several years after it was contacted by 1-800 Contacts and later sued. Thus, there is a data set showing the extent to which Memorial Eye ads appeared on SERPs generated by search queries for 1-800 Contacts’ trademark terms47 and whether those ad impressions led to consumer clicks for Memorial Eye. See CX8006 (Evans Expert Report) at 091-092. Based on the data for Memorial Eye, Dr. Evans projected the number of ads and clicks that would have resulted for the complete set of online rivals that were subject to the advertising restrictions. Dr. Evans’ model estimated 46 The Initial Decision lists criticisms of Dr. Athey’s model by Respondent’s expert, Dr. Anindya Ghose, ID at 158- 59, and, without substantive discussion, summarily concludes, “Although Respondent has identified some valid concerns regarding the underlying assumptions of . . . the Athey model . . . Respondent’s criticisms do not warrant the conclusion that the model [is] so faulty that [it] should be rejected entirely as unreliable.” ID at 160. The Initial Decision gives no indication which criticisms were valid and does not address Dr. Athey’s responses to the criticisms. We reject the ALJ’s conclusory statement.

Our substantive review of Dr. Ghose’s criticisms reveals that the concerns are not valid. Dr. Ghose criticized the model for using searches for generic terms as a proxy when creating ad layouts in the counterfactual world. In response to the criticism, Dr. Athey conducted reasonableness and robustness checks on modified ad layouts. Those checks show that the results Dr. Athey reported are robust, and actually are conservative. See CX8010 (Athey Rebuttal Report) at 033-035. Dr. Ghose also claims the appearance of ads by non-settling retailers in the counterfactual shows faulty assumptions. Dr. Athey explains that their appearance does not affect the results because the number of instances is not significant. Id. at 037-038. Dr. Ghose contends Dr. Athey’s model does not consider whether the settling parties increased advertising spending on generic searches when they could not bid on 1-800 Contacts’ trademark terms. Dr. Ghose’s suggestion is contrary to the evidence in this case; advertisers indicated that they bid based on the return on investment for each keyword rather than spending a fixed amount on search advertising regardless of the keywords that were permitted. See, e.g., CX9039 (Clarkson Dep.) at 176; {CX9024 (Holbrook Dep.) at 36-37; CX9014 (Batushansky Dep.) at 119-120}. Finally, Dr. Ghose criticizes the model for failing to include an analysis of additional conversions in the counterfactual world. A model need not estimate everything in order to be valuable; in particular, it need not quantify the number of additional conversions when it estimates the number of additional ad impressions. In any case, the record clearly demonstrates that online sellers obtain sales when they advertise on searches for 1-800 Contacts’ trademark terms. IDF 605, 611, 619, 644- 46, 714-16, 720.

47 Memorial Eye did not bid on 1-800 Contacts’ trademark terms; its ads were displayed in response to search queries for 1-800 Contacts’ trademark terms as a result of Memorial Eye bidding on generic terms such as “contacts” in broad match or phrase match. IDF 617.

295 1-800 CONTACTS, INC. Opinion of the Commission that, absent the Challenged Agreements, between January 2010 and June 2015, 114 million additional ads for competitors would have been displayed in response to queries containing 1- 800 Contacts’ trademark terms. IDF 755. The model also estimated that, absent the Challenged Agreements for the same period, clicks for 1-800 Contacts’ competitors’ ads would have increased by 145,000 and sales for the competitors would have increased by 12.3 percent. IDF 756.48 We find that this evidence directly shows that the Challenged Agreements were effective in restricting truthful advertising from being presented to consumers. The Google data showed that competitors largely ceased bidding on 1-800 Contacts’ trademark terms, and the but-for models from Drs. Athey and Evans predict the substantial number of ads that were not displayed. In addition, the models show that information those advertisements would have conveyed was valued by consumers who would have clicked on the ads and made additional purchases. Together, this evidence directly shows the Challenged Agreements cut off advertising in a way that interfered with the operation of competitive forces in the online sale of contact lenses and disrupted consumers’ mechanisms for comparing and selecting between alternative online sources.

Respondent and the Dissent also dispute that IFD finds that a restriction on truthful advertising is sufficient as evidence of actual anticompetitive effects. According to Respondent, in that case, the withholding of x-rays from insurance companies was an express restriction on output because x-rays were a service customers wanted. See RAB at 23. We disagree. X-rays were taken to assess the need for, and to guide the provision of dental treatment. X-rays were not offered as a separate product independent of dental treatment. Moreover, the Supreme Court’s analysis focused on the informational role of x-rays and the harm to market mechanisms 48 The ALJ’s assessment of Dr. Evans’ model was comparable to his assessment of Dr. Athey’s model. The Initial Decision lists critiques by Respondent’s expert, Dr. Ghose, but provides no substantive discussion or evaluation of those critiques and ignores Dr. Evans’ responses. The ALJ again summarily stated, “Although Respondent has identified some valid concerns regarding the assumptions of . . . the . . . model, Respondent’s criticisms do not warrant the conclusion that the model[] [is] so faulty that [it] should be rejected entirely as unreliable.” ID at 160. Again, we reject the Initial Decision’s conclusory analysis. A substantive review of the model and the critiques reveals that the criticisms provide no basis for finding the model unreliable. Dr. Ghose argues that {Memorial Eye} is not representative of other online sellers and it therefore was improper to extrapolate data for {Memorial Eye} to other online sellers. See RX0733 (Ghose Expert Report) at ¶¶ 161-64. Dr. Evans responds that the difference between Memorial Eye and other online sellers is that Memorial Eye did not implement negative keywords when it was threatened with litigation by 1-800 Contacts, whereas other online sellers did. Consequently, the differences between Memorial Eye and other online sellers identified by Dr. Ghose reflect this fact. The increased number of ads for Memorial Eye displayed by search engines compared to other sellers actually provides the basis for the analysis; it is not evidence that Memorial Eye is unrepresentative. See CX8009 (Evans Rebuttal Report) at 073-075. Dr. Ghose also claims the Evans model failed to account for ad activity found in the real world and, therefore, the model overstated the number of incremental impressions and clicks in the counterfactual world. We find there is insufficient evidence this occurred. Dr. Evans designed his methodology to estimate the activity of rival online sellers, and excluded ad impressions that were irrelevant, such as impressions from firms that do not sell contact lenses, i.e., companies that bid on “1-800” because they sell contact information for people and businesses via 1-800 telephone numbers. See CX8009 (Evans Rebuttal Report) at 076-077. VOLUME 166 Opinion of the Commission that would flow from withholding that information. See IFD, 476 U.S. at 461-62 (“A concerted and effective effort to withhold (or make more costly) information desired by consumers for the purpose of determining whether a particular purchase is cost justified is likely enough to disrupt the proper functioning of the price-setting mechanism of the market that it may be condemned even absent proof that it resulted in higher prices or, as here, the purchase of higher priced services, than would occur in its absence.”).

b. Increased Online Contact Lens Prices In addition to evidence of reduced advertising, Complaint Counsel presented direct evidence of a price effect, which provides a persuasive, independent basis for Complaint Counsel’s prima facie case.49 As the ALJ found, “the evidence in this case proves . . . that at least some consumers have paid, or will pay, prices that are higher than they would otherwise be, absent the Challenged Agreements.” ID at 153. As we previously discussed, the record contains evidence that the Challenged Agreements reduced the number of competitor ads, and increased sales for 1-800 Contacts while reducing the sales for its rivals. E.g., IDF 710-11, 717-19, 723, 725, 727, 730, 749-50 (citing Athey model’s results that without the Challenged Agreements, consumer clicks on competitor ads would increase by 3.5 clicks per 100 searches and clicks on 1-800 Contacts’ ads would decrease by 2 clicks per 100 searches), IDF 756 (citing Evans model’s result that absent the Challenged Agreements sales by competitors would have increased by 12.3 percent).

At the same time, prices charged by 1-800 Contacts were on average {15-20 percent} higher than those of its online competitors. IDF 692 (citing CX0547 at 032, in camera (1-800 Contacts document showing that prices from three major online rivals were {16 percent} lower than 1-800 Contacts’ prices in 2006 and {26 percent} lower than 1-800 Contacts’ prices in 2011); CX0295 at 063, in camera (showing that in January 2014, 1-800 Contacts’ prices were higher than other online contact lens retailers by {22 percent} per box, {15 percent} for a sixmonth supply and {14 percent} for a twelve-month supply; RX1228 at 036 (2015 analysis showing that 1-800 Contacts’ prices were higher than those of other online retailers; CX8007 (Athey Expert Report) at 013-014, 045-051, Exh. D-1 to D-7, in camera (calculating that 1-800 Contacts’ prices were {21} percent higher than online competitors’ prices, on average, for its top ten selling products for the period 2010 to 2016); see also CX8003 (Mitha Decl.) at ¶ 4 (“In general, 1-800 Contacts’ prices are higher than Lens Discounters’ by a significant amount. In the past, we have found that 1-800 Contacts’ prices were almost double Lens Discounters’ prices for some products.”).

On the facts of this case, we find the evidence that the Challenged Agreements insulate 1- 800 Contacts from normal competitive forces and divert sales from low-priced sellers to a high- 49 The Dissent avers that “actual, sustained, and substantial or significant price effects” are required to meet the burden of showing direct evidence of anticompetitive effects. Dissenting Statement at 30. The Dissent claims the direct evidence presented does not meet the legal standard and goes on to recount a list of ways in which a direct effects showing can be satisfied. The Dissent’s view of direct effects evidence, however, is unduly cramped. Courts have recognized instances where parties colluded to withhold information (see, e.g., IFD, 476 U.S. 447) and cases where an agreement prevents consumers from gaining access to lower-costing alternatives (see, e.g., Realcomp, 635 F.3d at 831-34). The record here supplies more than enough evidence to clear that bar. 297 1-800 CONTACTS, INC. Opinion of the Commission priced seller is direct evidence of an increase in price. The higher prices that consumers are paying do not reflect a producer selling a differentiated product, such as a product with new technology or additional features that offer more than the products of low-priced sellers. Instead, the higher prices are a consequence of 1-800 Contacts shielding itself from competitive pressure by preventing consumers from obtaining information that would enable comparison shopping. The economic principles and evidence regarding consumer search previously discussed, see supra Section V.A.1, provide the explanation. The record shows that many consumers are unaware of the price difference between 1-800 Contacts and its online competitors. IDF 694 (citing RX1228 at 36 (based on a consumer survey, AEA analysis stated, “Actual price variances [are] much more than perceived price variances”)). Restricting the advertising presented to such consumers at the critical time when they are about to make a purchase impedes their ability to compare prices, which leaves them unaware of alternatives to 1-800 Contacts’ higher-priced products.

Further evidence that the Challenged Agreements had actual price effects comes from 1- 800 Contracts’ price-matching policy, whereby it offered to meet or beat any price offered by online, or certain other, rivals. See IDF 436 (in 2011 1-800 Contacts’ ad copy stated “We Beat Any Online Price”), 437 (referencing 1-800 Contacts’ policy in 2014 to meet or beat rivals’ prices), 438 (quoting 1-800 Contacts’ 2016 policy stating, “We’ll beat any price on every product we carry by 2%”). But to take advantage of the price matching policy, a customer had to contact 1-800 Contacts. IDF 439. By reducing rivals’ ads and consumer clicks on those ads, the settlement agreements necessarily reduced access to the type of information that consumers needed to trigger 1-800 Contacts’ price matches.50 The Challenged Agreements thus directly interfered with consumers’ ability to trigger discounts.

2. 1-800 Contacts’ Response to the Direct Evidence of Anticompetitive Effects Because Complaint Counsel have demonstrated anticompetitive effects, the burden now shifts to Respondent. 1-800 Contacts challenges the factual support for the direct evidence of anticompetitive effects and proffers procompetitive justifications for the restraints. a. Respondent’s Challenges to the Direct Evidence Respondent and the Dissent challenge the direct evidence of anticompetitive effects on several grounds. First, Respondent argues that Complaint Counsel have presented only the theories of experts, not direct evidence of price effects. We reject this characterization. The opinions of Complaint Counsel’s experts derive from the facts in the record and econometric 50 In fact, many 1-800 Contacts customers are unaware that other online contact lens sellers exist. See IDF 697 (citing RX0041 at 0019 (in consumer survey prepared for Berkshire Partners, which was considering acquisition of 1-800 Contacts in 2012, more than one-third of respondents said they initially purchased from 1-800 Contacts because it was the only online site the consumer was aware of), 698 (due diligence for Berkshire Partners in 2012 concluded, “1-800 likely benefits from a sizable segment of uninformed buyers who are simply unaware of the other (and growing) low-priced choices on the internet.”). Display of rivals’ search ads would tend to counter this ignorance.

VOLUME 166 Opinion of the Commission analysis of those facts. The experts use known facts to quantify the impact of the advertising restrictions on the ads that would otherwise appear and on the consumer responses—including clicks and purchases—thereto. They provide empirical evidence, not economic theory isolated from facts, and the underlying facts are in the record.

Respondent and the Dissent next challenge the premise of higher prices, arguing that 1- 800 Contacts offers a higher quality of service, so there is no reason to conclude that its prices are higher on a quality-adjusted basis. RAB at 25-26. Certainly, customer service can be a differentiating factor when a firm sells a commoditized product. See CX8007 (Athey Expert Report) at 015. But the record shows that without the Challenged Agreements, consumers would have shifted purchases from 1-800 Contacts to its rivals, which reveals customer preferences for the price/quality combination offered by rivals.51 At least for these customers, 1-800 Contacts was offering a higher price, even after adjusting for quality. Other aspects of the record show that 1-800 Contacts’ service levels do not fully explain its higher prices. Professor Athey testified that “[D]direct facts and market data support that there is a price premium [for 1-800 Contacts] and that that price premium is not fully accounted for by service differentials.” IDF 740 (quoting Athey, Tr. 797). This testimony reflects numerous market facts.

Other online sellers judge that they offer comparable service to 1-800 Contacts. See, e.g., {CX9000 (Batushansky I.H.) at 65 (Web Eye Care president testified, “we offer the same great products and we feel that our service is on par with theirs”)}; CX9039 (Clarkson Dep.) at 88 (AC Lens president testifying that AC Lens is “pretty fanatical about service, by trying very hard to make the process convenient and quick, . . . getting orders shipped the day they arrive” and having net promoter scores “consistent with the highest on the internet”). The competitors’ view of service levels was shared by independent evaluators. The investment memorandum prepared by Berkshire Partners as part of the consideration of 1-800 Contacts stated, “[W]e are concerned that 1-800’s premium pricing positioning versus its competitors is unsustainable in the medium- to long-term given the commodity-like nature of contact lenses and 1-800’s insufficiently distinguishable service.” CX1109 at 003. Other evidence supports the conclusion that 1-800 Contacts’ higher prices are not fully explained by the firm’s service level. Some statements by 1-800 Contacts’ employees express doubt that its service level is sufficient to justify the price premium. See, e.g., CX1086 (email expressing concern that ads by lower priced competitors would lead to reduced 1-800 Contacts sales; comment in the email chain states, “The only other option I see is trying to convince customers that our existing prices are better than they really are or worth the cost. Tough challenge considering that we sell the exact same thing as everyone else.”). Similarly, some of 51 Indeed, internal 1-800 Contacts documents suggest that once consumers make a purchase from another online retailer, they are unlikely to make their next purchase from 1-800 Contacts. Based on a small sample of 54 consumers whose most recent purchase was from another online retailer, 17 percent reported that they were likely to make their next purchase from 1-800 Contacts and 71 percent reported that they were not likely to do so. See CX1117 at 023. Thus, the customer service differential asserted by 1-800 Contacts did not support a return of customers who had purchased from another online contact lens retailer. 299 1-800 CONTACTS, INC. Opinion of the Commission 1-800 Contacts’ documents question the firm’s supposed quality advantage. See CX1117-022 (“Other online suppliers achieve satisfaction scores as high as us”). Finally, the need for 1-800 Contacts to offer a price-match policy suggests that the service differential is insufficient to offset the price premium.

Respondent and the Dissent also argue that Complaint Counsel have not shown that 1- 800 Contacts’ price was supracompetitive. RAB at 22. We find Complaint Counsel’s showing sufficient. Proof of an anticompetitive effect does not require an econometric model to estimate a precise competitive price in order to establish that the existing price is supracompetitive. Complaint Counsel have, in fact, shown that the price consumers paid was higher with the Challenged Agreements than it would have been had the market been allowed to function without the advertising restraints. In addition to the direct evidence of actual price effects discussed above—the diversion of purchases from low-priced rivals to 1-800 Contacts and the withholding of information needed to trigger 1-800 Contacts’ price match—Dr. Athey testified that “if consumers become more informed, it will be difficult [for 1-800 Contacts] to sustain a price premium and . . . they would thus face a choice, either lose market share in the online channel, and particularly in the search channel, or lower their price. . . . [M]ore likely than not, prices – prices would fall. It’s also possible that [1-800 Contacts] could keep their prices high and – but consumers would use more price match, which would lead to a reduction in the effective price by 1-800 even if the list price stayed high.” Athey, Tr. 797-98 Similarly, Respondent and the Dissent argue that because 1-800 Contacts’ profit margins {remained constant}, even as the number of settlement agreements increased, the evidence contradicts an inference that the agreements raised prices to supracompetitive levels. RAB at 22, 26. We disagree. As an initial matter, measuring profit margins in an economically meaningful manner is difficult, and Respondent’s assertion gives us no basis to conclude they were properly measured. Moreover, {constant} margins do not necessarily mean prices did not rise; without competitive pressures, costs may have risen as prices increased, {keeping margins constant}. Finally, if 1-800 Contacts started with a profit margin reflecting supracompetitive prices, there is no reason to expect its margin to increase. In fact, 1-800 Contacts was the incumbent online seller, with a dominant share of online sales throughout this period. See IDF 69; CX0055-009 (2004 1-800 Contacts strategy memo identifying “Market Leadership” as a strength and stating that 1-800 Contacts “leads US phone/internet retail market” in “size” and has “20 x unaided brand awareness of online competitors”); CX0526-007; Coon, Tr. 2668-70. Consequently, 1- 800 Contacts’ {constant} profit margin is consistent with a conclusion that the Challenged Agreements prevented the growth of online rivals when they entered the market, thus preventing the erosion of 1-800 Contacts’ supracompetitive margins.

Consequently, we find that Complaint Counsel have established a prima facie case of anticompetitive harm through direct evidence of the restriction of truthful advertising and through direct evidence of price increases.

VOLUME 166 Opinion of the Commission b. Respondent’s Procompetitive Rationales for the Advertising Restraints Respondent may rebut Complaint Counsel’s prima facie case by establishing procompetitive justifications that outweigh the anticompetitive harms. Respondent has identified two justifications—the settlement of costly litigation and trademark protection—that we have found cognizable and facially plausible. See supra Section V.A.2. But, as discussed above in our analysis of the challenged restraints as inherently suspect, Respondent fails to sufficiently support its asserted justifications, and Complaint Counsel have demonstrated that the challenged advertising restraints are not reasonably necessary to achieve the asserted benefits. See supra Sections V.A.5 and V.A.3.a. In these circumstances, direct evidence of anticompetitive harm provides a second, independent basis for concluding that Respondent has engaged in unfair methods of competition in violation of Section 5 of the FTC Act. C. Analysis of the Challenged Agreements for Effects on Search Engines In addition to harm to consumers, the Complaint alleges that the Challenged Agreements harm search engines by, inter alia, unreasonably restraining price competition in certain search advertising auctions, preventing search engine companies from displaying to users the array of advertisements that are most responsive to a user’s search, and impairing the quality of service provided to consumers by search engine companies. Compl. ¶ 31a-d. Despite the allegations in the Complaint and the presentation of evidence on the issue, and contrary to Commission rules,52 Judge Chappell determined that the “Initial Decision need not, and does not, . . . determine whether or not the Challenged Agreements have anticompetitive effects in the form of harm to search engines.”53 ID at 166.

Our review of the record reveals that Complaint Counsel have presented a prima facie case of anticompetitive harm to search engines based on direct evidence of actual harm.54 52 Commission Rule of Practice 3.51(c) states “The initial decision shall include a statement of findings of fact . . . and conclusions of law, as well as the reasons or basis therefor, upon all the material issues of fact, law, or discretion presented on the record . . .” 16 C.F.R. §3.51(c).

53 The ALJ’s decision not to address an independent theory of liability based on effects for search engines is particularly troubling because Judge Chappell omitted provisions from the proposed order that addressed “conduct, such as price-fixing and market allocation.” The ALJ reasoned that the provisions are “too far removed from the unlawful conduct found to exist in this case to conclude that the provisions . . . are justified as reasonably related, fencing-in provisions.” ID at 195-96. The deleted provisions addressed conduct related to the allegations regarding search engines. They are unrelated to the unlawful conduct found by the ALJ only because he failed to address all of the Complaint’s allegations. Complaint Counsel have not requested restoration of the deleted provisions, and in any event, we believe the Order without those provisions provides an effective remedy to harm against search engines.

54 Alternatively, we could evaluate the Challenged Agreements under the inherently-suspect framework. For a restraint that “operates as an absolute ban on competitive bidding,” “no elaborate industry analysis is required to demonstrate the anticompetitive character of such an agreement.” Natl Soc’y of Prof’l Eng’rs v. United States, 435 U.S. 679, 692 (1978). Economic learning clearly shows that cooperative bidding strategies among rivals impair competition, by raising what they can charge for goods or services or reducing what they pay when bidding to buy from a third party. See CX8006 (Evans Expert Report) at 070 & n.167. Indeed, in many contexts, bid rigging may 301 1-800 CONTACTS, INC. Opinion of the Commission Absent a valid procompetitive justification, this provides a third, independent basis to find liability in this case.

Under the terms of the Challenged Agreements, 1-800 Contacts and its online rivals agreed to refrain from bidding in particular search-advertising auctions. Online rivals agreed not to bid when the consumer’s search is for 1-800 Contacts’ trademark terms, and 1-800 Contacts reciprocally agreed not to bid on the trademark terms of its rivals. The Challenged Agreements thus reduce the number of bidders participating in the auctions because the parties have agreed not to compete.

The record shows that the Challenged Agreements resulted in actual harm to search engines. Witnesses from both Google and Bing explained that a reduction in the number of search-advertising auction participants offering relevant ads55 reduces the price paid by the auction winners and reduces the revenue for the search engine. Google’s Director for Ads Quality testified that when advertisers that previously appeared on the SERP stop appearing, {“they would be reducing auction pressure or pricing pressure on the ads that remained, so ads that remained on the page would start paying less per click because there’s less competition for the ad position that it just won.”} Juda, Tr. 1157, in camera. Bing’s partner scientist in charge of Bing Ads similarly stated, {When relevant bidders agree to drop out of a search advertising auction, it may reduce the competitiveness of the auction, and the prices and revenues that result from our [generalized second-price] auction model. . . . [B]ecause each advertiser does not pay the amount of its maximum bid, but rather just enough to keep its position in Bing’s ranking of advertisements, removing even one relevant bidder from an auction can reduce the price paid by one or be condemned as a per se offense. See, e.g., United States v. MMR Corp., 907 F.2d 489, 496 (5th Cir. 1990) (finding no conceptual distinction between bidding high and “backing away from bidding” as means for carrying out a potentially per se illegal agreement to rig bids); COMPACT v. Metro. Govt, 594 F. Supp. 1566, 1575-77 (M.D. Tenn. 1984).

As we previously discussed, Respondent has advanced legitimate procompetitive justifications, which, under the inherently-suspect framework, trigger a need for consideration of less anticompetitive alternatives to achieve the proffered procompetitive justifications or further factual findings and analysis regarding the likelihood of anticompetitive effects in the particular context. Respondent also has the burden of showing that the restraints in fact have the asserted procompetitive virtues. We already have addressed the availability and workability of alternative settlement terms that would be less restrictive of competition, see supra Section V.A.3.a., as well as Respondent’s failure to show that its asserted justifications are valid, not merely plausible. See supra Section V.A. 5. A further showing of the likelihood of anticompetitive effects in the particular context would involve assessing the evidence that we discuss in the text below. Consequently, both modes of analysis rely on the same evidence, and we limit our full exposition regarding anticompetitive effects for search engines to the latter mode of analysis. 55 Under the second-price auction used by search engines, if additional bidders enter the auction, but all of them have ads determined by the search engine algorithm not to be sufficiently relevant to consumers, then the increased number of bidders would not affect the price paid by the highest ranked advertiser. If some of the bidders have sufficient relevance (i.e., a higher second highest AdWords score), the price paid by the advertiser with the first position would be higher. IDF 219.

VOLUME 166 Opinion of the Commission more other bidders. . . . Further, . . . Bing observes that, when a particular keyword attracts more competition, advertisers often increase their bids. If advertisers do not increase their bids to meet new competition, their ads will often fall to a lower position on the page, or even fall off the page entirely. Bing observes that some of the affected advertisers will raise their maximum bids to preserve their positions on the SERP.} CX8005 (Iyer Decl.) at 006, in camera; IDF 243.

Here, we know the ads are sufficiently relevant to affect prices. During the period from 2002 through 2016, Google served advertisements for nine of the fourteen online contact lens sellers who were parties to the Challenged Agreements based on those firms’ bids on the 1-800 Contacts trademark terms before they entered into the Challenged Agreements. IDF 653. This demonstrates that Google determined the ads were sufficiently relevant to be displayed, which indicates the ads would have affected the cost-per-click prices charged to the advertisers. Juda, Tr. 1151, in camera ({“any ad that appears on the search engine result page almost by definition is going to be impacting the CPC [cost-per-click] of the advertiser above it, . . . [T]ypically ads that appear on the search page by definition influence the CPC [cost-perclick] of other ads on the page, the lone exception being the very first ad on the page.”}); see also IDF 219 (describing the price effect of an additional bidder in a second-price auction used by search engines).

The record contains direct evidence of these price and revenue effects. 1-800 Contacts’ internal documents acknowledge that one effect of the Challenged Agreements was reduced search advertising costs. A 2009 email from 1-800 Contacts’ former Senior Search Marketing Manager explained that one part of 1-800 Contacts’ “[t]rademark keyword management process” was to “[e]nforce trademark policy to remove competitors which in turn drives down how much we pay per click.” CX0935; see also CX0051 at 007 (Presentation on Search Overview describing bid management for trademarks: “• Keep competitors & affiliates off • Low competition = low price”); CX0658 at 001 (weekly marketing report stating, “Compared with recent weeks, we saw fewer competitors showing on our [trademark] keywords this week, which helped drop our spend for these terms.”); CX0915 (July 28, 2008 email from 1-800 Contacts’ Senior Search Marketing Manager stating, “TM CPCs [trademark costs-per-click] . . . jumped up by 18% from last week and pushed us to our most costly week yet for trademarks. There were more advertisers on our marks this week (both local and national retailers), which increased competition and CPCs [costs-per-click] for our top terms.”). Dr. Evans’ model, which estimated the net change in the number of rival ad impressions that would have been shown without the advertising restrictions, showed that the bidding restrictions in the Challenged Agreements reduced 1-800 Contacts’ cost-per-click on its trademark keywords. The model estimated that the agreements reduced the prices paid by 1-800 Contacts by {49 to 59} percent. CX8006 (Evans Expert Report) at 076 ¶ 168; Evans, Tr. 1648- 50, in camera. Dr. Evans concluded that “[t]he empirical analysis of the impact of the agreements on 1-800 Contacts’ costs of bidding on its [brand name keywords] confirms” that “agreements among competitors not to enter into auctions would have a material impact on 303 1-800 CONTACTS, INC. Opinion of the Commission price.” CX8006 (Evans Expert Report) at 077 ¶ 169. The lower prices paid by 1-800 Contacts are a result of agreements with its competitors not to bid at auctions, and cause a competitive injury to the search engine.56 The Challenged Agreements also harm both the search engines and consumers by removing advertisements that otherwise would have been displayed, thereby decreasing the quality of the search engines’ product. Search engines seek to show the most relevant ads to consumers; after all, search engines receive payment only when a consumer clicks on an ad. Juda, Tr. 1072. Having access to a larger number of relevant ads allows search engines to better fill SERPs with relevant ads that are valued by consumers. Bing’s partner scientist in charge of Bing Ads explained that reducing the number of bidders {could reduce the quality of the advertisements that Bing can show our users. . . . In general, the greater the number of bidders, the more choices Bing has to fill the available advertising space on the SERP. If the number of bidders is reduced, Bing may not have enough relevant, quality ads to fill the available space. This also reduces the number of choices Bing is able to present to its users. Even if Bing has enough advertisers to fill the ad space, the removal of a relevant advertiser may result in a reduction of the average quality of the ads shown to the user[.]} CX8005 (Iyer Decl.) at 005 ¶¶ 31-32, in camera.

Dr. Evans’s model estimated that without the Challenged Agreements, Google would have displayed more than 100 million additional ads between January 2010 and June 2015. CX8006 (Evans Expert Report) at 010. Dr. Evans concluded that this reduction in the number of relevant ads displayed reduced the quality of the product offered by search engines and diminished the value of search engine service to consumers. Id. at 078. Dr. Athey’s model similarly showed that many additional ads would have been displayed to consumers if the Challenged Agreements were not in place—with the number of competitor ads per search on 1- 800 Contacts’ trademark terms more than tripling. IDF 749; CX8010 (Athey Rebuttal Report) at 072. Her model also showed that those additional ads were valued by consumers; the model showed that consumers would have increased their clicks on competitors’ ads. See CX8007 (Athey Expert Report) at 029-034.

Consequently, we find that Complaint Counsel have satisfied their initial burden and established a prima facie case of anticompetitive harm to search engines through direct evidence of reduced auction prices and reduced quality of SERPs presented to consumers. The burden now shifts to Respondent. Here, Respondent challenges the factual basis underlying the direct 56 Contrary to Respondent’s argument that Complaint Counsel failed to prove anticompetitive harm to search engines because Complaint Counsel failed to define a relevant antitrust market for paid search advertising, RRB at 18-19, proof of actual detrimental effects does not require market definition or proof of market power. See IFD, 476 U.S. at 460-61; American Express, 138 S. Ct. at 2285 n.7; see also supra Section V.B. VOLUME 166 Opinion of the Commission evidence of anticompetitive effects. In addition, Respondent again proffers its procompetitive justifications.

Respondent argues that the price effects for search engines occur only if “all other things [are] equal.” RRB at 19. Respondent argues that Complaint Counsel failed to demonstrate any impact on search engines’ revenue because the bidding restrictions would merely have caused advertisers to shift their bids to other keywords. Respondent would have us assume that denying advertisers access to their first-choice of keywords and forcing them to turn to what they consider less desirable alternatives has no effect on their search advertising spending and no effect on the quality of search engine results. The record contradicts Respondent’s argument. 1- 800 Contacts paid less per click as a result of the Challenged Agreements. See, e.g., CX0935; CX0051 at 007; CX0658 at 001; CX0915. Also, advertisers indicated that they bid based on the return on investment for each keyword rather than spending a fixed amount on search advertising regardless of the keywords that were permitted. See, e.g., CX9039 (Clarkson Dep.) at 176; {CX9024 (Holbrook Dep.) at 36-37; CX9014 (Batushansky Dep.) at 119-120}. Preventing 1- 800 Contacts’ online rivals from bidding on their first choices for keywords, leaving them to bid only for keywords that they value less, reduced those retailers’ demand for search advertising, reduced their purchases of search advertising, and reduced the search engines’ revenues. While avoiding litigation costs through settlement and trademark protection are cognizable and facially plausible justifications, see Section V.A.2, reliance on those justifications falters for the reasons articulated above. See supra Sections V.A.3.a and V.A.5. Consequently, without an offsetting, valid procompetitive justification, the anticompetitive harm to search engines caused by the Challenged Agreements is a further, independent basis for concluding that Respondent has engaged in unfair methods of competition in violation of Section 5 of the FTC Act.

VI. REMEDY The Commission is empowered to enter an appropriate order to prevent a recurrence of the violation. 15 U.S.C. § 45(a)(2); FTC v. Colgate-Palmolive Co., 380 U.S. 374, 395 (1965) (the Commission is permitted “to frame its order broadly enough to prevent respondents from engaging in similarly illegal practices” in the future). It has considerable discretion in fashioning an appropriate remedial order, so long as the order bears a reasonable relationship to the unlawful conduct found to exist. See FTC v. Natl Lead Co., 352 U.S. 419, 428 (1957); Jacob Siegel Co. v. FTC, 327 U.S. 608, 611, 613 (1946). “The Commission is not limited to prohibiting the illegal practice in the precise form in which it is found to have existed in the past,” but “must be allowed effectively to close all roads to the prohibited goal, so that its order may not be by-passed with impunity.” FTC v. Ruberoid Co., 343 U.S. 470, 473 (1952). To remedy Respondent’s violation of Section 5, the ALJ issued an Order that bars 1-800 Contacts from agreeing with any seller of contact lens products to limit participation in online search advertising auctions (including restricting the use of keywords or requiring the use of negative keywords) or to limit online search advertising. ID at 203 (ALJ Order Paragraph II) . The ALJ’s Order contains a carve-out clause regarding future litigation. The carve-out confirms that that Order does not prohibit Respondent from initiating or prosecuting a lawsuit; 305 1-800 CONTACTS, INC. Opinion of the Commission communicating to any seller its intention to initiate or prosecute a lawsuit; or implementing or enforcing an order entered by any court of law, including an order approving a litigation settlement. Id. The ALJ’s Order also requires Respondent to cease enforcing existing agreements that are inconsistent with the Order. ID at 203 (ALJ Order Paragraph III). The ALJ’s Order contains a number of notification requirements in connection with Respondent’s future litigation and settlements. ID at 203 (ALJ Order Paragraph IV). Respondent argues that the ALJ’s Order encroaches on Article III courts’ authority to enforce the existing settlements. It asks the Commission to delete all restrictions in the ALJ’s Order on continued judicial enforcement of the existing settlements, while only barring 1-800 Contacts from entering into similar agreements in the future without judicial approval. RAB at 42-43. Respondent also argues that the ALJ’s Order violates 1-800 Contacts’ Fifth Amendment rights by retroactively depriving it of the ability to enforce its trademark rights, in violation of the Due Process and Takings Clauses. RAB at 43-45.

Complaint Counsel also ask the Commission to modify the ALJ’s Order. CCB at 47-50. They urge the Commission to restore the original language that they had proposed for the careout and that the ALJ subsequently changed. Specifically, they would remove the language that provides that the ALJ’s Order does not prohibit Respondent from implementing or enforcing the order entered by any court of law, “including an order approving a litigation settlement,” and would replace this with language providing that the Commission’s Order does not prevent Respondent from implementing or enforcing the order issued by any court of law “at the conclusion of a contested litigation.” Id. at App. B ¶ II.A-B (emphasis omitted). A. Enforcement of the Challenged Agreements Respondent asserts that the ALJ’s Order improperly trespasses on Article III courts’ authority. We disagree. The ALJ’s Order restricts Respondent from enforcing or attempting to enforce the requirements in an existing agreement or court order that are inconsistent with the remedial provisions imposed by the Commission.57 This does not direct or limit a court; it only restrains 1-800 Contacts. Our challenge here has focused on 1-800 Contacts’ conduct in entering and policing private agreements, and our remedy governs 1-800 Contacts’ conduct in continuing to enforce those agreements. The fact that a small number of 1-800 Contacts’ private agreements have been embodied in consent orders does not remove them from our administrative review; the private agreements they entail remain subject to antitrust scrutiny and the Commission’s remedial authority. Cf. Local No. 93, Intl Assn of Firefighters v. City of Cleveland, 478 U.S. 501, 519-22 (1986) (distinguishing giving effect to an obligation created by litigants’ private agreement from giving effect to the power of federal courts unilaterally to impose that obligation); In re Lipitor Antitrust Litig., 868 F.3d 231, 264-66 (3d Cir. 2017) (incorporation into a consent order of a private settlement agreement did not inoculate it from antitrust scrutiny under Noerr-Pennington principles).

57 The Order also requires 1-800 Contacts to take whatever action is necessary to vacate or nullify the provisions in its existing agreements or court orders that are inconsistent with the Commission’s remedial order. VOLUME 166 Opinion of the Commission Moreover, Respondent’s proposed modification of the ALJ’s Order would allow it to continue enforcing restrictions that already have been found unlawful under the FTC Act. Provisions that have been found to violate the antitrust laws are unenforceable. See Kaiser Steel Corp. v. Mullins, 455 U.S. 72, 80 (1982) (defense to an action based on contract is appropriate “where the judgment of the Court would itself be enforcing the precise conduct made unlawful by the Act”) (quoting Kelly v. Kosuga, 358 U.S. 516, 520 (1959)). Moreover, under the FTC Act, the Commission “is directed to prevent persons, partnerships or corporations . . . from using unfair methods of competition in and affecting commerce,” 15 U.S.C. § 45(a)(2) (emphasis added), and, upon finding a violation, “shall issue and cause to be served on [the respondent] an order requiring such person, partnership or corporation to cease and desist from using such method of competition . . . .” 15 U.S.C. § 45(b) (emphasis added). Given that the Commission has found that Respondent’s agreements violate the FTC Act, an order directing Respondent to cease enforcing the unlawful provisions is consonant with, and indeed integral to, the governing statutory scheme.

Respondent also claims that the Order violates its Fifth Amendment rights. Specifically, Respondent asserts that condemnation of the Challenged Agreements establishes a new trademark rule, and retroactive application of that rule to 1-800 Contacts’ settled lawsuits is inequitable and violates the Due Process and Takings Clauses. There are a number of problems with this argument.

First, we are not establishing a new trademark rule; indeed, we make no ruling on any trademark issue at all. We hold only that, based on our assessment of existing trademark case law, Respondent has not presented sufficient evidence to establish the validity of a procompetitive benefit that might outweigh the anticompetitive harm of the Challenged Agreements, and that any such benefit could have been achieved by less anticompetitive means. Second, the Order does not apply retroactively. It does not levy fines, determine damages, or impose any other sanctions for Respondent’s entry into and prior enforcement of the Challenged Agreements. Rather, the Order prohibits 1-800 Contacts from enforcing existing settlements in the future and from entering into new agreements containing the unlawful terms. Injunctive relief is inherently forward-looking. That it arises from past conduct does not render it retroactive. See Landgraf v. USI Film Prods., Inc., 511 U.S. 244, 273–74 (1994) (“When the intervening statute authorizes or affects the propriety of prospective relief, application of the new provision is not retroactive. . . . [R]elief by injunction operates in futuro.”) (quotation marks omitted); Russell v. Dunston, 896 F.2d 664, 668 (2d Cir. 1990) (the fact that prospective relief arises out of a past injury does not render an otherwise forward-looking injunction retroactive).58 Third, the Order is not novel, either in substance or in effect. It should not surprise Respondent that its agreements with competitors to restrict advertising and bidding were subject to an antitrust challenge. Antitrust has long barred rivals’ agreements regarding advertising and 58 Even as to the future, the Order preserves 1-800 Contacts’ ability to defend its trademark rights. The Order states, “[N]othing in [Paragraphs II.A or II.B] shall prohibit Respondent from . . . initiating or prosecuting a lawsuit . . . .” Final Order ¶¶ II. A-B.

307 1-800 CONTACTS, INC. Opinion of the Commission bidding restrictions. See supra Sections V.A.1.a and V.A.3. In fact, the District Court that rejected 1-800 Contacts’ trademark claims against Lens.com gave Respondent a clear warning in 2010: “Were this actually an agreement entered into by the parties, the court questions whether it would survive an antitrust challenge.” 1-800 Contacts, Inc. v. Lens.com, Inc., 755 F. Supp. 2d at 1188.59 Moreover, remedies requiring defendants to reverse an unlawful course of conduct, even if the defendant’s circumstances have changed, are common. See, e.g., Promedica Health Sys., Inc., 2012 WL 2450574, at *66-67 (F.T.C. June 25, 2012) (ordering divestiture after consummated merger notwithstanding the costs of unwinding already-consolidated services), petition for review denied, 749 F.3d 559 (6th Cir. 2014). Indeed, the Supreme Court confirms that “both within the settlement context and without, the Court has struck down overly restrictive . . . agreements.” Actavis, 570 U.S. at 150.

Respondent asserts, in effect, that it has a constitutional right to continue to enforce illegal agreements in perpetuity. It does not. As the Supreme Court stated: Federal regulation of future action based upon rights previously acquired by the person regulated is not prohibited by the Constitution. So long as the Constitution authorizes the subsequently enacted legislation, the fact that its provisions limit or interfere with previously acquired rights does not condemn it. Immunity from federal regulation is not gained through forehanded contracts. Fleming v. Rhodes, 331 U.S. 100, 107 (1947). We reject Respondent’s arguments that the Order violates the Fifth Amendment.

B. Enforcement of Future Court Orders As initially proposed by Complaint Counsel, the remedial order included a provision specifying that nothing in the subparagraph that bars 1-800 Contacts from agreeing with any seller of contact lens products to limit online search advertising prohibits Respondent from “implementing or enforcing the order entered by any court of law at the conclusion of a contested litigation.” The ALJ changed this carve-out to specify that nothing in the subparagraph prohibits Respondent from “implementing or enforcing the order entered by any court of law, including an order approving a litigation settlement.”60 Complaint Counsel contend that this modification permits recurrence of the very conduct found in this proceeding to be unlawful: “1-800 can file lawsuits, exact the same agreements with rivals, and place them before a court—where they will likely be approved.” CCB at 48. Respondent argued to the ALJ that the original language—limiting the exemption to court orders entered at the conclusion of 59 Respondent was aware even earlier that antitrust considerations might preclude enforceability of its settlements. In a letter dated January 7, 2008, Vision Direct’s counsel wrote to 1-800 Contacts’ trademark counsel to express “serious concerns regarding the enforceability of the Agreement, particularly as it relates to the implementation of negative key words” because such an agreement “creates an unacceptable risk of violating . . . Section 1 of the Sherman Act.” CX0141-001.

60 The ALJ also attached the same proviso to Subparagraph II.A, which bars 1-800 Contacts from agreeing with any seller of contact lens products to limit participation in search advertising auctions. VOLUME 166 Opinion of the Commission contested litigation—would interfere with the ability of Article III courts to issue orders approving settlements and dismissing litigation. ID at 193. We find Complaint Counsel’s concerns overstated. The ALJ’s carve-out allows “implementing or enforcing court orders.” It does not detract from the Order’s prohibition against entering agreements with sellers of contact lens products to limit participation in search advertising auctions or to limit search advertising. Moreover, the ALJ modified the proposed order in a second way, designed to mitigate concerns that courts will issue anticompetitive decrees: he added a provision requiring Respondent to “[p]rovide a copy of this Order to any court evaluating a request that a litigation settlement agreement relating to Search Advertising be approved by the court and/or incorporated into a court order.” ID at 203 (ALJ Order Paragraph IV.B.5). The ALJ’s Order thus preserves Respondent’s ability to implement court orders while ensuring that courts are made aware of the possible anticompetitive consequences before their orders are entered. Nevertheless, in addition to the ALJ’s modification, we also require Respondent to notify the Commission ten days before entering any stipulated order with a court and submitting a copy of the order at the time of notification. Such a notification provision will enable the Commission to intervene and apprise the court of any anticompetitive harm arising out of any stipulated order entered into by the Respondent. We find that the ALJ’s Order in conjunction with our notification provision confers adequate protection.61 VII. CHALLENGES TO THE LEGITIMACY OF THE FTC’S ENFORCEMENT PROCEEDING Finally, Respondent advances two arguments that contend that aspects of this enforcement proceeding lack legitimacy. One contention is that the Commission lacks a quorum. RAB at 46. That argument was advanced during a period when the FTC had two sitting Commissioners. Subsequently, however, additional Commissioners have joined the Commission, and the FTC currently has its full complement of five Commissioners to address this appeal. Respondent’s quorum arguments are therefore moot. Respondent also maintains that this proceeding is unconstitutional because “it was conducted by an ALJ, an ‘inferior Officer[]’ of the United States that Congress has improperly insulated from control by the executive branch by making Commissioners removable only for cause and authorizing them to remove ALJs only for cause.” RAB at 45 (citations omitted). Respondent did not raise this issue in its pleadings or while the matter was pending before the ALJ, but rather waited until the ALJ had ruled against it before first challenging the constitutionality of his functions in a single sentence on appeal. By waiting until this late date, Respondent has waived this claim. See In re Labmd, Inc., 2015 WL 5608167, at *2 (F.T.C. Sept. 14, 2015). By compressing its presentation of this broad issue into a single sentence, Respondent has failed to present a complete showing of constitutional harm. See Hospital Corp. of Am. v. FTC, 807 F.2d 1381, 1392-93 (7th Cir. 1986) (refusing to consider the merits of a for- 61 We do make one small additional change. The ALJ’s Order prohibits Respondent from entering into any agreement with a seller of contact lens products to “regulate” any search advertising. Lest this be interpreted to prohibit agreements to disclose the identity of the rival seller or to disclaim its affiliation with 1-800 Contacts, we have included an additional provision expressly permitting use of such less restrictive alternatives. 309 1-800 CONTACTS, INC. Opinion of the Commission cause termination claim when the hospital raising the constitutional challenge had “not laid a proper foundation for its assault” on the FTC’s structure). Such issues aside, the FTC’s ALJ occupies a different role than the Public Company Accounting Oversight Board (PCAOB) found to be improperly insulated from presidential control in Free Enterprise Fund v. PCAOB, 561 U.S. 477, 492-98 (2010), relied upon by Respondent. The FTC’s ALJ performs adjudicative rather than enforcement or policymaking functions, is subject to more Commission oversight, and is part of a well-established statutory structure that has been in place for more than 70 years. In addition, if the Administrative Procedure Act’s “good cause” standard for removal is properly construed—i.e., to allow removal of an ALJ for failure to perform adequately or to follow agency policies, and to limit the Merit Systems Protection Board’s role to determining whether a factual basis exists for the agency’s proffered grounds for removal—the APA gives the President a constitutionally adequate degree of control over ALJs. See Brief for Respondent Supporting Petitioners at 48-53, Lucia v. SEC, 138 S. Ct. 2044 (2018) (No. 17-130). Moreover, unlike in Lucia v. SEC, where the Court found that the ALJ was unconstitutionally appointed by SEC staff members, the FTC’s ALJ was appointed by the Commission, which is a “Head[] of Department[].” Lucia v. SEC, 138 S. Ct. 2044, 2050 (2018).

In this opinion, we have evaluated traditional concerns of antitrust law—the anticompetitive harms that flow when rivals agree to restrict truthful advertising and to limit their participation in auctions—in a contemporary context involving online shopping and advertising via internet search engines. Our analysis has accounted for and given weight to justifications based on trademark protection as well as the benefits of settling costly litigation. We hold that Complaint Counsel have shown competitive harm by demonstrating the inherently suspect nature of the restraints at issue. We have determined that Respondent has asserted cognizable and plausible procompetitive justifications, requiring Complaint Counsel to make a further showing. Complaint Counsel have made that showing, both by demonstrating the availability of less anticompetitive alternatives to the challenged restraints and by showing in greater detail that those restraints are indeed likely in the particular context to harm competition. In contrast, Respondent has failed to establish that its justifications are not merely plausible, but in fact valid. We also hold that Complaint Counsel have shown competitive harm by providing direct evidence that the challenged agreements resulted in actual anticompetitive effects. Respondent, however, failed to rebut Complaint Counsel’s direct evidence and could not provide sufficient efficiency justifications that would outweigh the evidence of anticompetitive effects. Consequently, we conclude that the advertising restrictions in the Challenged Agreements between Respondent and 14 of its rival online sellers of contact lenses constitute unfair methods of competition, in violation of Section 5 of the FTC Act, and we require Respondent to cease and desist from enforcing the unlawful provisions in its existing agreements and from entering into similar agreements in the future.

VOLUME 166 Final Order FINAL ORDER The Commission has heard this matter upon the appeal of Respondent from the Initial Decision, and upon briefs and oral argument in support thereof and in opposition thereto. For the reasons stated in the accompanying Opinion of the Commission, the Commission has determined to sustain the Initial Decision with certain modifications. IT IS ORDERED that the Initial Decision of the administrative law judge be, and it hereby is, adopted as the Findings of Fact and Conclusions of Law of the Commission, to the extent not inconsistent with the findings of fact and conclusions contained in the accompanying Opinion. Other findings of fact and conclusions of law of the Commission are contained in the accompanying Opinion.

IT IS FURTHER ORDERED that the following Order to cease and desist be, and it hereby is, entered:

I.

IT IS ORDERED that, as used in this Order, the following definitions shall apply: A. “1-800 Contacts” means 1-800 Contacts, Inc., its directors, officers, employees, agents, representatives, successors, and assigns; and any joint ventures, subsidiaries, partnerships, divisions, groups, and affiliates in each case controlled by 1-800 Contacts, and the respective partners, directors, officers, employees, agents, representatives, successors, and assigns of each.

B. “Commission” means the Federal Trade Commission.

C. “Communicate,” “Communicating,” or “Communication” means the exchange, transfer, or dissemination of any information, without regard to the manner or means by which it is accomplished.

D. “Entering Into” means entering into, adhering to, participating in, maintaining, implementing, enforcing, inviting, offering or soliciting. E. “Keyword” means a word or phrase used to instruct a Search Engine to display specified Search Advertising.

F. “Negative Keyword” means a word or phrase used to instruct a Search Engine not to display specified Search Advertising.

G. “Person” means both natural persons and artificial persons, including, but not limited to, corporations and unincorporated entities.

H. “Search Advertising” means online advertisements displayed on a Search Engine Results Page in response to a user query.

311 1-800 CONTACTS, INC. Final Order I. “Search Engine” means a computer program, available to the public, that enables Persons to search for and identify websites and sources of information on the World Wide Web.

J. “Search Engine Results Page” means a web page displayed by a Search Engine in response to a user query.

K. “Seller” means any Person that markets or sells any contact lens product and includes its employees, agents, and representatives.

L. “Trademark Infringement Claim” means a lawsuit threatened or filed in the United States of America purporting to enforce rights under a trademark. II.

IT IS FURTHER ORDERED that Respondent, directly or indirectly, or through any corporate or other device, in connection with the advertising, marketing, sale, or distribution of contact lenses in or affecting commerce, as “commerce” is defined in Section 4 of the Federal Trade Commission Act, 15 U.S.C. § 44, shall cease and desist from: A. Entering Into any combination, conspiracy, or agreement with a Seller to prohibit, restrict, regulate, or otherwise place a limitation on the ability of a Seller to participate in a Search Advertising auction, or to provide instructions to a Search Engine regarding the nature and extent of a Seller’s participation, including but not limited to, prohibiting or restricting the use of a Keyword or requiring the use of a Negative Keyword.

Provided that nothing in this Paragraph II.A shall prohibit Respondent from (a) initiating or prosecuting a lawsuit, (b) communicating to any Seller Respondent’s intention to initiate or prosecute a lawsuit, or (c) implementing or enforcing an order entered by any court of law, including an order approving a litigation settlement.

B. Entering Into any combination, conspiracy, or agreement with a Seller to prohibit, restrict, regulate, or otherwise place a limitation on any Search Advertising; provided, however, that nothing in this Paragraph II.B shall prohibit Respondent from entering into or complying with a written agreement providing that a: 1. Seller shall not include in the text of any Search Advertising (a) a false or deceptive claim, (b) a representation that Respondent is the source of the goods or services advertised therein, (c) a representation that the Seller is affiliated with or sponsored by Respondent, or (d) a name that is identical to or confusingly similar to any trademark owned by Respondent; or 2. Seller’s Search Advertising shall clearly identify the Seller (for the avoidance of doubt, including the name of the Seller in the URL, website VOLUME 166 Final Order address, or domain name shall constitute clear identification of the Seller); and Provided further that nothing in this Paragraph II.B shall prohibit Respondent from (a) initiating or prosecuting a lawsuit, (b) communicating to any Seller Respondent’s intention to initiate or prosecute a lawsuit, or (c) implementing or enforcing the order entered by any court of law, including an order approving a litigation settlement.

C. Entering Into any combination, conspiracy, or agreement with a Seller to prohibit, restrict, regulate, or otherwise place any limitation on truthful, non-deceptive, and non-infringing advertising or promotion.

D. Attempting to engage in any conduct that is prohibited by Paragraph II of this Order.

Provided, however, that nothing in this Paragraph II shall prohibit Respondent from entering into or complying with a written agreement with a Seller to require that Search Advertising disclose the Seller’s identity and/or lack of affiliation with Respondent or disclose that the Search Advertising is not sponsored by Respondent.

III.

IT IS FURTHER ORDERED that Respondent shall:

A. Cease and desist from enforcing or attempting to enforce any and all provisions, terms, or requirements in an existing agreement or court order that impose a condition on a Seller that is not consistent with Paragraph II of this Order. B. Within sixty (60) days after the date this Order is issued, take whatever action is necessary to vacate or nullify any and all provisions, terms, or requirements in any court order or agreement that impose a condition on a Seller that is not consistent with Paragraph II of this Order.

IV.

IT IS FURTHER ORDERED that Respondent shall:

A. Within thirty (30) days from the date this Order is issued: 1. Distribute by first-class mail, return receipt requested or by electronic mail with return confirmation, a copy of this Order and the Complaint to each of its officers, directors, and managers;

313 1-800 CONTACTS, INC. Final Order 2. Send by first-class mail, return receipt requested or by electronic mail with return confirmation, on Respondent’s official letterhead, the statement attached to this Order as Appendix A to each Person:

(a) To whom Respondent communicated regarding that Person’s involvement as a plaintiff or defendant in any actual or potential Trademark Infringement Claim; and (b) With whom Respondent entered into any agreement prohibited by Paragraph II of this Order.

B. For a period of five (5) years from the date this Order is issued: 1. Provide to Commission staff a copy of any Communication by Respondent with any Person regarding that Person’s suspected trademark infringement no later than ten (10) days after Communicating with such Person;

2. Send by first-class mail, return receipt requested or by electronic mail with return confirmation, on Respondent’s official letterhead, the statement attached to this Order as Appendix A to each Person referenced in Paragraph IV.B.1. of this Order no later than the time Respondent initially Communicates with such Person;

3. Provide to Commission staff a copy of any agreement (or description, if the agreement is not in writing) that Respondent enters into with a Seller relating to Search Advertising, no later than thirty (30) days after it enters into such agreement;

4. Provide to Commission staff notice and a copy of any proposed stipulated order to settle litigation with provisions that prohibit, restrict, regulate, or otherwise place a limitation on any Search Advertising or on the ability of a Seller to participate in a Search Advertising auction, no later than ten (10) days before requesting entry of that order;

5. Distribute by first-class mail, return receipt requested or by electronic mail with return confirmation, a copy of this Order and the Complaint to each Person who becomes an officer, director, or manager and who did not previously receive a copy of this Order and Complaint, no later than ten (10) days after the date such Person assumes his or her position; and, 6. Provide a copy of this Order to any court evaluating a request that a litigation settlement agreement relating to Search Advertising be approved by the court and/or incorporated into a court order.

VOLUME 166 Final Order C. Retain documents and records sufficient to record Respondent’s compliance with its obligations under this Paragraph IV.

V.

IT IS FURTHER ORDERED that Respondent shall file a verified written report with the Commission setting forth in detail the manner and form in which it intends to comply, is complying, and has complied with this Order:

A. No later than ninety (90) days from the date this Order is issued, and B. One (1) year from the date this Order is issued and annually thereafter for four (4) years on the anniversary of the date on which this Order is issued, and at such other times as the Commission may request.

VI.

IT IS FURTHER ORDERED that Respondent shall notify the Commission at least thirty (30) days prior to:

A. Any proposed dissolution of Respondent;

B. Any proposed acquisition, merger, or consolidation of Respondent; or C. Any other change in Respondent, including, but not limited to, assignment and the creation or dissolution of subsidiaries, if such change might affect compliance obligations arising out of this Order.

VII.

IT IS FURTHER ORDERED that, for the purpose of determining or securing compliance with this Order, and subject to any legally recognized privilege, and upon written request and upon five (5) days’ notice to Respondent, that Respondent shall, without restraint or interference, permit any duly authorized representative of the Commission: A. Access, during office hours of Respondent and in the presence of counsel, to all facilities and access to inspect and copy all books, ledgers, accounts, correspondence, memoranda, and all other records and documents in the possession, or under the control, of Respondent relating to compliance with this Order, which copying services shall be provided by Respondent at its expense; and B. To interview officers, directors, or employees of Respondent, who may have counsel present, regarding such matters.

315 1-800 CONTACTS, INC. Final Order VIII.

IT IS FURTHER ORDERED that this Order shall terminate on November 7, 2038. By the Commission, Commissioner Phillips dissenting and Commissioner Wilson not participating.

Appendix A [Letterhead of 1-800 Contacts] [Name and Address of the Recipient] Dear (Recipient):

As you may know, the Federal Trade Commission issued an administrative complaint in 2016 against 1-800 Contacts, Inc. (“1-800 Contacts”) challenging several agreements between 1- 800 Contacts and other contact lens sellers that restrict the ability of such sellers to purchase trademark keywords in search advertising auctions, or to place search advertising triggered by those keywords on internet search engine results pages.

The Federal Trade Commission has issued a Decision and Order (“Order”) against 1-800 Contacts in connection with its complaint. This Order provides, in part, that 1-800 Contacts may not prohibit competing sellers of contact lenses from engaging in truthful, non-deceptive advertising or solicitation through the display of search advertising. Specifically, 1-800 Contacts may not:

1. Enter into, enforce, or attempt to enforce any agreement between or among 1-800 Contacts and a contact lens seller to restrict the ability of the seller to participate in any internet search advertising auction, including restricting the use of keywords or requiring the use of negative keywords; or 2. Enter into, enforce, or attempt to enforce any agreement with a contact lens seller that otherwise places any limitation on any search advertising. The Order further requires 1-800 Contacts to take whatever action is necessary to have vacated all court orders or other restraints related to trademark infringement claims initiated to accomplish any of the above-listed prohibited activities.

VOLUME 166 Dissenting Statement The Order does not prohibit 1-800 Contacts from entering into an agreement with a seller of contact lenses that requires certain disclosures in the text of an advertisement, including a clear identification of the seller placing the advertisement. For more specific information, you should refer to the FTC order itself. The Federal Trade Commission’s Complaint and Decision and Order are available on the Commission’s website, http:\\www.ftc.gov.

DISSENTING STATEMENT OF COMMISSIONER NOAH JOSHUA PHILLIPS The majority’s decision in this case deems “inherently suspect” and then condemns agreements to settle legitimate trademark infringement litigation. Applicable precedent requires the more thorough rule of reason analysis, with more credence given to the intellectual property at the heart of the case. The majority make a separate holding that the settlements are anticompetitive based on a showing of direct effects, but the evidence upon which they rely fails, both as a matter of law and as a matter of fact, to meet the legal requirement that such effects must be actual, sustained, and significant or substantial. I fear the majority’s approach will foster uncertainty and undermine trademark policy, and so I respectfully dissent. Neither the necessary judicial experience nor economic learning exist to apply a truncated antitrust analysis to the facts of this case. A fair reading of relevant case law makes clear that the full rule of reason should apply to the trademark settlement agreements between 1-800 Contacts and thirteen alleged trademark infringers (the “Trademark Settlements”).1 In supporting their choice of analytical framework, the majority avoid entirely the fact that the agreements at issue settle intellectual property claims. They then judge and discard entirely the value of those claims, a methodological error with a result that judicial experience and economic learning have taught us for decades to avoid—i.e., an unclear rule that is difficult to administer and harder still to predict, and that may capture and will chill procompetitive behavior. The majority couch their holding as a limited one dealing with restraints on the opportunity to make price comparisons—an overstated conclusion—but their decision not to grapple fairly with the trademark context of the agreements results in a rule that appears to be one of the following:

1 I use the phrase “Trademark Settlements” to refer to the agreements settling trademark infringement litigation between 1-800 Contacts and the following thirteen contact lens retailers: (1) AC Lens, (2) Coastal Contacts, (3) Contact Lens King, (4) Empire Vision, (5) EZ Contacts, (6) Lenses for Less, (7) Lensfast, (8) Memorial Eye, (9) Standard Optical, (10) Tram Data, (11) Vision Direct, (12) Walgreens, and (13) Web Eye Care. The phrase “Trademark Settlements” does not include the sourcing and services agreement between 1-800 Contacts and Luxottica (the “Luxottica Agreement”) because that agreement did not resolve trademark infringement litigation and, therefore, should be analyzed separately. See Section 0, infra. 317 1-800 CONTACTS, INC. Dissenting Statement  all advertising restrictions are inherently suspect, regardless whether they protect intellectual property rights, a rule supported by the logic of the opinion but which the majority disclaim expressly; or  a standard of review under which the Commission will review as inherently suspect settlements of what it considers weak trademark infringement claims, leaving open the question of how it will analyze infringement claims that the Commission adjudges to be strong.

The former rule will treat clearly pro-competitive conduct as presumptively unlawful. The latter will require the Commission and federal courts to litigate (or re-litigate) inherently fact-specific intellectual property infringement claims in every antitrust challenge to a settlement agreement, a difficult process we have long eschewed. It will also create uncertainty for parties considering settlement, deterring enforcement and, in the case of trademarks, reducing the incentive to build brands.

Precedent offers—indeed, requires—a better approach: apply the full rule of reason to antitrust challenges to trademark settlement agreements like those at issue here, giving appropriate credence to the fact that the conduct at issue is the settlement of legitimate (i.e., nonsham) trademark infringement claims. Such a rule would provide guidance to the market, increase certainty, encourage brand investment, and enhance competition. I. Background Jonathan Coon started the business that would become 1-800 Contacts in 1992 from his college dormitory room with just $50 to his name, seeking to reduce prices, improve service, and provide a better customer experience for contact lens consumers. IDF 30-33, 43;2 Coon, Tr. 2649:9-12, 2651:12-20. Over the next 26 years he would succeed, building a company (and a brand) from essentially nothing to one of the largest contact lens retailers in the country, while introducing American consumers to mail-order contact lenses (and later ordering contacts online), driving down prices, and attracting competition from small and large companies alike. That growth required a combination of a massive investment in advertising and a constant quest to improve the customer experience. That is the type of conduct that antitrust and trademark law should, and do, encourage.

2 For the sake of convenience and consistency, I use the same abbreviations as the majority for the following documents:

Compl.: Complaint ID: Initial Decision IDF: Initial Decision Finding of Fact Stip.: Joint Stipulation Regarding Search Engine Mechanics and Glossary of Terms RAB: Respondent’s Brief on Appeal I also use the following abbreviations in citations:

Op.: Opinion of the Commission IH: Investigational Hearing VOLUME 166 Dissenting Statement A. 1-800 Contacts Invested a Tremendous Amount to Build Its Brand. Trademarks encourage innovation and brand investment, giving more information to customers and attracting competition. See Section 0(A)(4)(b), infra. 1-800 Contacts has a long history of taking risks to invest in its brand. In July 1995, when Mr. Coon and his business partner John Nichols renamed their company 1-800 Contacts and obtained the associated telephone number, the company’s sales more than doubled in the first month. IDF 36-37; Coon, Tr. 2654:13-19, 2658:19-25, 2661:20-2662:16. It cost Mr. Coon and Mr. Nichols approximately $163,500 to obtain the telephone number “1-800-CONTACTS”, but they only had $10,000 in the bank at the time, so they used that entire sum to make an upfront payment and agreed to pay the remainder in monthly installments of approximately 10% of the company’s total monthly revenue. Coon, Tr. 2658:19-2660:25.

After it started marketing itself as 1-800 Contacts, the company saw an increase of 20% to 25% in customer acquisition and retention. IDF 51. The initial advertising campaign was in print, but shortly thereafter the company started advertising on television. IDF 50, 52. Television advertising had an immediate and significant impact, growing the business by approximately 50% in just a few months. IDF 53. Ever since, television has generally been the largest category of marketing spend in 1-800 Contacts’ advertising budget. See RX0739 (Murphy Expert Report) at 092.

1-800 Contacts’ approach to promoting itself was—and continues to be—designed to generate brand awareness and new orders through “a multichannel integrated marketing” strategy. IDF 60-61. This strategy has included “print advertising, television advertising, radio advertising, internet display advertising, affiliate marketing, social media advertising, and search engine optimization, in addition to internet search advertising.” IDF 62. Of particular relevance to this case, there is a positive correlation between 1-800 Contacts’ television advertisements and traffic to 1-800 Contacts’ website via searches for its trademarked terms. IDF 63; CX9017 at 045 (Blackwood Dep. 176:2-12); CX9032 at 063 (L. Schmidt Dep. 246:25-247:13); RX0736 (Goodstein Expert Report) at 008; see also CX9031 at 025-026 (C. Schmidt Dep. 95:25-97:15) (testifying that 1-800 Contacts saw an increase in the amount of paid search advertising on its trademarked terms in response to broad scale advertising, such as television and radio). Research conducted by 1-800 Contacts found that 40 percent of the traffic to its website from paid trademark search was directly related to television advertising. CX9017 at 059 (Blackwood Dep. 230:1-23).

1-800 Contacts has spent hundreds of millions of dollars to generate brand awareness and new orders. From 2002 to 2014 (just 13 of the 26 years the company has existed), 1-800 Contacts spent more than {$363 million} on advertising, of which {$186 million} (or more than {51}%) went to television advertising and almost {$87 million} (or {23.9}%) to all internet advertising (not just paid search advertising). IDF 64-65; RX0739 (Murphy Expert Report) at 092. In 2014 alone (the most recent year for which data is available), 1-800 Contacts’ marketing budget was {$33.8 million}; {$15 million} (or {44.4}%) of that total budget went to television advertising and {$12.4 million} (or {36.9}%) to all internet advertising (not just paid search advertising). IDF 64-66; RX0739 (Murphy Expert Report) at 092. As these numbers show, 319 1-800 CONTACTS, INC. Dissenting Statement television is 1-800 Contacts’ principal means of advertising because it drives growth in terms of brand awareness and identifying and finding customers. See CX9001 at 016-017 (Bethers IH 60:15-61:3). Other online contact lens retailers generally have not invested in broad scale (e.g., television) advertising. CX9029 at 004 (Bethers Dep. 10:3-11:13); CX9035 at 023 (Coon Dep. 88:2-6); RX0736 (Goodstein Expert Report) at 009; see also Bethers, Tr. 3614:15-24 (“[W]e don’t see anyone that invests in broad-scale advertising like us [1-800 Contacts]”). Paid internet search advertising through Google comprised between {6.5}% and {12.3}% of 1-800 Contacts’ total advertising budget between 2004 and 2014.3 See IDF 66; RX0739 (Murphy Expert Report) at 027-028, 092, 140. Paid search advertising through Google using 1-800 Contact’s trademarked terms constituted no more than {1.7}% of 1-800 Contacts’ total advertising budget and no more than {14}% of 1-800 Contacts’ paid search advertising expenses in any year between 2004 and 2014.4 RX0739 (Murphy Expert Report) at 027-028, 092, 140. In 2014 (the most recent year for which data is available), paid search advertising through Google constituted {10.5}% of 1-800 Contacts’ total advertising budget of {$33.8 million} and paid search advertising through Google on its trademarked terms accounted for {0.9}% of that total budget (and only {9}% of 1-800 Contacts’ expenses on paid search advertising). RX0739 (Murphy Expert Report) at 027-028, 092, 140. This massive endeavor—the kind of conduct trademark law is intended to foster—did more than benefit 1-800 Contacts: it pioneered the mail-order contact lens business and then the online contact lens business to the direct benefit of consumers in the form of reduced prices and increased convenience and choice.

B. 1-800 Contacts Has Been Committed to Improving the Customer Experience Since Its Founding.

As Administrative Law Judge D. Michael Chappell (the “ALJ” or “Judge Chappell”) found in the Initial Decision, “1-800 Contacts’ business objective from the company’s inception was to make the process of buying contact lenses simple and it tries to distinguish itself from other contact lens retailers by making it faster, easier, and more convenient to get contact lenses.” IDF 43; see also Coon, Tr. 2712:11-2713:7 (testifying that the company’s strategy of distinguishing itself on service stemmed from a recognition that it would be easy for another retailer to match prices but it is very difficult to “create a brand and provide great service”). This contrasts with other online contact lens retailers, which generally do not seek to distinguish themselves on the basis of customer experience, customer service, or simplicity. See CX9029 at 004 (Bethers Dep. 9:12-11:13). 1-800 Contacts did not limit itself to competing on price because it found that many customers valued speed and convenience just as much as price. Coon, Tr. 2705:16 -2708:1; see also RX1117 at 028 ({customer research conducted by 1-800 Contacts 3 1-800 Contacts either did not spend money on paid search advertising in 2002 and 2003, or that data is unavailable. See RX0739 (Murphy Expert Report) at 092.

4 As with paid search advertising, 1-800 Contacts either did not spend money on paid search advertising using its own trademarked terms in 2002 and 2003, or that data is unavailable. See RX0739 (Murphy Expert Report) at 092. VOLUME 166 Dissenting Statement between 2010 and 2014 showing that in selecting a contact lens retailer, important factors for many consumers include the convenience and speed of delivery}). 1-800 Contacts has made significant investments in providing high quality service to customers, including a dedicated call center, prompt shipping within two business days, quality control measures in inventory, prescription verification, and a 100% guaranteed return policy. CX9031 at 024 (C. Schmidt Dep. 90:2-92:3); Coon, Tr. 2690:20-2692:15. 1-800 Contacts stocks more contact lenses in inventory than any other contact lens retailer, which allows it to fill 98% of all orders from inventory on hand; answers most calls with a live person by the third ring and most emails within 10 minutes; has live customer support personnel available to answer text messages; offers click-to-chat customer service; and replaces torn lenses for free. IDF 44-46; Coon, Tr. 2690:20-2692:15; RX0904 at 016.

In addition, 1-800 Contacts designed its website with the same goals as Mr. Coons founded the company: to make the contact lens buying experience better for customers. See IDF 39. The website was as simple and efficient as possible, minimizing “the amount of time spent on the website and the number of clicks a consumer had to make to purchase contact lenses.” Id. Over time, the company continued to improve its website and developed a mobile application to ensure that customers could purchase contact lenses as quickly and easily as possible. See, e.g., IDF 40-42.

1-800 Contacts’ relentless investment in its brand and in improving its customer service are recognized. Many third parties—including J.D. Power and Associates, StellaService Elite, and Foresee—have recognized or given awards to 1-800 Contacts for its customer service. IDF 47; see also RX0736 (Goodstein Expert Report) at 016, Table 2 (listing other awards received by 1-800 Contacts, including awards for its customer service). But that has not stopped 1-800 Contacts from continuing to invest in improving its service to enhance the customer experience. See, e.g., IDF 48.

The service and brand investments made by 1-800 Contacts have resulted in millions of consumers purchasing contact lenses from 1-800 Contacts over the phone and online. They are precisely the types of investments that trademark law exists to protect and encourage. And, according to multiple witnesses, they created precisely the value that other retailers sought to derive by bidding on 1-800 Contacts’ trademarked terms. See, e.g., CX9033 at 017 (Mohan Dep. 61:9-12) (Walmart executive testifying that 1-800 Contacts’ trademarks were more valuable as search terms “[b]ecause a lot more people know the brand.”); CX9039 at 040 (Clarkson Dep. 155:25-156:8) (AC Lens executive testifying that the value it receives from paid trademark search advertising depends on the strength of the competitors’ brand); id. at 026 (97:20-98:3) (noting 1-800 Contacts “unmatched brand awareness”).

321 1-800 CONTACTS, INC. Dissenting Statement C. The Trademark Settlements Resolved Legitimate and Contested Trademark Infringement Claims.

1. The Context Surrounding the Trademark Settlements.

The Trademark Settlements resolved trademark infringement claims brought by 1-800 Contacts against certain other online contact lens sellers, which bought advertisements using 1- 800 Contacts’ trademarks as keywords—i.e., when consumers searched for “1-800 Contacts”, the search engine would display advertisements for the other sellers. As early as 2002, online retailers of contact lenses expressed concern that bidding for advertisements using third parties’ trademarks might be illegal. See, e.g., IDF 583 (“In 2002, AC Lens decided not to use 1-800 Contacts’ trademarks as keywords for paid search advertising because of legal concerns.”); Clarkson, Tr. 325:6-23 (AC Lens executive testifying that “it was unclear to me what the legal situation was relative to advertising on other companies’ trademarks” and that he had a concern about advertising on other companies’ trademarks “for a long time”); CX9003 at 024 (Clarkson Dep. 90:21-91:10) (“I think I had a general sort of concern that [paid trademark search advertising] may not be legal anyway.”). 1-800 Contacts itself had a policy that pre-dated the Trademark Settlements not to use other companies’ trademarked terms as keywords to trigger paid search advertisements, in part attributable to a concern about the propriety of using other companies’ trademarks as keywords. CX9031 at 016 (C. Schmidt Dep. 57:7-59:1); CX9001 at 027-028 (Bethers IH 104:4-105:20).

Prior to April 2004, Google—the largest search engine since before the first Trademark Settlements—did not permit advertisers to bid on keywords that contained a trademark owned by a third party. See IDF 137, 287. Microsoft, which owns Bing—the second-largest online search engine after Google—had the same policy until 2011. See IDF 298. 1-800 Contacts executives met with Google representatives in April 2004, the same month that Google changed its policy and began allowing advertisers to bid on the trademarks of other companies. See Schmidt, Tr. 2900:12-2901:1. At this meeting, 1-800 Contacts understood Google’s position to be that while Google would no longer resolve trademark disputes directly, it offered negative keywords as an effective tool to prevent or inhibit future trademark infringement. Schmidt, Tr. 2904:2-16, 2905:16-25; CX9031 at 010 (C. Schmidt Dep. 33:20- 34:21). Negative keywords prevent an advertisement from being triggered by the words or phrases comprising the negative keywords. Stip. at 2. According to 1-800 Contacts, Google representatives specifically suggested that 1-800 Contacts resolve its disputes directly with its competitors by telling them to implement 1-800 Contacts’ trademarks as negative keywords. CX9031 at 010-011 (C. Schmidt Dep. 33:20-34:20, 36:13-37:3); CX9013 at 044-045 (Aston Dep. 170:8-20, 171:10-172:3, 173:5-20).

Following Google’s policy change in April 2004, 1-800 Contacts continued to protect its trademarks vigorously because, among other things, failure to police a trademark could render a trademark unenforceable. Hogan, Tr. 3265:4-3266:9; see also RX0734 (Hogan Expert Report) at 013 (citing Malaco Leaf, AB v. Promotion In Motion, Inc., 287 F. Supp. 2d 355, 364-65 (S.D.N.Y. 2003) (“[T]trade dress may become generic, meaning commonly used and not entitled to protection, as a result of the trademark owner’s failure to police it”) (citation, brackets, and VOLUME 166 Dissenting Statement quotation marks omitted); Bachellerie v. Z. Cavaricci, Inc., 762 F. Supp. 1070, 1077 (S.D.N.Y. 1991) (failure of plaintiff to enforce its mark against third-party users “diminishes the strength of the mark”)). Other trademark owners acted in a similar manner. See RX0734 (Hogan Expert Report) at 083-086; RX0926 at 001 (listing cases involving the “purchase of another party’s trademark as a keyword for internet advertising”). Some of these attempts by trademark owners to protect their marks ultimately led to litigation.

In the initial years of paid search advertising litigation, between 2004 and 2009, it was unclear whether courts would recognize a cause of action under a theory that bidding on trademarked terms as keywords constituted a “use in commerce” under the Lanham Act, a critical predicate to establishing a trademark infringement claim. IDF 333; RX0734 (Hogan Expert Report) at 059-060. Only three of the Trademark Settlements were signed during this period: Vision Direct (executed in June 2004), Coastal Contacts (executed in October 2004), and EZ Contacts (executed in May 2008). See IDF 306, 314, 344. On April 3, 2009, however, the legality of 1-800 Contacts’ competitors’ bidding on advertisements with 1-800 Contacts’ trademarks as keywords—precisely the conduct ended by the Trademark Settlements—became even more dubious when the Second Circuit issued an opinion holding that using trademarks as keywords in paid search advertising “fits literally within the terms specified by [the Lanham Act,] 15 U.S.C. § 1127” as a “use in commerce”. Rescuecom Corp. v. Google Inc., 562 F.3d 123, 129-30 (2d Cir. 2009); see also id. at 127 (“The allegations of Rescuecom’s complaint adequately plead a use in commerce.”). Following Rescuecom, federal circuit courts came to agree that bidding on trademarked terms as keywords for paid search advertising constituted a “use in commerce” for the purposes of trademark law, see IDF 333, eliminating a threshold defense in trademark infringement litigation. For advertisers bidding on other companies’ trademarks, this shifted the focus to whether, in particular cases, the use was likely to cause confusion among customers, see IDF 333; Hogan, Tr. 3256:11-19, a highly fact-specific inquiry necessitating litigation. See Section 0(A)(4)(a), infra. The legal risks rose, increasing the incentive for alleged trademark infringers to settle rather than endure a full trial on the merits, given the fact-specific nature of the inquiry into trademark confusion. See, e.g., Hogan, Tr. 3260:21-3261:4. In the wake of the Rescuecom decision and the resulting change in legal exposure, 1-800 Contacts entered nine of the thirteen Trademark Settlements between December 2009 and February 2011. IDF 348; CX0315 (Lensfast, Dec. 2009); RX0028 (AC Lens, Mar. 2010); CX0323 (Contact Lens King, Mar. 2010); CX0320 (Lenses for Less, Mar. 2010); CX0319 (Empire Vision, May 2010); CX0321 (Tram Data, May 2010); CX0322 (Walgreens, June 2010); CX0324 (Web Eye Care, Sept. 2010); RX0408 (Standard Optical, Feb. 2011). One month after Rescuecom, 1-800 Contacts entered a second settlement agreement with Vision Direct to address Vision Direct’s alleged violations of the 2004 settlement agreement for failing to implement negative keywords. See IDF 345-347; CX0314 at 004 (“The 2004 Settlement Agreement shall remain in full force and effect except that the Parties’ sole obligations with respect to the use of negative keywords shall be to comply with the terms of this Settlement Agreement.”); see also CX0316 (Order of Permanent Injunction, 1-800 Contacts, Inc. 323 1-800 CONTACTS, INC. Dissenting Statement v. Vision Direct, Inc., No. 08-cv-1949 (S.D.N.Y May 15, 2009)). Only one Trademark Settlement came after the initial wave of settlements following Rescuecom: Memorial Eye settled in November 2013, principally because of the legal uncertainty about its failure to implement negative keywords on 1-800 Contacts’ trademarked terms. See, e.g., IDF 349, 351; Holbrook, Tr. 1942:12-13 (“We knew that the [negative keyword] broad matching issue had not firmly been put to rest by the court.”); CX9024 at 017 (Holbrook Dep. 63:13-18) (“We also knew that in the appellate court, I believe it was, that the appellate court had been silent on the [negative keyword] broad matching issue, which was to us the most important thing. It was a big deal. So there was a lot of legal uncertainty because of that still hanging out there.”); see also IDF 617 (finding that Memorial Eye did not bid on 1-800 Contacts’ trademarked terms as keywords in paid search advertising).

The Trademark Settlements resolved increasing legal risk for putative bidders on trademarked keywords. No one, even today, contends that the trademark claims asserted by 1- 800 Contracts were shams or legal claims asserted to achieve an otherwise anticompetitive end. See RX0680 at 013 (“Complaint Counsel therefore does not contend that the lawsuits constituted ‘sham’ litigation as defined by the Supreme Court in PRE.”) (referring to Prof’l Real Estate Investors, Inc. v. Columbia Pictures Indus., Inc., 508 U.S. 49 (1993)); RX0678 at 008 (“Complaint Counsel does not contend that the lawsuit, 1-800 Contacts, Inc. v. Memorial Eye, P.A., was Sham Litigation.”); id. (“Complaint Counsel does not contend that the lawsuit, 1-800 Contacts, Inc. v. Lens.com, was Sham Litigation.”); see also 1-800 Contacts, Inc. v. Memorial Eye, P.A., No. 08-cv-983, 2010 WL 988524, at *6 (D. Utah Mar. 15, 2010) (“[T]he Court finds that Plaintiff’s claim is not baseless”); Lens.com, Inc. v. 1-800 Contacts, Inc., No. 12-cv-352, 2014 WL 12596493, at *1 (D. Utah Mar. 3, 2014) (“Because the district court and the Tenth Circuit agree that the underlying action was not baseless, this court agrees that Lens’ claims, all of which center on the proposition that 1-800 engaged in sham litigation, should be dismissed with prejudice.”).

2. The Relevant Terms Contained in the Trademark Settlements. The Trademark Settlements resolved legitimate intellectual property infringement claims. They were bilateral: 1-800 Contacts entered each Trademark Settlement separately and with a single counterparty to protect each settling party’s trademarks. No material amount of money changed hands.5 Users who searched for 1-800 Contacts’ trademarks would not see an advertisement for the other settling party (although they might see an “organic” search result, 5 Certain Trademark Settlements contained token amounts of monetary consideration, but nothing approaching the millions of dollars at issue in FTC v. Actavis, Inc., 570 U.S. 136 (2013), and always in the usual direction (i.e., from the defendant to the plaintiff). See CX0311 at 002 (Vision Direct paid $1 in monetary consideration in 2004); CX0313 at 002 (EZ Contacts paid $29,000 in monetary consideration); CX0314 at 001 (Vision Direct paid $475,000 in 2009 for “partial reimbursement of 1-800 Contacts’ attorneys’ fees”); CX0315 at 001 (Lensfast made a $20,000 payment); CX0323 at 001 ($8,000 payment by Contact Lens King); CX0324 at 001 ($2,000 payment by Web Eye Care); cf. Actavis, 570 U.S. at 145 (“[The branded manufacturer] agreed to pay millions of dollars to each generic”). None of the payments split monopoly rents, cf. Actavis, 570 U.S. at 154; indeed, the majority of Trademark Settlements had no monetary component. See CX0310 (Coastal Contacts); RX0028 (AC Lens); CX0320 at 002 (Lenses for Less); CX0319 (Empire Vision); CX0321 (Tram Data); CX0322 (Walgreens); RX0408 (Standard Optical); CX0326 (Memorial Eye).

VOLUME 166 Dissenting Statement depending on relevance, see Stip. at 5). The parties were adjusting to an evolving market and increased legal risk, achieving by contract (with implementing guidance from Google)6 what had previously been the stated policy of the two most popular search engines.7 See CX9031 at 010- 011 (C. Schmidt Dep. 33:20-35:2, 35:23-36:2, 36:13-37:3); CX9013 at 044 (Aston Dep. 172:1-3) (“They [Google] instructed us [1-800 Contacts] to have the offenders add those specific trademarked terms into their negatives for their -- for their AdWords campaigns.”); id. at 044- 045 (Aston Dep. 170:8-20, 171:10-19, 173:5-20). First, the Trademark Settlements prohibited both 1-800 Contacts and the counterparty from bidding on each other’s trademarked terms as keywords. IDF 363. Second, twelve of the thirteen Trademark Settlements required both parties to implement negative keywords to prevent their advertisements from appearing in response to searches for the other party’s trademarked terms. IDF 364; ID at 1; Compl. ¶ 24. It is important to keep in mind what the Trademark Settlements did not require. The Trademark Settlements did not prevent 1-800 Contacts or other online contact lens retailers from engaging in any form of non-infringing advertising. There were no restrictions on the settling parties’ ability to advertise offline (e.g., through print, television, or radio); to advertise using other forms of electronic/online advertising (e.g., internet display advertising, affiliate marketing, social media advertising, and search engine optimization); or to engage in paid search advertising as long as the advertisement did not appear in response to a search for one of the settling parties’ trademarks. Nothing prevented the parties from buying advertisements to respond to consumers’ searches for generic terms or phrases, such as “contacts”, “contact lenses”, “cheap contacts”, “inexpensive contacts”, or “discount contacts”. See, e.g., IDF 367. And the parties to the Trademark Settlements did, in fact, engage in many of these other types of advertising. See IDF 497-561 (describing the importance of paid search advertising generally—i.e., not just for trademarked keywords—to contact lens retailers, and noting that most retailers advertise in forms other than paid search advertising); see also Op. at 6-7 (noting the importance of paid search advertising generally—i.e., not just for trademarked keywords—to contact lens retailers). Neither the majority’s opinion nor the Initial Decision identifies what portion of the marketing budgets of the counterparties to the Trademark Settlements comprises trademark search advertising (as opposed to paid search advertising generally). Most of the Trademark Settlements specifically permit non-infringing uses like comparative advertising and parodies. For example, the 2004 settlement agreement between Vision Direct and 1-800 Contacts stated that the acts prohibited by the agreement “shall not include (i) use of the other Party’s Trademarks on the Internet in a manner that would not constitute an infringing use in an non-Internet context, e.g., the use on the Internet of comparative advertising, parodies, and similar non-Infringing uses . . . .” CX0311 at 004; see also IDF 369 (citing CX0311 at 004 (Vision Direct 2004); CX0313 at 004 (EZ Contacts); CX0315 at 004 (Lensfast); CX0319 at 002 (Empire Vision); CX0320 at 004 (Lenses for Less); CX0321 at 002 (Tram Data); CX0323 at 003 (Contact Lens King); CX0324 at 003 (Web Eye 6 The April 2004 meeting between 1-800 Contacts and Google predated all of the Trademark Settlements, the first of which was executed in June 2004. See CX0311 (Vision Direct Trademark Settlement, dated June 24, 2004). 7 All but one of the Trademark Settlements incorporated Google’s advice to use negative keywords to ensure that the settling parties’ trademarks were protected. See Compl. ¶ 24; CX0310 (1-800 Contacts’ Trademark Settlement with Coastal Contacts did not include a provision requiring the implementation of negative keywords). 325 1-800 CONTACTS, INC. Dissenting Statement Care); RX0028 at 002 (AC Lens); RX0408 at 003 (Standard Optical)); see also Op. at 9 (citing IDF 369 for the proposition that ten of the thirteen Trademark Settlements contained a clause permitting non-infringing uses).

The Trademark Settlements likewise place no restrictions on the content that any of the settling parties may include in their advertisements. The settling parties are free to advertise lower prices and higher quality whenever and, in general, wherever they like. And, of course, the restrictions in the Trademark Settlements impact only those consumers who search specifically for 1-800 Contacts’ trademarks, the vast majority of which searches are navigational, i.e., searches performed by the consumer with the intent to locate 1-800 Contacts’ website. RX0733 (Ghose Expert Report) at 032, 050.

The Trademark Settlements sought to balance 1-800 Contacts’ legitimate interests in protecting its trademarks with competitors’ (and consumers’) interests in truthful advertising. II. The Majority Fail to Show That the Trademark Settlements Are Anticompetitive. The majority deem the Trademark Settlements anticompetitive by applying the “inherently suspect” framework, which truncates the traditional rule of reason analysis, and, alternatively, by finding direct anticompetitive effects. Governing precedent supports neither approach on the facts adduced, and in neither analysis do the majority grapple adequately with the intellectual property rights at the heart of this case. A. The Trademark Settlements Are Not Inherently Suspect. 1. Categorizing Conduct as Inherently Suspect Is a Drastic Step. The Supreme Court has made clear time and again that “abandonment of the ‘rule of reason’ in favor of presumptive rules (or a ‘quick-look’ [i.e., inherently suspect] approach) is appropriate only where ‘an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect on customers and markets.’”8 FTC v. Actavis, Inc., 570 U.S. 136, 159 (2013) (quoting Cal. Dental Assn v. FTC, 526 U.S. 756, 770 (1999)). The per se and inherently suspect “standards are exceptional . . . and their application is reserved for the most patently anticompetitive restraints.” Craftsmen Limousine, Inc. v. Ford Motor Co., 491 F.3d 380, 387 (8th Cir. 2007), cert. denied, 552 U.S. 1040 (2007). “[T]he Supreme Court has cautioned that presumptions of anticompetitiveness should not be lightly invoked.” Princo Corp. v. Intl Trade Commu, 616 F.3d 1318, 1339 (Fed. Cir. 2010) (en banc) (citing Broadcast Music, Inc. v. Columbia Broad. Sys., Inc. (“BMI”), 441 U.S. 1, 8-9 (1979)), cert. denied, 563 U.S. 987 (2011); see also id. (“Quick-look analysis applies to ‘naked restraint[s] on price and output’”) (brackets in original) (quoting Cal. Dental, 526 U.S. at 769-70). In our rulings, the Commission has recognized as much. See, e.g., In re N. Tex. 8 “Quick look” is the federal judiciary’s equivalent to the Commission’s “inherently suspect” framework. See, e.g., N. Tex. Specialty Physicians v. FTC (“NTSP”), 528 F.3d 346, 360-61 (5th Cir. 2008) (“The ‘inherently suspect’ paradigm that the FTC employed in the present case is a ‘quick-look’ rule-of-reason analysis.”), cert. denied, 555 U.S. 1170 (2009).

VOLUME 166 Dissenting Statement Specialty Physicians, 140 F.T.C. 715, 719, 733 (2005), aff’d in relevant part sub nom., N. Tex. Specialty Physicians v. FTC, 528 F.3d 346 (5th Cir. 2008), cert. denied, 555 U.S. 1170 (2009).9 The Trademark Settlements do not approximate conduct that the Commission or courts have previously found to be inherently suspect, much less per se illegal. Those precedents make abundantly clear that the Commission should not treat the Trademark Settlements as presumptively unlawful. That is especially so given the trademark rights involved, an issue that none of the cases on which the majority rely even consider.10 2. We Lack an Adequate Basis to Declare the Trademark Settlements Inherently Suspect.

In California Dental, the progenitor for the Commission’s “inherently suspect” framework,11 the Supreme Court outlined the test for when it is appropriate to truncate the rule of reason analysis: only “when the great likelihood of anticompetitive effects can easily be ascertained.” Cal. Dental, 526 U.S. at 770 (citations omitted). “[W]here . . . any anticompetitive effects of given restraints are far from intuitively obvious,” however, “the rule of reason demands a more thorough enquiry . . . .” Id. at 759. “The object is to see whether the experience of the market has been so clear, or necessarily will be, that a confident conclusion about the principal tendency of a restriction will follow from a quick (or at least quicker) look, in place of a more sedulous one.” Id. at 781.

Lower courts and the Commission have elaborated upon the market experience necessary to apply the “inherently suspect” framework. Interpreting California Dental, the D.C. Circuit held in Polygram that “[i]f, based upon economic learning and the experience of the market, it is obvious that a restraint of trade likely impairs competition, then the restraint is presumed unlawful . . . .” Polygram Holding, Inc. v. FTC (“Polygram II”), 416 F.3d 29, 36 (D.C. Cir. 2005), aff’g sub nom., In re Polygram Holding, Inc. (“Polygram I”), 136 F.T.C. 310 (2003). We likewise stated that inherently suspect conduct “ordinarily encompasses behavior that past judicial experience and current economic learning have shown to warrant summary condemnation.” Polygram I, 136 F.T.C. at 344-45. That judicial experience and economic learning are absent here.

9 The majority apparently do not view application of the “inherently suspect” framework as exceptional. Their opinion suggests that as long as they consider the specific procompetitive justifications of the challenged conduct, it does not matter whether the “inherently suspect” label is applied. See Op. at 41. The majority’s view not only discounts any value of trademarks generally and relies on assessments in each case of the value of the trademarks at issue, see id. at 38-41, it also gives short shrift to the precedent instructing that application of the “inherently suspect” label is exceptional. See, e.g., Cal. Dental, 526 U.S. at 769-81. 10 I disagree with the majority’s attempts to distinguish the two relevant cases that involve intellectual property, Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50 (2d Cir. 1997), and Actavis. See Section 0(A)(3), infra (discussing Clorox); Sections 0(A)(2)(a)(i), 0(A)(4)(a), 0(C), infra (discussing Actavis). 11 See NTSP, 528 F.3d at 361 (“The FTC formulated its ‘inherently suspect’ analysis after the issuance of California Dental Association”) (citing Polygram Holding, Inc. v. FTC (“Polygram II”), 416 F.3d 29, 35-36 (D.C. Cir. 2005)).

327 1-800 CONTACTS, INC. Dissenting Statement a) We Lack Sufficient Judicial Experience to Presume the Trademark Settlements Are Unlawful.

The facts of this case do not fit neatly into the jurisprudence on advertising restraints. The cases upon which the majority rely involve complete advertising bans or limitations on the content that advertisements could contain, neither of which is present here. Such restraints prevent price signals from reaching the market, whereas the Trademark Settlements are alleged only to reduce the opportunity of certain consumers—specifically, those searching for 1-800 Contacts’ trademarks—to see advertisements paid for by other sellers in response to those searches. In addition, none of the cases the majority cite implicate intellectual property rights, the presence of which necessarily changes the analysis because the Commission must account for a competing federal policy.

i. California Dental Supports the Application of the Traditional Rule of Reason Here.

The California Dental experience, sunny and painful though it must have been, makes clear that we should not truncate the traditional rule of reason here. In that case, the California Dental Association adopted a policy that “effectively prohibited members from advertising price discounts in most cases, and entirely precluded advertising regarding the quality of services.” In re Realcomp II Ltd. (“Realcomp I”), Dkt. No. 9320, 2007 WL 6936319, at *20 (F.T.C. Oct. 30, 2009), aff’d sub nom., Realcomp II, Ltd. v. FTC (“Realcomp II”), 635 F.3d 815 (6th Cir. 2011), cert. denied, 565 U.S. 942 (2011). Limitations on price and quality advertising have a more obvious direct effect on the price setting mechanism of the market because they prevent information about price and quality from spreading. See, e.g., Cal. Dental, 526 U.S. at 773 (“The explanation proffered by the Court of Appeals for the likely anticompetitive effect of the [California Dental Association]’s restrictions on discount advertising began with the unexceptionable statements that price advertising is fundamental to price competition, and that restrictions on the ability to advertise prices normally make it more difficult for consumers to find a lower price and for dentists to compete on the basis of price”) (internal citations, quotation marks, and alterations omitted). Yet the Court applied the traditional rule of reason, because there was an insufficiently strong and obvious connection between the restraint and the price setting mechanism of the market for dental services. See Cal. Dental, 526 U.S. at 759, 774-78. The test, as the majority correctly note, is whether “the normal linkage between advertising restrictions and price/output effects in the underlying product market [i]s attenuated”. Op. at 42. The link between the restraints here and price or output effects is far more attenuated than that in California Dental. As a threshold matter, Complaint Counsel did not demonstrate any output effect. See ID at 153 n.36 (“Complaint Counsel does not contend that the Challenged Agreements reduced the output of contact lenses.”). The Trademark Settlements permit advertising, including on price and quality. They do not restrict the content of advertisements that 1-800 Contacts or the counterparties can run in innumerable contexts, including in response to search queries. And, of course, the Trademark Settlements do not bind sellers of contact lenses that are not parties to those agreements. In all of these ways, information about prices continued to reach the market. For a subset of potential contact lens customers—who search specifically for VOLUME 166 Dissenting Statement “1-800 Contacts”—the Trademark Settlements reduce one avenue for discovering products offered by certain other sellers of contact lenses. But, even for those customers not looking for 1-800 Contacts’ website,12 the cost of additional discovery is minimal: another search, a scroll down the results page, a moment’s hesitation. Given that the California Dental Court applied the traditional rule of reason to analyze restraints with a more obvious anticompetitive impact, a fortiori, the restraints here should not be analyzed under a harsher standard. Actavis supports this conclusion.13 In that case, the Supreme Court rejected a “quick look” (i.e., inherently suspect) approach when analyzing three reverse payment settlements resolving Hatch-Waxman patent infringement litigation. See Actavis, 570 U.S. at 158-59. It did so even though the alleged conduct at issue was far more harmful to competition than anything at issue here, as well-established economic evidence demonstrated. In particular, the FTC alleged that Solvay, a maker of branded pharmaceuticals, paid millions of dollars to Actavis and other generic pharmaceutical manufacturers to delay their entry into the market for Androgel (a transdermal gel formulation of testosterone). Id. at 145; see also id. at 154 (describing the settlement payments as potentially “a share of [the brand’s] monopoly profits that would otherwise be lost in the competitive market”). The anticompetitive price effects caused by such settlements were well-established by studies conducted by the Commission. See, e.g., Brief for the Petitioner at 8, Actavis, 570 U.S. 136 (No. 12-416); Fed. Trade Commu, Pay-for-Delay: How Drug Company Pay-Offs Cost Consumers Billions, at 8 (2010). Compared to a limited restriction within one channel of advertising, the complete exclusion of generic competition from the market in exchange for a share of the brand’s monopoly profits—keeping prices at supracompetitive levels—is clearly worse for consumers. While Actavis may not, as the majority contend, “stand for the proposition that no restriction in a settlement agreement . . . can be inherently suspect”, Op. at 35, it clearly does not support treating less egregious restrictions as presumptively unlawful.

The majority attempt to distinguish California Dental by limiting its holding to professional services. See id. at 21-22, 42. But the Court did not do so, applying its rule to situations that “fail[ ] to present a situation in which the likelihood of anticompetitive effects is [ ] obvious”. Cal. Dental, 526 U.S. at 771. It has continued to rely upon that case outside of the professional services context. In Actavis, the Court applied California Dental to find that reverse payment settlements did not meet the criteria necessary to abandon “the ‘rule of reason’ in favor 12 According to Respondent’s expert, Dr. Anindya Ghose, “the academic literature and the data [ ] indicate that the vast majority of consumers searching for 1-800 Contacts’ trademark do so with navigational intent.” RX0733 (Ghose Expert Report) at 060.

13 I agree with the majority’s conclusion that the Supreme Court’s ruling in Actavis does not immunize the Trademark Settlements from liability. See Op. at 12-16. That said, I do not believe the majority opinion applies Actavis properly to the facts of this case. In Actavis, the Supreme Court rejected the “scope of the patent” test, which would have rendered all settlements of patent infringement claims immune to antitrust liability. Actavis, 570 U.S. at 147. There are four issues from Actavis worthy of note here: the Supreme Court (1) created an exception, (2) did not assess the underlying infringement claim, (3) called for traditional rule-of-reason treatment of the reverse payment settlement agreement at issue there, and (4) saw indicia of anticompetitive conduct in the reverse payment settlement that are not present here. For the reasons stated elsewhere in this dissenting statement, we should follow Actavis and (a) refrain from making a judgment on the underlying infringement claim and (b) apply the traditional rule of reason. 329 1-800 CONTACTS, INC. Dissenting Statement of presumptive rules (or a ‘quick-look’ approach)” because it was not the case that “‘an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect on customers and markets.’” Actavis, 570 U.S. at 159 (quoting Cal. Dental, 526 U.S. at 770). Other courts have similarly applied the logic of California Dental beyond the professional services context. See, e.g., California ex rel. Harris v. Safeway, Inc., 651 F.3d 1118, 1137-39 (9th Cir. 2011) (en banc). While the potential procompetitive benefits of the advertising restrictions in the context of professional services helped persuade the Court to apply the rule of reason, see Cal. Dental, 526 U.S. at 771-73, the broader takeaway is that grappling with countervailing considerations gave it pause before classifying as presumptively unlawful restraints more obviously problematic than those at issue here. See, e.g., Polygram I, 136 F.T.C. at 340 (“The Court [in California Dental] concluded that . . . in the absence of any empirical evidence supporting the theoretical basis for a presumption of anticompetitive effects, [the defendant]’s identification of plausible procompetitive justifications precluded the ‘indulgently abbreviated’ review of the Ninth Circuit.”) (citing Cal. Dental, 526 U.S. at 774-78). In this case, the plausibility of the benefits that the protection of intellectual property rights bring to competition “rules out the indulgently abbreviated review” provided by the majority. Cal. Dental, 526 U.S. at 778.14 “The obvious anticompetitive effect that triggers abbreviated analysis has not been shown.” Id. ii. Polygram Does Not Support an “Inherently Suspect” Approach.

The majority rely on Polygram to support their categorization of the Trademark Settlements as inherently suspect. Polygram involved a worldwide and total ban on advertising. See Polygram I, 136 F.T.C. at 354-58, 372; cf. id. at 340 (distinguishing California Dental because the restrictions at issue in California Dental “did not ban advertising completely”). In addition to agreeing not to advertise at all, the Polygram defendants agreed not to discount the albums they were selling. Polygram II, 416 F.3d at 37. That is, they fixed prices—conduct long condemned as per se illegal. Id. Treating the price fixing agreement and the complete advertising ban together,15 the D.C. Circuit focused on the former: “An agreement between joint venturers to restrain price cutting and advertising with respect to products not part of the joint venture looks suspiciously like a naked price fixing agreement between competitors, which would ordinarily be 14 It is no answer at this stage in the analysis to say that 1-800 Contacts’ underlying infringement claims were weak, a fact-specific judgment we should avoid for the reasons I discuss below. See Section 0(A)(4)(a), infra. Were it so, the analytical framework we apply, a legal question, would depend on a highly-factual inquiry. 15 Even if the advertising restrictions at issue in Polygram were treated separately from the price fixing agreement (contrary to the D.C. Circuit’s approach), that case still does not support a finding that the Trademark Settlements are inherently suspect. In Polygram, the Commission and the D.C. Circuit found that both restraints (advertising and price fixing) were severable from the underlying joint venture. See Polygram II, 416 F.3d at 37; Polygram I, 136 F.T.C. at 359. This was a critical analytical step toward the finding that the agreement was inherently suspect because—without the underlying joint venture—the restraints became standalone (i.e., naked) agreements between direct competitors not to compete in significant ways. See Polygram II, 416 F.3d at 37; Polygram I, 136 F.T.C. at 359, 361, 363, 366. Nobody has suggested that the advertising limitations at issue here are somehow severable from the Trademark Settlements. Thus, even assuming that Polygram held that the advertising ban at issue there, standing alone, was inherently suspect (which the D.C. Circuit did not), the same logic cannot apply here because the alleged advertising restraint is not severable from the Trademark Settlements. VOLUME 166 Dissenting Statement condemned as per se unlawful.” Id. (emphasis added). It was precisely because the agreement looked like price fixing—“behavior that past judicial experience . . . ha[d] shown to warrant summary condemnation”, Polygram I, 136 F.T.C. at 344-45—that the D.C. Circuit upheld the Commission’s decision to find the agreement presumptively unlawful. See Polygram II, 416 F.3d at 37-38.

There is no price fixing here. Nor is there an advertising ban. 1-800 Contacts and the counterparties to the Trademark Settlements were free to engage in any type of advertising they saw fit, including paid keyword search advertising, as long as they did not implicate each other’s trademarks. The Trademark Settlements do not look “suspiciously” like any per se illegal conduct,16 so Polygram does not support applying the “inherently suspect” framework here. iii. Other Case Law Supports Application of the Rule of Reason.

The remaining cases cited by the majority for our judicial experience likewise do not support an “inherently suspect” approach on the facts adduced here. Critically, none involve intellectual property. And all involve advertising restrictions that bear no resemblance to the Trademark Settlements because the restraints at issue were: (1) complete bans on advertising17; (2) restrictions on the content of advertisements (i.e., limitations or bans on the ability to advertise price or quality)18; or (3) restrictions akin to per se violations of the Sherman Act.19 The distinction between the restrictions at issue in those cases and the Trademark Settlements is significant, because it is obvious how a complete ban on advertising (without implicating intellectual property rights) and these other types of restrictions could be anticompetitive. Far less obvious is how some consumers not seeing advertisements in response to searches for certain trademarked terms has the same effect. That is precisely the line drawn in California Dental, and there should be no doubt on which side the Trademark Settlements fall. 16 The majority apparently want to have it both ways with respect to whether they believe the Trademark Settlements are analogous to per se illegal conduct. In one breath, they suggest that the Trademark Settlements are analogous to per se illegal bid rigging, see Op. at 14, but in the next they analyze the Trademark Settlements’ alleged harm to search engines under the rule of reason, see id. at 50-54. As discussed in more detail below, there is insufficient evidence to conclude that Trademark Settlements harmed search engines, much less constituted per se illegal bid rigging. See Section 0(E), infra.

17 See, e.g., Bates v. State Bar of Ariz., 433 U.S. 350 (1977) (state bar rule prohibiting all advertising by lawyers in newspapers or other media); Polygram II, 416 F.3d at 33 (agreement to prohibit discounts and advertising); In re Am. Med. Assn, 94 F.T.C. 701, 1979 WL 199033, at *231 (Oct. 12, 1979) (“[I]t is fair to say that almost all advertising and promotional activity is proscribed, with a few narrowly circumscribed exceptions.”). 18 See, e.g., Cal. Dental, 526 U.S. at 762 (dental association rules effectively prohibited price advertising in most cases and entirely prohibited quality advertising); Morales v. Trans World Airlines, Inc., 504 U.S. 374, 388-89 (1992) (state restrictions on airlines fare (i.e., price) advertising); In re Mass. Board of Registration in Optometry, 110 F.T.C. 549, 1988 WL 1025476, at *27-*29 (June 13, 1988) (complete ban on truthful advertising of discount prices and other categories of advertising).

19 See, e.g., BMI, 441 U.S. at 4 (agreements to fix prices); Natl Collegiate Athletic Assn v. Bd. of Regents of the Univ. of Okla., 468 U.S. 85, 98-101 (1984) (horizontal price fixing and restrictions on output); NTSP, 528 F.3d at 352 (horizontal price fixing); Blackburn v. Sweeney, 53 F.3d 825, 828-29 (7th Cir. 1995) (horizontal agreement to allocate markets among competitors); United States v. Gasoline Retailers Assn, Inc., 285 F.2d 688, 689-91 (7th Cir. 1961) (criminal prosecution for conspiracy to fix prices). 331 1-800 CONTACTS, INC. Dissenting Statement b) We Lack Sufficient Economic Learning to Presume the Trademark Settlements Are Unlawful.

The economic studies cited by the majority do not examine paid search advertising, see Op. at 20-21, much less how restraints upon it interact with the trademark policies at issue here. The majority instead state that “the behavior of consumers and advertiser-sellers in response to this type of advertising is the same as for other types of advertising”, id. at 35, an assertion that is both unsupported and inconsistent with the majority’s position that “search-based keyword advertising” occurs in a “relatively new context”, id. at 29; see also id. at 2 (“This phenomenon is comparatively recent”). The economic evidence upon which the majority rely is insufficient to expand the scope of what we consider “inherently suspect” to include the Trademark Settlements.20 3. The Majority Should Not Have Truncated Their Rule of Reason Analysis.

Applicable precedent makes clear that the Trademark Settlements should be analyzed under the traditional rule of reason. And the cases on which the majority rely fail to provide support for truncating that analysis by applying the “inherently suspect” framework. As noted, those cases do not involve trademarks, or intellectual property of any kind. That is relevant— indeed, decisive—because trademarks often limit advertising in one way or another, and the logic of the majority’s analysis would support a rule that stigmatizes conduct protecting those rights, which is clearly procompetitive, as presumptively unlawful. Consider a situation in which a company uses a competitor’s trademark in an advertisement in a way that clearly creates confusion and, thus, infringes on a valid trademark. The mark owner sues and the parties settle, barring the conduct in question. The settlement restrains advertising. Some consumers are deprived of the opportunity to see an advertisement for a lower-priced competing product, the nub of the majority’s theory in this case. And the alleged infringer, which sells that competing product, reaches fewer customers because it is unable to use the more desirable advertising scheme. While the majority eschew the result, see Op. at 40, their logic would treat this settlement agreement as “inherently suspect” (i.e., presumptively illegal).

The answer is to follow the one case cited by the parties that considers a trademark settlement in the context of antitrust law: the Second Circuit’s decision in Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50 (2d Cir. 1997). Clorox, the only truly analogous case, and far more so than any case upon which the majority rely, makes clear that the Trademark Settlements should be evaluated using a traditional rule of reason analysis with appropriate recognition of trademark policy.

20 The majority also appear to require 1-800 Contacts to prove that paid search advertising is different from other types of advertising. See Op. at 34-35. This places the burden of proof on the wrong party; it is Complaint Counsel’s burden to show that paid search advertising operates the same as other types of advertising. VOLUME 166 Dissenting Statement a) Summary of Clorox.

Clorox involved an antitrust challenge brought by Clorox (the then-current owner of the Pine-Sol trademark) against Reckitt (the then-current owner of the Lysol trademark) regarding a trademark settlement agreement executed by the parties’ predecessors-in-interest. See Clorox, 117 F.3d at 52. The agreement restricted how Clorox could advertise Pine-Sol products and what products Clorox could sell under the Pine-Sol brand. Id. at 53-54. After acquiring the Pine-Sol mark, Clorox sued Reckitt claiming that the settlement agreement was anticompetitive because it restricted Clorox’s ability to compete using the Pine-Sol mark and served no legitimate trademark purpose because there was no longer a likelihood of consumer confusion between the marks. Id. at 54.

The Second Circuit started its analysis with the proposition that trademark settlements are “common, and favored, under the law.” Id. at 55 (citing J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 18:25 (4th ed. 1996) [hereinafter “McCarthy 4th Edition”]) (other citations omitted).21 The court presumed that “arms-length [trademark settlement] agreements are pro-competitive”, id. at 60, and that “[e]fforts to protect trademarks, even aggressive ones, serve the competitive purpose of furthering trademark policies”, id. at 61. The rule declared by the Second Circuit was not absolute and would not apply where a trademark settlement was a pretext for a per se violation of the antitrust laws.22 Id. at 55-56 (“Unlike trademark agreements that in reality serve to divide markets and thus have been condemned as illegal per se under the antitrust laws, the agreement at issue here merely regulates the way a competitor can use a competing mark. Contrary to Clorox’s argument, the agreement does not effect any of the types of restraints that have historically been condemned as illegal per se, such as price fixing, market divisions, tying arrangements, or boycotts.”) (internal citations omitted); see also id. at 60 (“[I]n the absence of any evidence that the provisions relating to trademark protection are auxiliary to an underlying illegal agreement between competitors . . . and absent exceptional circumstances, we believe the parties’ determination of the scope of needed trademark protections is entitled to substantial weight.”). Determining that the trademark settlement at issue there “must” be examined under the rule of reason, id. at 56, the Clorox court gave appropriate weight to the value of trademark policy. It held that plaintiffs challenging trademark settlements under antitrust law face a “difficult task” of proving harm to competition. Id. at 56. That is so, the Second Circuit held, even when the underlying trademark settlement agreement “only marginally advances trademark policies”. Id. at 57. “[R]egardless of whether the agreement is entirely necessary to protect [the 21 See also Clorox, 117 F.3d at 60 (“[T]rademark agreements are favored in the law as a means by which parties agree to market products in a way that reduces the likelihood of consumer confusion and avoids time-consuming litigation.”).

22 The Second Circuit’s finding that the rule of reason applies unless the challenged conduct is “auxiliary to an underlying illegal agreement between competitors”, Clorox, 117 F.3d at 60, is reminiscent of Polygram, where then- Chief Judge Douglas H. Ginsburg found that “under the Commission’s own framework, the rebuttable presumption of illegality arises not necessarily from anything ‘inherent’ in a business practice but from the close family resemblance between the suspect practice and another practice that already stands convicted in the court of consumer welfare.” Polygram II, 416 F.3d at 37.

333 1-800 CONTACTS, INC. Dissenting Statement defendant’s] trademark rights”, the Second Circuit held that a plaintiff still was required to show that an alleged anticompetitive restraint “may significantly harm competition as a whole”. Id. at 57 (emphasis added) (citation omitted). After performing a rule of reason analysis, the Second Circuit held that the trademark settlement agreement at issue there did not violate the antitrust laws. Id. at 60-61.

b) Applying Clorox to the Trademark Settlements.

Clorox is on all fours with this case: an ex post antitrust challenge to an agreement that settled trademark infringement litigation.23 And the restraint at issue here does not involve a settlement that is a pretext for a per se violation of the antitrust laws, so it does not fall into the exception to the rule of reason described by the Second Circuit. See Id. at 55-56, 60. As a result, the Commission should analyze the Trademark Settlements under the traditional rule of reason— without treating the Trademark Settlements as inherently suspect—as the Second Circuit did in Clorox.

Complaint Counsel “faces a difficult task” to show that the Trademark Settlements “significantly harm competition as a whole”, see id. at 56, 57 (emphasis added), a burden they have not met here.24 The inquiry is not simply whether the Trademark Settlements limited competition; some impact on competition is acceptable as a predictable result of the trademark policy, as the Clorox court addressed directly:

It may well be that the restrictions in the [trademark settlement] agreement prevent Clorox from competing as effectively as it otherwise might. . . . The antitrust laws do not guarantee competitors the right to compete free of encumbrances, however, so long as competition as a whole is not significantly affected. . . . [T]he fact that Clorox can still compete despite the [trademark settlement] Agreement, and that numerous other companies are also capable of competing against Reckitt, seriously undermines Clorox’s [antitrust] claim.

23 As discussed below, see Section 0(A)(4)(a), infra, I disagree with the majority’s characterization of the Trademark Settlements as “unusual”. See Op. at 13-14. The Trademark Settlements, much like the agreement at issue in Clorox, “merely regulate[ ] the way a competitor can use a competing mark.” Clorox, 117 F.3d at 55-56. The majority do not dispute that bidding on a trademarked keyword constitutes a “use” under the Lanham Act. In addition, the Second Circuit in Clorox held that courts should give “substantial weight” to the scope of agreements settling trademark infringement litigation. Id. at 60.

24 Even if the majority were correct that the Trademark Settlements constitute a naked restraint of trade, they still may not be anticompetitive. As a leading antitrust treatise noted, “even a ‘naked’ horizontal market-division agreement is competitively harmless if it occurs in a competitive market in which the defendants are merely a few among several serious players or if the restraint does not suggest a significant potential for reducing marketwide output.” PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW, ¶ 2046(b)(4) (emphasis added) (discussing Clorox as an example that fits this general statement). The market for the retail sale of contact lenses is clearly competitive and, according to Judge Chappell, “Complaint Counsel does not contend that the Challenged Agreements reduced the output of contact lenses.” ID at 153 n.36. VOLUME 166 Dissenting Statement Id. at 59 (citations omitted). The limited advertising restrictions contained in the Trademark Settlements may well prevent 1-800 Contacts and the counterparties to the Trademark Settlements from competing free of encumbrances. The record reflects that competitors’ advertisements may be less effective without the use of 1-800 Contacts’ trademarks. But these restrictions do not significantly affect competition as a whole because the counterparties to the Trademark Settlements are still capable of competing against 1-800 Contacts—including by selling to whomever they wish, advertising aggressively, and even buying advertisements on search engines, just not all advertisements—as are numerous other sellers of contact lenses, including other online retailers (e.g., Lens.com), independent eye care professionals (“ECPs”), optical retail chains (e.g., Visionworks), mass merchants (e.g., JCPenny), and club stores (e.g., Costco).25 To paraphrase the Clorox court, the fact that the counterparties can still compete despite the Trademark Settlements, and that numerous other companies are also capable of competing against 1-800 Contacts, seriously undermines Complaint Counsel’s claim. c) The Majority Fail to Distinguish Clorox.

Clorox is the most directly applicable precedent, and the majority’s attempts to distinguish it are not convincing. They point to the purported strength of the trademark infringement claim in Clorox, contrasting it with what they believe were weak claims asserted by 1-800 Contacts. See Op. at 26-27. As discussed below (see Section 0(A)(4)(a), infra), precedent, both parties in this case, the ALJ, and good policy all counsel against the Commission substituting its own view of the quality of non-sham intellectual property infringement claims for the business judgment of the contracting parties.26 Even if the majority’s assessment of 1-800 Contacts’ infringement claims were accurate, Clorox remains applicable for at least two reasons. First, the Clorox court made clear that its rule applied even to weak trademark claims. Clorox, 117 F.3d at 57 (noting that its analysis should apply “[e]ven if [a settlement] agreement only marginally advances trademark policies”). Second, as authoritative antitrust commentators have noted, the trademark claims at issue in Clorox were, in fact, not strong at all. The authors of one prominent treatise questioned “whether the Pine-Sol name manifested a confusing similarity to the older Lysol name”, noting that the Patent and Trademark Office examiner’s conclusion that the marks were similar was “somewhat dubious”. PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW: AN ANALYSIS OF ANTITRUST PRINCIPLES AND THEIR APPLICATION, ¶ 2046(b)(4) (2018).

The majority also distinguish Clorox as involving only two competitors, whereas the “Challenged Agreements covered 14 different online contact-lens retailers that account for 79 percent of online contact lenses in the United States” and “cover the landscape of online contact- 25 The majority have not defined a relevant product market (see Section 0(D), infra), so they cannot claim that competition from companies other than the pure play online retailers do not compete directly with 1-800 Contacts. 26 Consider the converse: a trademark infringement claim that everyone agrees is strong. Would Clorox then apply? If so, then the majority appears willing to put the factual cart (claim strength) before the analytical horse (inherently suspect). And, if Clorox still would not apply, would the majority deem “inherently suspect” a settlement of an unquestionably strong trademark infringement claim? 335 1-800 CONTACTS, INC. Dissenting Statement lens retailers”.27 Op. at 33. In the absence of a properly defined relevant product market (see Section 0(D), infra), however, neither the numerosity of the Trademark Settlements nor what portion of “online contact-lens retailers” they cover is meaningful—it is far from clear, in other words, that settlements with fourteen companies here are meaningfully different from the one settlement involving two companies at issue in Clorox. The record reflects that sales by online retailers account for only 17% of total contact lens sales in the United States, IDF 491, and the Trademark Settlements do not include certain large online sellers, such as Lens.com, that account for at least 21 percent of online sales. See Op. at 8, 33. If the majority believe that there is some smaller relevant market in which 1-800 Contacts has market power, they should define that market.

The majority go on to argue that “[p]redictably, Clorox was unable to muster much evidence of consumer harm.” Id. at 33. But they focus on the wrong reason that the lack of such evidence was predictable. The Second Circuit in Clorox noted the consensus that “trademarks are by their nature non-exclusionary” because “unlike other intellectual property rights, [a trademark] does not confer a legal monopoly on any good or idea”. Clorox, 117 F.3d at 56. Trademark owners cannot prevent others from manufacturing and selling identical goods under a different mark and, as a result, “the opportunity for effective antitrust misuse of a trademark, as distinguished from collateral anti-competitive activities on the part of the manufacturer or seller of the goods bearing the mark, is so limited that it poses a far less serious threat to the economic health of the nation.” Id. (quoting Carl Zeiss Stiftung v. V.E.B. Carl Zeiss, Jena, 298 F. Supp. 1309, 1314 (S.D.N.Y. 1969), aff’d in relevant part, 433 F.2d 686 (2d Cir. 1970), cert. denied, 403 U.S. 905 (1971); and citing McCarthy 4th Edition, § 31:96). Thus, the difficulty of showing harm was not specific to Clorox and it is not specific to this case;28 rather, it applies to trademark cases generally.

Contrary to Clorox, and citing Actavis, the majority believe that the Commission should second guess the form and scope of all settlements of trademark infringement litigation. See Op. at 33-34. This approach misses the mark in two important ways. First, the record reflects that non-use agreements are standard means of settling trademark disputes, see RX0734 (Hogan Expert Report) at 096, and bidding on trademark terms as keywords is a recognized “use in commerce” under the Lanham Act, see, e.g., Rescuecom, 562 F.3d at 127, 129-30. Second, it was the form of the settlement in Actavis—namely the splitting of monopoly profits among the settling parties to the detriment of consumers—that led the Court to open the door to liability. See Actavis, 570 U.S. at 153-56. The Trademark Settlements include no splitting of monopoly profit—indeed, no material amount of money changed hands. See Section 0(0)(0), supra. Nor, again, are they associated with the kind of conduct—price fixing, etc.—that has raised the suspicion of courts. See Section 0(A)(2), supra.

27 This portion of the majority opinion is just one of the several instances in which the majority inappropriately group the Luxottica Agreement in with the Trademark Settlements. See Section 0, infra; see also Op. at 10 (defining the term “Challenged Agreements” to encompass the Trademark Settlements and the Luxottica Agreement). 28 As discussed below, I do not believe Complaint Counsel has met its burden to show direct anticompetitive effects in this case. See Section 0(B), infra.

VOLUME 166 Dissenting Statement According to the majority, Clorox involved labeling and, therefore, is not applicable here. See Op. at 14 (citing Clorox, 117 F.3d at 57); see also Draft Oral Arg. Tr. 43:19-44:6 (Complaint Counsel asserting that Clorox “was a case about labeling.”). The majority cite no case for the proposition that, for trademark law purposes, labeling and advertising are categorically different, nor am I aware of any. Courts apply the same fact-specific test to determine the likelihood of customer confusion regardless of whether the use of the trademark was on a label or in an advertisement. See, e.g., Pom Wonderful LLC v. Hubbard, 775 F.3d 1118, 1125-31 & n.7 (9th Cir. 2014) (analyzing both “labels and advertising materials” under the eight-factor test for likelihood of customer confusion developed in AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348- 49 (9th Cir. 1979), abrogated on other grounds by Mattel Inc. v. Walking Mountain Prods., 353 F.3d 792, 810 n.19 (9th Cir. 2003)). Leaving aside the law and considering the facts, the non-use agreement in Clorox operated in a manner similar to the Trademark Settlements, which themselves are a type of non-use agreement. The non-use agreement at issue in Clorox did not restrict Clorox or other firms from producing and selling products in direct competition with the Lysol brand as long as Clorox did not put the name “Pine-Sol” on those products. Op. at 14. And, likewise, the Trademark Settlements do not “in any way restrict [the other online contact lens retailers] from producing and selling products that compete directly with the [1-800 Contacts] brand,” so long as they do not advertise in response to searches for 1-800 Contacts’ trademarks. See Clorox, 117 F.3d at 57. That is a critical distinction under California Dental, because it demonstrates how price signals can continue to reach the market, making the link between the restraints and any price effect attenuated.

It also bears repeating that the Second Circuit in Clorox stated that the form and scope of trademark settlement agreements deserve “substantial weight” because the settling parties “are in the best position to determine what protections are needed” and “it is usually unwise for courts to second-guess such decisions.” Id. at 60. Thus, even if the form or effect of the Trademark Settlements differed substantially from those at issue in Clorox, the Commission should give the parties’ desired means of settlement deference because of the property right at issue and the absence of an auxiliary illegal agreement. Id.

4. The Majority’s Rule Will Have Negative Consequences. Treating the Trademark Settlements as “inherently suspect” yields an unclear rule that, regardless of interpretation, will, I fear, create uncertainty, dilute trademark rights, and dampen inter-brand competition. The majority couch their holding as a limited one dealing with restraints on the opportunity to make price comparisons, but, by adopting an analytical framework without accounting for the intellectual property at issue, they produce one of the following rules: either all advertising restrictions are inherently suspect, regardless whether they protect intellectual property rights, or the level of scrutiny applied to a particular restraint will depend on the strength of the trademark holder’s underlying infringement claim. The majority make it clear that they do not intend to label all advertising restrictions “inherently suspect”, see Op. at 22, but several parts of their analysis suggest precisely such a conclusion. First, their determination that the Trademark Settlements are “inherently suspect” avoids any mention whatsoever of trademarks. See id. at 18-22. The majority rely heavily upon 337 1-800 CONTACTS, INC. Dissenting Statement Polygram, but untether the advertising ban from the ban on discounting that led the D.C. Circuit to find liability. So the assertion stands alone, regardless of the existence of intellectual property. Second, the majority rely on precedents that do not involve trademarks, or intellectual property of any kind, and dismisses the one case—Clorox—that looks at a trademark settlement through the lens of antitrust law. In doing, they effectively declare that any advertising restraint is “inherently suspect”, regardless whether such restraint is intended or necessary to protect intellectual property.29 The majority cast this case as unique because the Trademark Settlements reduced the opportunity of some consumers to see some advertisements sometimes, but this description has no apparent limiting principle. Advertising is designed to grab attention, including through the use (or misuse) of trademarks. All of this raises a serious concern that the rule the majority today promulgates (i) is overbroad and (ii) will reach procompetitive conduct. Because the majority explicitly eschew a rule condemning all advertising restrictions, regardless whether they protect intellectual property rights, their reasoning suggests a rule under which the standard of review depends on (the Commission’s view of) the strength of the underlying trademark infringement claim. For infringement claims that the Commission deems weak or implausible, the challenged restraint will be deemed “inherently suspect”. This approach leaves open the question of how the majority would treat infringement claims that they believe are strong. Such a rule would put the factual cart ahead of the analytical horse, is wrong as a matter of law, and will require the Commission to litigate (or re-litigate) the underlying infringement claim—in every case—to determine what standard of review it will apply. That is precisely what happened here.

a) The Commission Should Not Litigate Inherently Fact-Intensive Infringement Claims.

The majority claim that they are not evaluating the underlying infringement claims. See Op. at 40 (“We are neither deciding matters of trademark law nor suggesting that to determine whether the Challenged Agreements unreasonably restrain competition, we need to conduct a mini-trial on the merits of the underlying trademark litigations.”). But that is not the approach reflected in their opinion. Instead of following Clorox and according trademarks their appropriate weight, the majority rest several key conclusions on the premise that 1-800 Contacts’ underlying trademark infringement claims were weak. The majority:  Ignore the presence of 1-800 Contacts’ intellectual property in their “inherently suspect” analysis, see id. at 18-22;

 Opine that customer confusion—part of 1-800 Contacts’ trademark infringement claims—is not at issue when evaluating 1-800 Contacts’ procompetitive justifications, see id. at 27;

29 For instance, the majority assert that “[r]estricting the availability of truthful information that guides consumer decisions in the marketplace is a competitive harm.” Op. at 43. VOLUME 166 Dissenting Statement  Distinguish Clorox based on a value judgment that the trademark infringement claims at issue in that case were somehow stronger than 1-800 Contacts’ infringement claims, see id. at 26-27;

 Dismiss 1-800 Contacts’ trademark infringement claims based on an abbreviated evaluation of consumer confusion,30 which is a deeply factual issue, see id. at 28- 29; and  Reject 1-800 Contacts’ trademark-related procompetitive justification based on their view of the strength of the underlying infringement claims, see id. at 37-40 & n.42.

The majority do all of this notwithstanding the universal agreement (from the ALJ, Complaint Counsel, and Respondent) that evaluating the relative strength of 1-800 Contacts’ infringement claims is unnecessary, improper, or both. ID at 171 (“[D]elving into the merits of 13 trademark lawsuits, after the fact, to determine whether or not 1-800 Contacts could ultimately have proven infringement, if even possible, would require unacceptable speculation and would constitute an unnecessary waste of judicial resources.”) (citing In re Schering-Plough Corp. (“Schering-Plough I”), 136 F.T.C. 956, 997 (2003), vacated on other grounds sub nom., Schering-Plough Corp. v. FTC (“Schering-Plough II”), 402 F.3d 1056 (11th Cir. 2005), cert. denied, 548 U.S. 919 (2006)); Draft Oral Arg. Tr. 49:5-13 (Complaint Counsel stating that “[i]t does not make a whit of difference whether 1-800 Contacts would have lost or won every single case it brought.”); id. at 59:14-23 (Complaint Counsel explaining why “we don’t need to evaluate the merits of the trademark claim”); RAB at 37-38. The reason the parties agree on this is clear: both precedent and sound policy counsel against having antitrust liability turn on ex post fact-intensive inquiries into the validity of non-sham intellectual property infringement claims. See Schering-Plough I, 136 F.T.C. at 997 (“An after-the-fact inquiry by the Commission into the merits of the underlying litigation is not only unlikely to be particularly helpful, but also likely to be unreliable.”). Complaint Counsel has taken a similar position in other litigation. See, e.g., Reply Brief for the Petitioner at 6, Actavis, 570 U.S. 136 (No. 12-416) (“We agree that the antitrust analysis of a Hatch-Waxman [reverse payment] settlement should not turn on a judicial assessment of the strength or scope of the particular patent involved in the case.”) (emphasis in original).

Actavis makes it clear that the Commission should not be in the business of evaluating the underlying infringement case when deciding an antitrust challenge; indeed, the Actavis Court explicitly declined do so. See Actavis, 570 U.S. at 159 (“To say this is not to require the courts to insist . . . that the Commission need litigate the patent’s validity”); id. at 158 (“[A] court, by examining the size of the payment, may well be able to assess its likely anticompetitive effects along with its potential justifications without litigating the validity of the patent”) (emphasis 30 The majority claim that their opinion does not hinge on the merits of the trademark infringement claim. Op. at 28 n.27. As this section demonstrates, the majority’s view of the strength of 1-800 Contacts’ infringement claims permeates their opinion. To the extent their opinion also applies to settlements of “strong” infringement claims, the majority do not answer the question of what standard would apply, or how that fact would bear on the analysis of a respondent’s procompetitive justifications.

339 1-800 CONTACTS, INC. Dissenting Statement added); id. at 153 (recognizing “the patent litigation problem”). The Court’s willingness to subject the reverse payment settlements to rule-of-reason analysis stemmed not from the underlying merits but from the “unusual” nature of the settlements, including large payments by the plaintiff-branded pharmaceutical manufacturer to the defendant-generic pharmaceutical manufacturer in exchange for the generic staying out of the market entirely, which kept prices high while the brand and generic manufacturers split the monopoly profits. See id. at 154. The Trademark Settlements are nothing like that: no material amount of money changed hands, so the settling parties did not divide monopoly profits at the expense of the consumer, and, most importantly for the present case, no supplier of contact lenses agreed to stay out of the market.31 The general rule of not evaluating the merits of non-sham intellectual property claims is particularly apropos in the trademark infringement context because the legal issues generally— and customer confusion in particular—involve fact-specific inquiries that should be decided by a judge or jury. As the Fourth Circuit held in a case that also involved alleged trademark infringement caused by paid keyword search advertising, “the likelihood of confusion issue . . . is ‘an inherently factual issue that depends on the facts and circumstances in each case.’” Rosetta Stone Ltd. v. Google, Inc., 676 F.3d 144, 153 (4th Cir. 2012) (quoting Lone Star Steakhouse & Saloon, Inc. v. Alpha of Va., Inc., 43 F.3d 922, 933 (4th Cir. 1995)); see also Hearts on Fire Co., LLC v. Blue Nile, Inc., 603 F. Supp. 2d 274, 288 (D. Mass. 2009) (refusing to grant the defendant’s motion to dismiss because where “a plaintiff has alleged a plausible likelihood of confusion based on the overall context in which a consumer performs his internet search, he has stated a claim for trademark infringement and may proceed on an initial interest theory.”) (internal reference omitted); Fair Isaac Corp. v. Experian Info. Solutions Inc., 645 F. Supp. 2d 734, 761 (D. Minn. 2009) (refusing to grant summary judgment because “genuine issues of material fact” remained regarding “whether Defendants’ purchase of keywords including [Plaintiff]’s trademarks, which caused Defendants’ websites to appear on the results page when a consumer ran an internet search consisting of those keywords, created a likelihood of confusion”), aff’d on other grounds, 650 F.3d 1139 (8th Cir. 2011); Soilworks, LLC v. Midwest Indus. Supply, Inc., 575 F. Supp. 2d 1118, 1132 (D. Ariz. 2008) (granting summary judgment on mark owner’s trademark infringement counterclaim because “the undisputed evidence in this case establishes that [the counterclaim-defendant] diverts the initial attention of potential Internet customers to its websites by using [the counterclaim-plaintiff’s] trademark in keywords and metatags.”).

Although the Commission should not evaluate the underlying infringement claim, the majority overstates the clarity of trademark law at the time of the Trademark Settlements. The record reflects that the parties entered the Trademark Settlements precisely because of the possibility that bidding on trademarked terms as keywords created liability for infringement, a reality exacerbated by the Rescuecom decision. See, e.g., IDF 333, 349; Holbrook, Tr. 1942:12- 13; CX9024 at 017 (Holbrook Dep. 63:13-18); Hogan, Tr. 3256:11-19, 3260:21-3261:4. 31 The majority claim that the Trademark Settlements are likewise “unusual” because they “reach[ ] farther than a cure based on rewording a label or an ad”. Op. at 14. As Clorox—the case the majority cite for this proposition— makes clear, even aggressive assertions of trademark rights are procompetitive. Clorox 117 F.3d at 60-61. A standard non-use restriction that goes farther than an ex post proposed remedy does not take us out of that category, much less provide a basis for antitrust liability.

VOLUME 166 Dissenting Statement The majority also address Soilworks only in passing. See Op. at 38. In that case, a federal district court granted summary judgment because it found that the mark owner (Midwest) had met its burden to show that the use of its trademarks as keywords in paid search advertising and metatags by the alleged infringer (Soilworks) caused initial interest confusion.32 Soilworks, 575 F. Supp. 2d at 1132. The Soilworks court considered, inter alia, the similarity between the keyword purchased by the alleged infringer and the trademark, the relatedness of the goods sold by the parties, and the marketing channels employed by the two companies.33 Id. at 1131. All of these factors would weigh in favor of a finding for 1-800 Contacts on a claim for initial interest confusion. And the district court’s holding in Soilworks could have applied equally to one of 1- 800 Contacts’ trademark infringement claims:

A person typing “soil sement” into a search engine presumably would be somewhat familiar with Midwest’s product and would be looking for the product or its maker, and yet would be directed by the keywords and metatags to Soilworks’ websites. The confusion—thinking one would be connected to Midwest when in fact Soilworks’ websites also appear in the search results—would entirely be caused by Soilworks’ use of Midwest’s mark.

Id. at 1132. The majority’s dismissal of Soilworks as “a single district court summary judgment decision from over ten years ago”, Op. at 38, fails to account for the fact that Soilworks was decided just ten months before the wave of Trademark Settlements that followed Rescuecom began, and was therefore precisely the type of case that the settling parties would have considered at the time they entered the Trademark Settlements. Like Rescuecom, Soilworks predated almost all of the Trademark Settlements. Those cases and other developments fed the legal uncertainty surrounding paid search advertising using trademarked keywords. Allegations of infringement based on trademark keyword bidding withstood dispositive motions. See id. at 38 & n.40.34 And a judge could have ordered the same 32 The district court in Soilworks distinguished initial interest confusion from source confusion: Although the core element of trademark infringement is whether the similarity of the marks is likely to confuse customers about the source of the products, the Ninth Circuit and other courts have recognized a variation of trademark infringement that does not require such confusion. Under the ‘initial interest confusion’ theory of trademark liability, ‘source confusion’ need not occur. Rather, initial interest confusion occurs when the defendant uses the plaintiff’s mark in a manner calculated to capture initial consumer attention. . . . When accomplished through the use of key words or metatags on the Internet, this wrongful conduct may involve no deception of the consumer. The consumer is simply led to the defendant’s website through the unseen keywords and metatags the defendant has purchased on the web.

Soilworks, 575 F. Supp. 2d at 1129-30, 1131 (internal quotation marks, alterations, citations, and footnotes omitted).

33 The district court also identified several other factors that other courts have used to evaluate consumer confusion, but found them “less relevant”, “of little import”, of “diminished importance”, “not directly relevant”, or “relatively unimportant” in the keyword/metatag context. Soilworks, 575 F. Supp. 2d at 1132 (citations omitted). 34 In addition to the cases cited by the majority, other infringement claims based on trademark keyword search advertising survived dispositive motions. See, e.g., Tokyo Broadcasting Sys. v. Am. Broadcasting Cos., Inc., No. 08- 341 1-800 CONTACTS, INC. Dissenting Statement relief that is contained in the Trademark Settlements. See, e.g., RX0679 at 005. Indeed, multiple federal judges later did. IDF 337 (“The court’s order prohibited LensWorld from purchasing 1-800 Contacts’ federally registered trademarks as keywords for any search engine advertising program and required LensWorld to implement certain negative keywords . . . where possible.”) (citation and internal quotation marks omitted); CX0144; CX0162; Pratt, Tr. 2558:5- 2559:4 (discussing CX0162). The parties may not have taken as dim a view of 1-800 Contacts’ trademark infringement claims as the Commission does today. Cases on the books at the time of the Trademark Settlements suggested that using trademarked terms as keywords could constitute infringement regardless of the content of the advertisement.35 See, e.g., Playboy Enters., Inc. v. Netscape Commc’ns Corp., 354 F.3d 1020, 1024-26 (9th Cir. 2004) (keywords and metatags); see also Brookfield Commc’ns, Inc. v. W. Coast Entm’t Corp., 174 F.3d 1036, 1057 (9th Cir. 1999) (domain names and metatags).

At most, the majority have shown that the legal status of using trademarked terms as keywords in paid search advertising was uncertain. When the settling parties entered the Trademark Settlements, courts did not “consistently reject[ ] the notion that buying or creating internet search terms” did not constitute trademark infringement. See Op. at 38-39 (quoting Tempur-Pedic N. Am., LLC v. Mattress Firm, Inc., No. 17-1068, 2017 WL 2957912, at *7 (S.D. Tex. July 11, 2017)). To the contrary, most courts viewed trademark infringement and customer confusion in the context of paid search advertising as fact specific inquiries that should be decided by judges and juries. See, e.g., Govt Emps. Ins. Co. v. Google, Inc., 330 F. Supp. 2d 700, 704 (E.D. Va. 2004). The risk of liability for trademark infringement became even more serious after the Second Circuit’s decision in Rescuecom. See Section 0(0)(C), supra. The complexity of the legal regime and the majority’s ex post determination of an inherently fact-specific question underscore the general rule that the Commission should not be in the business of litigating (or re-litigating) the underlying trademark infringement claim. cv-6550, 2009 WL 10668456, at *11 (C.D. Cal. Aug. 12, 2009); Govt Emps. Ins. Co. v. Google, Inc. (“GEICO”), 330 F. Supp. 2d 700, 704 (E.D. Va. 2004).

35 In a footnote, the majority cite a bevy of cases to support their claim that courts have consistently held that buying trademarked terms as keywords, standing alone, is insufficient to prove trademark infringement. See Op. at 39-40 n.43. However, almost all of those cases postdated the Trademark Settlements, so they could not have factored into the parties’ decision to settle 1-800 Contacts’ trademark infringement claims. See id. The few cases cited by the majority that predated the Trademark Settlements show—at most—that the legal landscape was uncertain, and support the fundamental proposition that trademark infringement and customer confusion are inherently fact-specific. See, e.g., GEICO, 330 F. Supp. 2d at 704 (“Whether defendant’s [trademark] uses . . . create a likelihood of confusion [is a] fact-specific issue[ ] not properly resolved through a motion to dismiss.”); Govt Emps. Ins. Co. v. Google, Inc., No. 04-cv-507, 2005 WL 1903128, at *4 (E.D. Va. Aug. 8, 2005) (“[T]he Fourth Circuit has emphasized that likelihood of confusion is a highly factual issue, the assessment of which depends largely on the particular circumstances of each case, . . . and that the likelihood of confusion standard does not require that a plaintiff prove actual confusion.”) (citations omitted). VOLUME 166 Dissenting Statement b) The Majority’s Approach Will Reduce Brand Investment Incentives.

Predicating antitrust liability on an ex post judgement about the strength of intellectual property infringement claims—or ignoring the context of their protection entirely—not only will reduce clarity in the law, but also threatens to chill the procompetitive investment that is one of the hallmarks of trademark law. As Complaint Counsel’s expert, Dr. Evans, put it: Trademarks help companies convey information to consumers about themselves and their products. They enable companies, for example, to use a brand name to signal to consumers that the company provides a high quality product or offers particular attributes that consumers care about. Protecting trademark rights encourages investment in this sort of brand-building activity, which in turn generates valuable market information, promotes competition and ultimately benefits consumers. Moreover, trademark policy prevents the spread of misinformation as when a company claims falsely that it produces the same brand of a competitor or tries to confuse consumers into thinking they do by using similar words.

CX8006 (Evans Expert Report) at 135. In other words, trademark protection gives companies an incentive to maintain their reputations and improve quality, which promotes competition. Park ‘N Fly, Inc. v. Dollar Park and Fly, Inc., 469 U.S. 189, 193 (1985) (“[T]rademarks desirably promote competition and the maintenance of product quality”); William M. Landes & Richard A. Posner, Trademark Law: An Economic Perspective, 30 J.L. & ECON. 265, 269 (1987) (“[T]rademark protection encourages expenditures on quality”); J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 2:4 (5th ed. June 2018) [hereinafter “McCarthy 5th Edition”].

Competition is not the only benefit of trademark protection; by encouraging brand investment, it also fosters innovation and gives more information to customers. See, e.g., Landes & Posner, supra, at 269 (“In short, a trademark conveys information that allows the consumer to say to himself, ‘I need not investigate the attributes of the brand I am about to purchase because the trademark is a shorthand way of telling me that the attributes are the same as that of the brand I enjoyed earlier.’”) (footnote omitted). “An important purpose underlying trademark law is the protection of the trademark owner’s investment in the quality of the mark and the quality of the goods or services the mark identifies. . . . ‘By contrast, if there were no trademarks . . . a manufacturer would gain little or nothing from improving his product’s quality. . . . The result would be a race to produce inferior products, rather than competition to produce better ones.’” McCarthy 5th Edition, supra, § 2:4 (quoting Richard Craswell, FTC Policy Planning Issues Paper: Trademarks, Consumer Information and Barriers to Competition, at 7 (1979)). The procompetitive benefits of trademarks are precisely why courts like the Second Circuit have encouraged zealous trademark enforcement, and declined to impose upon mark owners the fear of treble antitrust damages. See, e.g., Clorox, 117 F.3d at 61 (“Efforts to protect 343 1-800 CONTACTS, INC. Dissenting Statement trademarks, even aggressive ones, serve the competitive purpose of furthering trademark policies.”); Drop Dead Co. v. S. C. Johnson & Son, Inc., 326 F.2d 87, 96 (9th Cir. 1963) (“[T]he bringing of infringement suits based on colorable similarity rather than on exact identity . . . constitute[s] the sort of aggressive competition and promotion that anti-trust law seeks to protect”), cert. denied, 377 U.S. 907 (1964); see also Car-Freshner Corp. v. Auto Aid Mfg. Corp., 438 F. Supp. 82, 87 (N.D.N.Y. 1977) (“[T]he acts of the plaintiffs in registering and enforcing the trademark in issue . . . merely represent fair and aggressive competition which does not constitute a violation of the antitrust laws”) (citation omitted). Zealous protection is precisely what 1-800 Contacts did here.

The crux of the majority’s antitrust story underscores the point. The search engine results pages that appear in response to searches for “1-800 Contacts” were the supposed “critical battleground”36 for competition precisely—and only—because of 1-800 Contacts’ brand investment. See, e.g., CX9033 at 017 (Mohan Dep. 61:9-12); CX9039 at 026, 040 (Clarkson Dep. 97:20-98:3, 155:25-156:8). In other words, 1-800 Contacts engaged in the type of brand investment envisioned by trademark policy, and, combined with its excellent service and constant efforts to improve the customer experience, built a brand that customers trust. The company then sought zealously to protect its brand.

Assigning liability—and the potential for treble damages, no less—to this conduct will not only chill brand investment, it will chill the very competition the majority seeks to protect. c) The Policy Favoring Litigation Settlements Supports Application of the Traditional Rule of Reason.

Trademark policy is not the only one at stake. “Few public policies are as well established as the principle that courts should favor voluntary settlements of litigation by the parties to a dispute.” Am. Sec. Vanlines, Inc. v. Gallagher, 782 F.2d 1056, 1060 (D.C. Cir. 1986) (citations omitted); accord Williams v. First Natl Bank of Pauls Valley, 216 U.S. 582, 595 (1910); St. Louis Mining & Milling Co. v. Montana Mining Co., 171 U.S. 650, 656 (1898); TBK Partners, Ltd. v. W. Union Corp., 675 F.2d 456, 461 (2d Cir. 1982). That is because settlements of complex litigation allow the settling parties to avoid “a litany of direct and indirect costs”. Schering-Plough II, 402 F.3d at 1075. Consistent with this precedent, both parties’ experts agreed that settlements are economically efficient. See CX9042 at 050 (Evans Dep. 196:22-24); RX0739 (Murphy Expert Report) at 053; Murphy, Tr. 4207:22-4208:25; RX0737 (Landes Expert Report) at 017.

The majority’s rule effectively makes non-use agreements—the most common means of settling trademark infringement litigation,37 and favored in their own right on policy grounds38— “inherently suspect”, opening the door to reviewing and/or litigating many more trademark 36 Draft Oral Arg. Tr. 39:20-22; see also Op. at 14, 30, 32, 34 (describing the search engine results pages displayed in response to searches for 1-800 Contacts’ trademarked terms as the “key moment” or “crucial moment” of competition).

37 See RX0734 (Hogan Expert Report) at 096.

38 Trademark non-use agreements are “usually entered into to settle an infringement dispute”, are “not against public policy”, and “are routinely upheld and enforced.” McCarthy 5th Edition, supra, § 18:82 (footnote omitted). VOLUME 166 Dissenting Statement settlements. This will increase the risk of settling trademark infringement litigation, which is efficient in part because it reduces risk. This is particularly so where, as here, the real issue is the highly fact-specific question of confusion. The Second Circuit explained the point in Clorox: [T]rademark agreements are favored in the law as a means by which parties agree to market products in a way that reduces the likelihood of consumer confusion and avoids time-consuming litigation. Parties such as Clorox, Sterling, and their predecessors, are in a position to structure such agreements in the way that the parties believe best accommodates their interests in light of trademark law. Accordingly, in the absence of any evidence that the provisions relating to trademark protection are auxiliary to an underlying illegal agreement between competitors—such as the territorial market division condemned in Timken [Roller Bearing Co. v. U.S., 341 U.S. 593 (1951)]—and absent exceptional circumstances, we believe the parties’ determination of the scope of needed trademark protections is entitled to substantial weight. At the time of the execution of such an agreement, the parties are in the best position to determine what protections are needed and how to resolve disputes concerning earlier trademark agreements between themselves. . . . In the absence of evidence to the contrary it is reasonable to presume that such arms-length agreements are pro-competitive.

Clorox, 117 F.3d at 60.39 A rule requiring the post hoc evaluation of intellectual property infringement claims will be difficult for us to apply, but also, and more importantly, for private parties to self-administer. What level of infringement confidence is required? Are plaintiffs only allowed to settle trademark infringement claims that they know they are going to win? That certainly can’t be the 39 The Ninth Circuit rejected a challenge to a trademark settlement agreement for similar reasons: If the merits of a cause of action underlying a [trademark] compromise agreement could, as a matter of course, be inquired into in an action to enforce the settlement, neither settlement nor the policies it promotes would be fostered. The parties would be subjected to the expense, delay, and uncertainty they sought to avoid through settlement; the court would be burdened with trial of the underlying dispute and the preparation which precedes it.

MWS Wire Indus., Inc. v. Cal. Fine Wire Co., Inc., 797 F.2d 799, 802 (9th Cir. 1986); see also T & T Mfg. Co. v. A. T. Cross Co., 449 F. Supp. 813, 827 (D.R.I. 1978) (“[T]he Court must balance the public interest against confusion, one of the significant purposes of trademark law, against the interest in enforcing contracts and protecting the reliance they induce. [¶] The Court must also add into this balance the interest in encouraging extra-judicial settlement of trademark litigation. Insisting that a court review a settlement to assure that no public confusion will result would make such agreements of little value to the parties. Parties would sensibly conclude that they might better litigate the issue of confusion to conclusion rather than reach a settlement which might later be found to be unenforceable. Such a premium on litigation would lead to a further drain on judicial resources. Moreover, we note the advantage of allowing business persons to determine whether their self-interest is better served by making such contracts or not.”) (emphasis added), aff’d, 587 F.2d 533 (1st Cir. 1978), cert. denied, 441 U.S. 908 (1979). 345 1-800 CONTACTS, INC. Dissenting Statement standard. Regardless, we are ill-equipped to judge. Clarity may only result from substantial litigation that follows the majority’s opinion, animated by the prospect of treble damages. B. The Evidence That the Trademark Settlements Had Direct Anticompetitive Effects Is Insufficient.

If the Trademark Settlements are not “inherently suspect”, which they are not, Complaint Counsel can meet their initial burden of proof under the rule of reason in one of two ways: “an indirect showing based on a demonstration of defendant’s market power” or “direct evidence of ‘actual, sustained adverse effects on competition’”. Realcomp I, 2007 WL 6936319, at *31 (quoting FTC v. Indiana Federation of Dentists (“IFD”), 476 U.S. 447, 461 (1986)) (other citations omitted). The majority take only the direct approach; they do not attempt an indirect showing of market power. See Section 0(D), infra. To meet the initial burden of proof with direct evidence, a plaintiff must show adverse effects on competition that are actual, sustained, and significant or substantial. See Realcomp I, 2007 WL 6936319, at *31; Op. at 17 (“[T]he plaintiff has the burden to prove that the challenged restraint has, or is likely to have, a substantial anticompetitive effect that harms consumers.”); Ohio v. Am. Express Co. (“Amex”), 138 S. Ct. 2274, 2284 (2018) (“Under [the rule of reason] framework, the plaintiff has the initial burden to prove that the challenged restraint has a substantial anticompetitive effect that harms consumers”); Realcomp II, 635 F.3d at 831-32 (“substantial consumer harm”); Clorox, 117 F.3d at 57 (requiring the plaintiff to show that the trademark settlement agreement “may significantly harm competition as a whole”). Complaint Counsel have not met that burden with its showing on direct effects.

1. In the Context of a Trademark Settlement Agreement, a Restriction on Advertising Is, by Itself, Insufficient to Show Direct Effects. The majority first argue that Complaint Counsel established direct effects by showing that advertising was limited by the Trademark Settlements. But the Supreme Court held in California Dental that restrictions on advertising, by themselves, are insufficient to show anticompetitive harm.40 See Cal. Dental, 526 U.S. at 776. The relevant inquiry is whether an advertising restriction limited output of the underlying product or service. See id. (“The question is not whether the universe of possible advertisements has been limited (as assuredly it has), but whether the limitation on advertisements obviously tends to limit [output of the underlying product or service].”).

Other than California Dental, the only cases cited by the majority for the proposition that a reduction in advertising, by itself, is sufficient to show direct effects are Indiana Federation of Dentists (“IFD”) and Realcomp, see Op. at 42-43, neither of which supports that proposition. Indeed, neither case involved advertising, a point the majority apparently concede. See id. at 43. The majority rely on IFD for the proposition that a concerted effort to withhold “information”—a broad and nebulous category—constitutes a competitive harm and, therefore, 40 As discussed above, the majority’s attempts to distinguish California Dental fail. See Section 0(A)(2)(a)(i), supra.

VOLUME 166 Dissenting Statement any limitation on “information” constitutes direct evidence of anticompetitive effects. See id. at 42-43, 46 n.49. They misread the case. In IFD, the Supreme Court considered “a horizontal agreement among the participating dentists to withhold from their customers a particular service that they desire”, specifically, providing x-rays to insurers. IFD, 476 U.S. at 459. Thus, IFD is a case about agreeing not to provide a service, not about information or advertising. The Commission and the D.C. Circuit recognized as much in Polygram. See Polygram I, 136 F.T.C. at 335 (describing the restraint at issue in IFD as “an agreement among dentists to withhold from their customers a desired service”); Polygram II, 416 F.3d at 36 (“[I]n IFD, the Supreme Court ruled a horizontal agreement to withhold services could not be sustained”). As did the Supreme Court in California Dental, see Cal. Dental, 526 U.S. at 770, and other courts in the years since IFD. See, e.g., Major League Baseball Props., Inc. v. Salvino, Inc., 542 F.3d 290, 317 (2d Cir. 2008); Intl Healthcare Mgmt. v. Hawaii Coal. for Health, 332 F.3d 600, 606 (9th Cir. 2003). Even assuming the majority’s categorization of IFD were accurate (which it is not), nothing in IFD supports a finding that all restrictions on information (much less advertising), standing alone, constitute direct evidence of anticompetitive effects.41 The defendant in IFD implemented an outright ban on providing x-rays to dental insurers, whereas the Trademark Settlements merely raise the search costs (marginally) to a certain set of customers for information still very much available.

The majority’s reliance on Realcomp as an “information” restraint case is similarly misplaced. See Op. at 43. The conduct at issue there was a policy that prohibited the dissemination of property listing information to competitors through Realcomp’s multiple listing services (“MLS”). Realcomp II, 635 F.3d at 819. This prevented competing realtors from offering listings (i.e., their product) to their customers. See id. In other words, the restraint foreclosed access to a necessary input and directly reduced downstream output, see Realcomp I, 2007 WL 6936319 at *25, “restrict[ing] the ability of members to offer consumers products that create ‘price pressure’ on more expensive products”, id. at *5. The restraint limited output, not advertising, so the anticompetitive effect (i.e., a reduction in output) was obvious. See Realcomp II, 635 F.3d at 829-30. Realcomp cannot support a finding that reductions in advertising or information, without a concomitant reduction in output, constitute direct anticompetitive effects. According to the majority, any restriction on truthful advertising—indeed, even less, the restriction of truthful information that might impede a consumer’s ability to discover a lower price—constitutes direct evidence of anticompetitive harm. See, e.g., Op. at 43 (“Restricting the availability of truthful information that guides consumer decisions in the marketplace is a competitive harm.”). If all a plaintiff need show to establish direct effects is the existence of a restriction on advertising—regardless of justification, size, or effect—then all limits on truthful 41 Even the portion of IFD quoted by the majority does not support their position. See Op. at 43 (“As the Supreme Court explained in IFD, ‘a concerted and effective effort to withhold (or make more costly) information desired by consumers for the purpose of determining whether a particular purchase is cost justified is likely enough to disrupt the proper functioning of the price setting mechanism of the market that it may be condemned even absent proof that it resulted in higher prices or . . . the purchase of higher priced services than would occur in its absence.’”) (quoting IFD, 476 U.S. at 461-62). The x-rays at issue allowed insurers to assess the appropriateness of claims for benefits. IFD, 476 U.S. at 455. There is no similar category of information withheld here. 347 1-800 CONTACTS, INC. Dissenting Statement advertising are, effectively, inherently suspect, a result the majority specifically disclaim.42 See id. at 22. And they must, as such a rule would inevitably treat conduct that would otherwise be considered competitively neutral or even procompetitive as presumptively illegal. See, e.g., Cal. Dental, 526 U.S. at 771 (“[I]t seems to us that the [California Dental Association]’s advertising restrictions might plausibly be thought to have a net procompetitive effect, or possibly no effect at all on competition.”). A trademark non-use agreement that applies to advertising is just one example.

As a matter of law, then, the majority’s attempt to establish direct effects by looking only at advertising fails. It also fails as a matter of fact. While advertisements in response to competitors’ trademarked search terms were limited, the majority fail to establish that the amount of advertising was reduced. See Section 0(E), infra. 2. There Is Insufficient Evidence of Direct Price Effects. While restrictions on advertising are not themselves enough, the majority are correct that a showing of actual, sustained, and substantial or significant price effects would suffice. See, e.g., Amex, 138 S. Ct. at 2284; Realcomp II, 635 F.3d at 831-32; Clorox, 117 F.3d at 57; Realcomp I, 2007 WL 6936319, at *31; Op. at 17. I disagree that Complaint Counsel have met that burden here.

The majority’s finding of direct price effects rests almost entirely on the unremarkable fact that 1-800 Contacts’ prices were higher than some of its competitors’ prices. See Op. at 46- 47. The majority find that “the higher prices are a consequence of 1-800 Contacts shielding itself from competitive pressure by preventing consumers from obtaining information that would enable comparison shopping.” Id. at 47. But Complaint Counsel failed to prove that the Trademark Settlements caused the price differential.

First, the record is clear that that price differential predated the Trademark Settlements. See, e.g., id. at 46; CX9001 at 021 (Bethers IH 79:23-80:8) (“[W]e were never trying to compete with our online competitors on price. We basically came back and said our online competitors are going to have lower prices than we do. And they did from the day I started with the company [in July 2003]. They were significantly below our retail price.”); CX0535 at 010 (2006 business plan stating that 1-800 Contacts’ “pricing strategy” was to “[p]rice below independent ECPs, close to retail chains, but above our online competitors and Costco”); see also Coon, Tr. 2708:22-2709:9 (noting that “[l]iterally from the beginning”, 1-800 Contacts’ strategy was to price at a discount from ECPs but slightly higher than other online contact lens retailers; that strategy has “never changed”); IDF 434 (“1-800 Contacts on average has retail prices for contact lenses below independent ECPs and retail optical chains, but higher than mass merchants, club stores, and other online retailers.”) (citation omitted).

42 Analytically, categorizing conduct as “inherently suspect” has the same result as holding that direct effects inhere in it. If the Trademark Settlements are inherently suspect, then it is hard to imagine what advertising restrictions would not be inherently suspect.

VOLUME 166 Dissenting Statement Second, Complaint Counsel has put forward no evidence that the price gap increased as a result of the Trademark Settlements. There is no clear causal connection between the price gap and the Trademark Settlements, especially considering that the gap existed before the Trademark Settlements. And there are at least two innocuous and equally plausible reasons why 1-800 Contacts’ prices are higher, including its superior service43 and customers’ preference for the 1- 800 Contacts brand.44 Both of these were likely facilitated and enhanced by 1-800 Contacts’ ability to earn a return on its brand.45 Without observable direct effects, the majority and Complaint Counsel rely on the claim that prices would have gone down but for the Trademark Settlements. But Complaint Counsel failed to quantify the amount that prices would have gone down in their but-for world. See, e.g., Evans, Tr. 1723:20-1724:3 (Complaint Counsel’s economic expert confirming that he did not quantify the extent to which 1-800 Contacts or any other company’s prices would have gone down in the absence of the Trademark Settlements); see also CX8007 (Athey Expert Report) at 036 (providing no empirical evidence for her conclusions). The law requires more: specifically, actual, sustained, and substantial or significant effects. Without quantification, we cannot know whether the harm meets that test.

The majority also claim that 1-800 Contacts maintained supracompetitive prices. See Op. at 49. But Complaint Counsel did not adduce legally sufficient proof. “[T]o support a claim that a defendant set supracompetitive prices through direct evidence, a plaintiff must often provide an 43 The majority assert that certain evidence counters a finding that the service differential explains the price gap. See Op. at 48. But superior service is just one of the reasons that 1-800 Contacts’ prices may be higher than its competitors’ prices. Regardless of how persuasive one may find the evidence on the service differential, it is insufficient to show that the price gap is the result of supracompetitive pricing. Also, the majority’s reliance on competitor testimony claiming that they “offer comparable service to 1-800 Contacts” is remarkable. See id. What competitor is going to get on the stand and testify under oath that its service is inferior? 44 See, e.g., McCarthy 5th Edition, supra, § 2:5 (noting that neither brand preference nor paying a premium for branded products is irrational); RX0739 (Murphy Expert Report) at 081 (“Economists studying price dispersion have shown that a variety of characteristics beyond access to information, such as consumer trust, retailer brand, market and category characteristics, can play an important role in explaining price dispersion.”) (footnote omitted); Borden, Inc., Proposed Order Modification with Statement to Aid Public Comment, 48 Fed. Reg. 9023, 9025 (proposed Mar. 3, 1983) (to be codified at 16 C.F.R. pt.13) (noting consumers’ willingness to pay a price premium as the result of a company’s “familiar and successfully advertised trademark”, which “reflected a marketplace judgment about interbrand competition, which ‘is the primary concern of antitrust law.’”) (quoting Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 51-52 n.19 (1977)); Complaint at 7, In re J.M. Smucker Co. & Conagra Brands, Inc., Dkt. No. 9381 (F.T.C. Mar. 5, 2018) (“Differences in shelf prices for branded and private label CV [i.e., canola and vegetable] oils reflect end consumers’ perception of meaningful product differentiation between branded and private label CV oils. End consumers who buy branded CV oils generally pay a significantly higher price for a branded CV oil than for a private label CV oil.”). 45 As the Supreme Court has noted:

Many decisions a manufacturer makes and carries out through concerted action can lead to higher prices. A manufacturer might, for example, contract with different suppliers to obtain better inputs that improve product quality. Or it might hire an advertising agency to promote awareness of its goods. Yet no one would think these actions violate the Sherman Act because they lead to higher prices. The antitrust laws do not require manufacturers to produce generic goods that consumers do not know about or want. The manufacturer strives to improve its product quality or to promote its brand because it believes this conduct will lead to increased demand despite higher prices. Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 896-97 (2007). 349 1-800 CONTACTS, INC. Dissenting Statement analysis of the defendant’s costs, showing both that the defendant had an ‘abnormally high pricecost margin’ and that the defendant ‘restricted output.’” Mylan Pharm. Inc. v. Warner Chilcott Pub. Ltd. Co., 838 F.3d 421, 434 (3d Cir. 2016) (quoting Geneva Pharm. Tech. Corp. v. Barr Labs. Inc., 386 F.3d 485, 500 (2d Cir. 2004)). As for the second prong of that test, Complaint Counsel does not contend that the Trademark Settlements reduced output. See ID at 153 n.36. Returning to the first prong, the majority do not even attempt to show that 1-800 Contacts’ price-cost margin was abnormally high—either before or after the Trademark Settlements.46 Instead, they rely on inferences and arguments unsupported by proven facts to show that 1-800 Contacts charged supracompetitive prices. As an initial matter, it is obvious that Complaint Counsel failed to meet their burden here because they did not proffer any evidence on margins.47 The only evidence in the record regarding 1-800 Contacts’ margins was proffered by 1-800 Contacts, and that evidence showed that 1-800 Contacts’ margins {remained constant} from 2003 to 2016 despite the Trademark Settlements. RX0739 (Murphy Expert Report) at 064, 107. The majority claim that “{constant} margins do not necessarily mean prices did not rise; without competitive pressures, costs may have risen as prices increased, {keeping margins constant}.” Op. at 49 (italicized emphasis added). This argument substitutes conjecture for actual evidence by providing one possible theory for {constant margins}. It is more likely that 1-800 Contacts’ {constant} margins were not affected by the Trademark Settlements. See, e.g., RX0739 (Murphy Expert Report) at 064 (stating that 1-800 Contacts’ margins have been “{effectively constant} over time” and did not increase as a result of the Trademark Settlements, which “tells us that the settlements {did not raise 1-800’s prices . . . relative to its costs}.”). Indeed, the founder of 1-800 Contacts testified that the company has had the same pricing and margin strategy since 1992. See CX9035 at 023 (Coon Dep. 86:15-87:14). Regardless of which explanation is more plausible, it is Complaint Counsel’s burden to prove direct effects, and they have provided no evidence on the topic of margins.

In an effort to show that 1-800 Contacts’ {constant} profit margins could be explained by 1-800 Contacts’ pre-Trademark Settlement supracompetitive prices, the majority attempt to put forward indirect evidence of market power. See Op. at 49. They claim that—because it was “the incumbent online seller” and had a large share of online sales—1-800 Contacts had market power, which allowed it to charge supracompetitive prices prior to the Trademark Settlements. See id. at 49. This argument fails as a matter of law. First, this is not an argument based on direct evidence of anticompetitive effects; rather, it is an attempt to shoehorn an indirect showing of market power into a direct effects analysis. See Geneva Pharm., 386 F.3d at 500 (“[The] plaintiffs’ assertion with regard to [the defendant]’s continuing high percentage market share is not direct evidence, but rather requires that we engage in the sort of inference more appropriate for market share analysis.”). Second, an indirect showing of market power based on market 46 The majority also fail to show that 1-800 Contacts’ margins would have been lower but for the Trademark Settlements; indeed, the record is devoid of evidence of counterfactual margins. 47 Given that Complaint Counsel bears the burden of proof to show direct effects, it is odd for the majority to argue that 1-800 Contacts somehow calculated its margins incorrectly without requiring any affirmative evidence from Complaint Counsel or any critique of 1-800 Contacts’ margin calculation itself. See Op. at 49. It appears that the majority shift the burden to disprove direct effects to 1-800 Contacts while relieving Complaint Counsel of its burden entirely.

VOLUME 166 Dissenting Statement shares requires a properly defined market, which is absent here. See AREEDA & HOVENKAMP, supra, ¶ 531 (“Market definition is the initial step in assessing a market’s structure.”) (footnote omitted); see also id. ¶ 532(a) (“Identifying a market and computing market shares provide an indirect means for estimating market power.”). Without a properly defined market, showing that 1-800 Contacts had market power based on its share of online sales is impossible. And without a showing of market power, the inference that 1-800 Contacts could have been charging supracompetitive prices also fails. As a result, it is equally (if not more) plausible that 1-800 Contacts’ {constant} margin is consistent with a finding that the Trademark Settlements had {no effect} on 1-800 Contacts’ margins, rather than supracompetitive prices as the majority claim. Cf. Op. at 49.

The majority also claim that “[p]roof of an anticompetitive effect does not require an econometric model to estimate a precise competitive price in order to establish that the existing price is supracompetitive.” 48 Id. While we may not need a “precise competitive price”, we do need evidence of substantial (or significant) anticompetitive harm to find that Complaint Counsel met its burden to show actual, sustained, and significant or substantial direct effects, especially in the presence of real efficiencies that would weigh against any such harm. If the econometrics are insufficient to quantify harm, there is always the option of showing market power indirectly; but the majority opt not to perform that analysis here. See Section 0(D), infra. Finally, the majority argue that the 1-800 Contacts price match policy provides evidence that the Trademark Settlements “had actual price effects”. Op. at 47. But the presence of a price match policy does not prove direct effects; it is equally consistent with a desire by 1-800 Contacts to price discriminate among its customers. And the mere existence of the policy itself signals to customers that they can buy their contact lenses from other suppliers at potentially lower prices.

C. The Majority Inappropriately Discount 1-800 Contacts’ Procompetitive Justifications for the Trademark Settlements.

Given that Complaint Counsel did not meet their initial burden under the inherently suspect framework or by showing direct effects (and because the majority opt not to attempt an indirect showing of market power), 1-800 Contacts need not put forward procompetitive justifications. Nevertheless, the majority fail to give appropriate credit to 1-800 Contacts’ proffered procompetitive justifications.

In their preliminary analysis of 1-800 Contacts’ procompetitive justifications, the majority recognize that the avoidance of litigation costs through settlement is a “legitimate” justification that is “cognizable and, at least, facially plausible”. Op. at 23. The majority also concede that avoidance of litigation costs is a well-recognized procompetitive justification. See 48 This claim contrasts markedly with the majority’s defense of the model put forward by Complaint Counsel’s expert to support the alleged advertising restrictions: “The opinions of Complaint Counsel’s experts derive from the facts in the record and econometric analysis of those facts. The experts use known facts to quantify the impact of the advertising restrictions on the ads that would otherwise appear and on the consumer responses—including clicks and purchases—thereto. They provide empirical evidence, not economic theory isolated from facts, and the underlying facts are in the record.” Op. at 48.

351 1-800 CONTACTS, INC. Dissenting Statement id. (citing Actavis, 570 U.S. at 153; Schering-Plough I, 136 F.T.C. at 1003; In re Tamoxifen Citrate Antitrust Litig., 466 F.3d 187, 202 (2d Cir. 2006)). Both sides’ experts recognize that settling lawsuits is generally economically efficient. IDF 355 (citing RX0739 (Murphy Expert Rep.) at 053, CX9042 at 050 (Evans Dep. 196:22-24)). “There is no question that settlements provide a number of private and social benefits as opposed to the inveterate and costly effects of litigation.” Schering-Plough II, 402 F.3d at 1075 (citation omitted). Despite this clear precedent and their acknowledgement that avoiding litigation costs is procompetitive, the majority claim that, to be considered “valid”, a respondent must show that any cost reduction achieved through settlement was passed on to customers. See Op. at 36-37. The majority cite no relevant case law for this proposition.49 Economic theory cannot fill the precedential void for the majority’s rule. Capital savings like reductions in litigation costs from settlements do not directly affect marginal costs, so it would be impossible to show that they were passed on directly to customers in the form of price reductions.50 Thus, under the majority’s analysis, savings resulting from settlements are “legitimate”, “cognizable”, and “facially plausible”, but could never be “valid”. See Op. at 23, 36-37. That cannot be the rule. The FTC and Supreme Court in Actavis recognized that the litigation costs saved through a settlement could be an “offsetting or redeeming virtue[ ]”. Actavis, 570 U.S. at 156. The Court explained that “[w]here a reverse payment reflects traditional settlement considerations, such as avoided litigation costs or fair value for services, there is not the same concern that a patentee is using its monopoly profits to avoid the risk of patent invalidation or a finding of noninfringement.” Id. In other words, the Supreme Court considered avoided litigation costs as a procompetitive justification. Id. Nowhere did the Court require a showing that savings be passed on to customers in order to be “valid”.

Regardless, the Trademark Settlements had the added benefit of protecting the settling parties’ intellectual property rights. As discussed above (see Section 0(A)(4), supra), trademarks promote interbrand competition, which the Supreme Court has identified as “the primary concern of antitrust law”. Bus. Elecs. Corp. v. Sharp Elecs. Corp., 485 U.S. 717, 724 (1988) (quoting Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 51-52 n.19 (1977)). The ability to 49 None of the cases cited by the majority for the proposition that settlement-related saved litigation costs must be passed through to consumers in order to be “valid” involved a settlement of any kind. See Op. at 37 (citing Chicago Prof’l Sports LP v. Natl Basketball Assn, 961 F.2d 667, 674 (7th Cir. 1992) (challenge to the NBA’s rule that certain television channels could not carry more than 20 games per season), cert. denied, 506 U.S. 954 (1992); Polygram I, 136 F.T.C. at 345 (challenge to joint venture agreement between competitors not to discount or advertise); FTC v. Penn State Hershey Med. Ctr., 838 F.3d 327, 350 (3d Cir. 2016) (challenge to a proposed merger between competing hospitals); Law v. Natl Collegiate Athletic Assn, 134 F.3d 1010, 1023 (10th Cir. 1998) (challenge to an NCAA rule limiting coaches’ compensation), cert. denied, 525 U.S. 822 (1998)). As a result, I do not find any of those cases as persuasive or as directly applicable to the present case as Actavis. 50 See, e.g., Dennis W. Carlton, Does Antitrust Need to be Modernized?, 21 J. ECON. PERSP. 155, 157 (2007) (“Under a consumer surplus standard, only the saving in marginal costs will carry weight because it will reduce prices, while the fixed-cost savings is not considered as a benefit to consumers. . . . Gains that lead to lower fixed costs today can encourage research and development, new products, and plants in the future. However, by focusing only on efficiencies that influence price over a short period, a government antitrust agency risks failing to credit the future efficiencies that will benefit consumers in the long run. To put it another way, the fixed-cost savings of today are the variable-cost savings in the future for new products.”). VOLUME 166 Dissenting Statement enforce and settle claims for infringement of those rights is essential to achieving their purpose. Thus, a reduction in—or elimination of—litigation costs as the result of a settlement is not just legitimate, it is also a valid procompetitive justification even without a showing that the specific reduction in litigation costs was passed on to consumers.

The majority’s only rebuttal to 1-800 Contacts’ argument that the trademark protections provided in the Trademark Settlements are procompetitive justifications is that 1-800 Contacts’ trademark infringement claims were weak. See Op. at 37-41. Evaluating the merits of the underlying infringement claims is inappropriate for the reasons explained above. See Section 0(A)(4)(a), supra. The majority’s concern about the merits of 1-800 Contacts’ infringement claims causes them to miss the forest for the trees. Nowhere in their evaluation of the trademarkrelated procompetitive benefits of the Trademark Settlements do the majority recognize how trademark protections and the vigorous enforcement of trademarks encourage brand investment and promote competition. In fact, the majority dismiss the benefits of trademark policy entirely. This is inappropriate as a matter of law and ignores the facts of this case, including the tremendous amount of investment 1-800 Contacts has made in building its brand, lowering the price of contact lenses, and offering customers superior service. It also raises the question of what the majority’s rule would mean for infringement claims they view as strong. D. The Majority Forego an Indirect Showing of Market Power. Because I do not believe that the majority have shown that the challenged conduct is inherently suspect or that Complaint Counsel have met their burden to show substantial direct anticompetitive effects, the only way for Complaint Counsel to meet its initial burden is through an indirect showing of market power.51 But the majority opt not to take that route here, instead relying exclusively on their claim that the Trademark Settlements are inherently suspect or caused direct anticompetitive effects. Even though the majority do not establish a relevant market, assumptions about the market permeate their opinion, providing ballast to a number of their premises. Without a properly defined product market, each of these arguments fails. For example, in their section on direct effects, the only support that the majority put forward for their claim that 1-800 Contacts charged supracompetitive prices prior to the Trademark Settlements was that “1-800 Contacts was the incumbent online seller, with a dominant share of online sales throughout this period.” Op. at 49 (citations omitted). For the reasons discussed above, see Section 0(B)(2), supra, any attempt to show that 1-800 Contacts charged supracompetitive prices as the direct result of its “share of online sales” requires a properly defined relevant market in which market power can be inferred from a high share. See AREEDA & HOVENKAMP, supra, ¶¶ 531-532. In other words, market definition is a prerequisite to inferring that 1-800 Contacts charged supracompetitive prices from its share of the market. 51 See, e.g., Realcomp I, 2007 WL 6936319, at *31 (stating that—absent a finding that a restraint is inherently suspect—a plaintiff can meet its initial burden “in either of two ways . . . an indirect showing based on a demonstration of defendant’s market power . . . [or] direct evidence of ‘actual, sustained adverse effects on competition’”) (quoting IFD, 476 U.S. at 461; and citing Tops Markets, Inc. v. Quality Markets, Inc., 142 F.3d 90, 96 (2d Cir. 1998); Law, 134 F.3d at 1019; United States v. Brown Univ., 5 F.3d 658, 668 (3d Cir. 1993)). 353 1-800 CONTACTS, INC. Dissenting Statement Without a relevant market, any claim that 1-800 Contacts had market power based on its “share of online sales” and, therefore, charged supracompetitive prices is unsupportable. The majority also “find that the agreements harm consumers and competition for the online sale of contact lenses.” Op. at 2 (emphasis added). It is impossible for the Trademark Settlements to harm competition in a limited line of commerce like “online sales” without a showing that such a limitation is appropriate. In other words, by failing to prove that “the online sale of contact lenses” is a properly defined antitrust market, the majority cannot claim that customers or competition in that market were harmed. Elsewhere, the majority use similar claims that 1-800 Contacts had a large share of “online sales” to imply that 1-800 Contacts was somehow a dominant seller of contact lenses online. See, e.g., id. at 4 (“In 2015, 1-800 Contacts accounted for approximately 54 percent of online sales, which is more than four times the sales of the second-largest online retailer.”) (citations omitted). However, because the majority opt not to define a relevant market, their attempts to show 1-800 Contacts was a dominant online seller—or even that they had a large share of contact lens sales—necessarily fail, as do any implications the majority would like to draw from those attempts. The majority similarly assert that the Trademark Settlements are problematic because they cover a large number of online contact lens retailers that make up a large percentage of online contact lens sales. See id. at 33 (“Challenged Agreements covered 14 different online contact-lens retailers that account for 79 percent of online contact lenses in the United States. . . . [T]he challenged agreements here cover the landscape of online contact-lens retailers resulting in harm to competition overall.”) (citations omitted). Because it relies on an indirect showing of market power, the majority’s conclusion that the Trademark Settlements caused “harm to competition” requires proof of a relevant antitrust market comprised of the online sale of contact lenses in the United States. Absent a proper showing such a market exists, statements like these are irrelevant to the antitrust analysis and do not support the majority’s assertion that the Trademark Settlements harmed competition.

Not only do these assumptions about the market support key aspects of the majority’s analysis, while lacking support themselves, they elide difficult questions about the market in this case. Significant participants in the online sales of contact lenses were not party to the Trademark Settlements, and the record reflects that customers purchased the majority (83%) of their contact lenses from other kinds of retailers, including independent ECPs, optical retail chains, mass merchants, and club stores. See IDF 491. Some were more expensive; some cheaper. Competition from these other retailers cannot be ignored, especially without a properly defined relevant market.

E. The Majority Have Not Shown That the Trademark Settlements Have Anticompetitive Effects for Search Engines.

The majority also would condemn the Trademark Settlements as unlawful because of their effects on firms owning search engines, such as Google (the search engine owned by VOLUME 166 Dissenting Statement Alphabet, Inc.) and Bing (owned by Microsoft Corp.).52 This legal theory is novel; none of the cases cited by the majority as involving advertising restrictions (e.g., California Dental and Polygram) considered such harm. If the theory is novel, the evidence that search engines have been harmed is all but absent.53 Microsoft {explicitly disclaimed any harm from the Trademark Settlements} and its {Principal Scientist} testified that the company is “{not aware of any specific harm to Microsoft or its users from the agreements that the Federal Trade Commission has challenged in this matter}.” RX0704 at 007. As to Google—the largest seller of paid search advertising (see Stip. at 5)—I am concerned this theory of liability fails adequately to take into account record evidence about the putative victim’s role in the alleged harm. As noted above, until 2004 Google itself banned as a matter of company policy the same conduct later barred by the Trademark Settlements (i.e., permitting advertisements for third parties to appear in response to searches for trademarked keywords). See Section 0(0), supra. When it changed its policy, Google assisted trademark owners, including 1-800 Contacts, to address the threat to their marks, advising them specifically that negative keywords were an effective tool to prevent or limit the opportunities for trademark infringement. See id. There is some irony, then, in claiming that Google was harmed. At the very least, the fact that Google once required and, later, affirmatively encouraged the allegedly anticompetitive conduct suggests the Trademark Settlements do not harm Google, a sophisticated and aggressively competitive seller of search-based online advertising, in any material way. In their analysis, the majority apply the rule of reason to consider the harm to search engines,54 finding direct evidence of decreases in (1) search engine advertising revenue; and (2) 52 It is odd for the Commission to address this issue at all. Judge Chappell did not analyze the effect of the Trademark Settlements on search engines and Complaint Counsel did not appeal this portion of the Initial Decision. See Op. at 50 & n.52.

53 I disagree with the conclusion the majority reach on the facts here, as explained in the text, but note that condemning actual bid rigging is a critical component of any robust antitrust regime. The Commission has a dual mission to protect consumers and to promote competition. What We Do, FED. TRADE COMM’N, https://www.ftc.gov/about-ftc/what-we-do. Promoting competition requires effective enforcement of the antitrust laws regardless of the identity of the harmed customer. See, e.g., FTC v. Staples, Inc., 190 F. Supp. 3d 100, 117-22 (D.D.C. 2016) (enjoining merger between the two largest office supply companies in the country because of the potential harm to large businesses, including some of “the most powerful companies in the world”) (citation omitted); id. at 126 (“Antitrust laws exist to protect competition, not a particular set of consumers”); FTC v. Sysco Corp., 113 F. Supp. 3d 1 (D.D.C. 2015) (enjoining merger between the two largest broadline foodservice distribution companies in the country primarily based on potential harm to businesses with a nationwide or multiregional footprint); see also FTC v. Penn State Hershey Med. Ctr., 838 F.3d 327, 338 (3d Cir. 2016) (reversing the district court’s denial of a preliminary injunction because the merger was likely to harm competition in the market for “general acute care (‘GAC’) services sold to commercial payors [i.e., insurers]”); FTC v. Advocate Health Care Network, 841 F.3d 460 (7th Cir. 2016) (same). And “there is near universal agreement that restrictive agreements among competitors, such as horizontal price-fixing (including bid-rigging) . . . can cause serious economic harm.” Office of General Counsel, U.S. Sentencing Commission, Primer: Antitrust, at 1 (March 2018), https://www.ussc.gov/sites/default/files/pdf/training/primers/2017_Primer_Antitrust.pdf (footnote omitted). 54 In a footnote, the majority argue that the Trademark Settlements could also be evaluated in terms of their impact upon search engines under an “inherently suspect” framework. See Op. at 50-51 n.54. But the facts of this case do not meet the standard for applying that standard. The Trademark Settlements govern what kind of advertisements can be bought, not the amount of advertisements that a company can buy; and a rudimentary observer might very well conclude such conduct has no effect on search engines. What is more, the majority do not cite sufficient 355 1-800 CONTACTS, INC. Dissenting Statement the number of advertisements displayed, which the majority claim reduced both the total output of advertisements and the quality of the search engines’ product, the search engine results page (“SERP”). See Op. at 50-54. Neither finding is sufficient to show direct effects under the Supreme Court’s standard, recently reiterated in Amex, that “[d]direct evidence of anticompetitive effects would be ‘proof of actual detrimental effects [on competition],’ such as reduced output, increased prices, or decreased quality in the relevant market”. Amex, 138 S. Ct. at 2284 (quoting IFD, 476 U.S. at 460) (other citations omitted); see also United States v. Microsoft Corp., 253 F.3d 34, 51 (D.C. Cir. 2001) (“[A] firm is a monopolist if it can profitably raise prices substantially above the competitive level. Where evidence indicates that a firm has in fact profitably done so, the existence of monopoly power is clear. Because such direct proof is only rarely available, courts more typically examine market structure in search of circumstantial evidence of monopoly power.”) (citations omitted).

The evidence does not support a finding of a direct price effect (here, a reduction in paid search advertising auction prices). The majority do not cite evidence of reductions in advertising budgets or the number of advertisements created or displayed by the contracting parties. Instead, the majority proffer a theory that “a reduction in the number of search-advertising auction participants offering relevant ads reduces the price paid by the auction winners and reduces the revenue for the search engine.” Op. at 51 (footnote omitted). While this might be correct with respect to certain auctions and SERPs involving trademarked keywords, there is no evidence that this is true with respect to the purchases by the parties to the Trademark Settlements generally, including purchases of other paid search advertising, online advertising more broadly, or advertising as a whole.55 But even accepting the specific auction as the relevant denominator, the economic evidence or judicial experience that would justify the application of a truncated rule of reason analysis. See id. While bid rigging has indeed been condemned as violating the antitrust laws, the Trademark Settlements are categorically different from the types of conduct that the FTC and DOJ consider per se illegal bid rigging. See Price Fixing, Bid Rigging, and Market Allocation Schemes: What They Are and What to Look For, An Antitrust Primer, U.S. DEP’T OF JUSTICE, https://www.justice.gov/atr/price-fixing-bid-rigging-and-market-allocation-schemes [hereinafter “DOJ Antitrust Primer”]; Bid Rigging, FED. TRADE COMM’N, https://www.ftc.gov/tipsadvice/competition-guidance/guide-antitrust-laws/dealings-competitors/bid-rigging. They lack what almost all forms of bid rigging have in common: “an agreement among some or all of the bidders which predetermines the winning bidder”. DOJ Antitrust Primer, supra, at 3. Nothing in the Trademark Settlements predetermined the winner of any auction. The Trademark Settlements also are not akin to per se illegal bid rigging because they were not intended to (and did not always) decrease auction prices, which happened (if at all) only incidentally as the result of the search engines’ use of an auction algorithm. Cf. Compact v. Metro. Govt of Nashville & Davidson Cty., Tenn., 594 F. Supp. 1567, 1575-76 (M.D. Tenn. 1984) (agreement to “fix the price of minority architect participation on public contracts” with the intent and “admitted purpose[ ]” of “eliminat[ing] competitive bidding between its members”). As a result, the Trademark Settlements do not bear the “close family resemblance” to classic bid rigging or rotation sufficient to apply “inherently suspect” analysis. See Polygram II, 416 F.3d at 36-37; see also United States v. Heffernan, 43 F.3d 1144, 1146-47 (7th Cir. 1994) (interpreting “bid rigging” as meaning “bid rotation”, the latter of which “eliminate[s] all competition rather than just price competition”) (citation omitted). The majority also suggest that Complaint Counsel’s initial burden under the inherently suspect and direct effects standards “rely on the same evidence”. Op. at 50-51 n.54. Suggesting that both standards utilize precisely the same evidence and failing to explain how the two analytical frameworks differ, I fear, will only exacerbate confusion in the law. As an expert antitrust agency, the Commission has a duty to help clarify the law, and its decisions certainly endeavor not to obfuscate antitrust analysis further.

55 The majority do not articulate what the appropriate scope of an “advertising” market would be. The majority’s analysis at best demonstrates a “direct effect” in the number of advertisements displayed in response to searches for VOLUME 166 Dissenting Statement record shows that it is not always true that if fewer advertisers participate in an auction that the price paid by the auction winner goes down. See, e.g., IDF 219 (“Under the second price auction used by Google, the number of bidders may or may not affect the actual [cost-per-click].”); Juda, Tr. 1205:5-10 (Google executive testifying that “[i]t is not always the case that more advertisers results in higher [cost-per-click]”); CX9019 at 015, 036 (Juda Dep. 55:9-13, 137:18-138:22) (Google executive testifying that, in certain circumstances, an “increase in [the number of] bidders would have zero influence on the price that that highest person was paying”, and that an additional bidder may or may not affect the cost-per-click of another advertiser in the auction). As {Rukmini Iyer, the Microsoft Principal Scientist} whom the majority cite for the proposition that reducing the number of search engine auction participants could reduce the prices paid by the auction winners (and thereby reduce search engine revenue), see Op. at 51-52, explained, “{[b]ased on my experience, the fact that a particular agreement reduced the number of bidders on a given keyword or set of keywords, standing alone, would not be sufficient to predict the specific effect of that agreement on the results of bidding on those keywords or the specific effect of that agreement on users or Microsoft.}” RX0704 at 006. The majority also claim price effects on the theory that—because advertisements limited by the Trademark Settlements had a higher return on investment (“ROI”)—advertisers would spend less in the absence of their availability. Op. at 53-54. That is a plausible assumption. But, especially given how important online advertising apparently was to the contracting parties, see id. at 6-7, 30-31, it is equally plausible they would have bought other advertisements, with no harm going to the owners of the search engines.

The majority’s ROI theory also discounts the value of advertising purchased for brandbuilding (as opposed to only for sales) purposes. If advertisers viewed online search advertising as a branding opportunity, removing certain keywords from the available pool would most likely shift advertising purchases to other keywords, because brand building is more about appearing frequently than achieving a set ROI with each appearance. The record is replete with evidence that advertisers evaluated online search advertisements on a brand-building basis (in addition to ROI). One witness explained that his company built its brand “primarily through the online search advertising.” CX9024 at 011 (Holbrook Dep. at 40:4-7); see also {CX9014} at 048 ({Batushansky Dep.} at {186:3-7}) ({agreeing that paid search advertising raises Web Eye Care’s brand recognition}); IDF 602 (“LensDirect believes there is value in showing an ad in the trademarked terms covered by the Trademark Settlements. Such a market seems implausible. Courts have rejected “search engine advertising” as a viable antitrust market because it is too narrow, but even that is far broader than the handful of trademarked keywords within search engine advertising at issue here. See, e.g., Lasoff v. Amazon.com Inc., No. C16-151, 2017 WL 372948, at *9 (W.D. Wash. Jan. 26, 2017) (“Because there is no basis for distinguishing the ‘search engine advertising’ market from the larger market of all internet advertising, the former is simply too narrow to form a meaningful ‘relevant market’ for purposes of antitrust liability.”) (quoting Person v. Google, Inc., No. C06-7297, 2007 WL 832941, at *4 (N.D. Cal. Mar. 16, 2007) (“The Court finds no basis for distinguishing the Search Ad Market from the larger market for Internet advertising. Search-based advertising is reasonably interchangeable with other forms of Internet advertising.”)); see also Statement of Commissioner Ohlhausen, Commissioner Wright, and Commissioner McSweeny Concerning Zillow, Inc. / Trulia, Inc., FTC File No. 141-0214 (Feb. 19, 2015), https://www.ftc.gov/system/files/documents/public_statements/625671/150219zillow mko-jdw-tmstmt.pdf.

357 1-800 CONTACTS, INC. Dissenting Statement response to a search for 1-800 Contacts, even if the ad is not clicked on, because it gives LensDirect brand visibility next to the larger players without any cost.”) (citation omitted); ID at 144 (“As LensDirect’s chief executive officer stated: ‘[T]he more times people see LensDirect, the better chance there is of them becoming a customer one day.’”) (citation omitted). While removing certain keywords from the available pool would most likely shift advertisement purchases to other keywords, the necessity of brand building gives additional reason to assume the money would continue to go to online search advertising, even with a lower ROI. Even if there were a reduction in advertising in response to searches for trademarked terms (which has not been proven), it is unclear that a reduction in the number of advertisements would negatively affect the quality of the search engine experience. As Complaint Counsel’s expert testified, there is significant literature explaining that search engines, as multi-sided platforms, must balance the advertisers’ desire to appear more frequently in SERPs and consumers’ desire to be bombarded with fewer ads. CX8006 (Evans Expert Report) at 024-025. Purchased advertisements are how the search engines monetize their platforms; whereas organic results are where the search engines place the links they deem most relevant to consumers. Consistent with this notion, {Microsoft’s Principal Scientist} testified that: {In general and within reason, the more an advertiser bids, the less relevant its advertisement needs to be to be displayed on the search results page. Accordingly, a paid search advertisement for a website may be sufficiently relevant to appear on the first page of search results for a user query even though an organic search result for the same website would not be sufficiently relevant to appear on the first page—or even the first several pages—of search results for the same user query.} RX0704 at 003. Thus, from the search engines’ perspective, it is not clear how the quality of the advertisements are lessened.56 F. The Trademark Settlements Were Appropriately Tailored. The majority rest their liability theory, in part, on the claim that the Trademark Settlements could have been narrower. See Op. at 25-30. This substitutes the Commission’s judgment for that of the parties, contrary to what Clorox requires. See Clorox, 117 F.3d at 60. 56 The majority assert that the Trademark Settlements prevented some consumers from clicking on advertisements that did not appear because of the agreements, presumably generating less value for the search engine. But it is not clear from the search engines’ perspective (i.e., the theory of harm at issue here) why a consumer searching for “1-800 Contacts” is less likely to click through under the Trademark Settlements. They might be faced with a more obviously responsive advertisement (e.g., one for 1-800 Contacts), and thus more likely to click through on that advertisement than on an advertisement for another vendor. Indeed, record evidence indicates that most searches for the trademarked terms at issue were, in fact, navigational—that is, consumers typed in “1-800 contacts” because they wanted to reach 1-800 Contacts’ website. RX0733 (Ghose Expert Report) at 007 (“[C]onsumers who searched for 1-800 Contacts’ trademarks typically did so with a navigational intent.”); id. at 060 (“[T]he academic literature and the data [ ] indicate that the vast majority of consumers searching for 1-800 Contacts’ trademark do so with navigational intent.”).

VOLUME 166 Dissenting Statement But the Trademark Settlements also were appropriately tailored to achieve their objective. The searches that the Trademark Settlements prohibit are precisely those searches that implicate 1- 800 Contacts’ trademarks. They are also the searches through which users are most likely attempting to reach the 1-800 Contacts website (i.e., searches for 1-800 Contacts’ trademark). See, e.g., RX0733 (Ghose Expert Report) at 060 (“[T]he academic literature and the data [ ] indicate that the vast majority of consumers searching for 1-800 Contacts’ trademark do so with navigational intent.”). Indeed, 1-800 Contacts considered navigational searches (i.e., paid searches for its trademarks) as “direct traffic” to its website (as opposed to indirect traffic). IDF 577. As a result, the settling parties structured the Trademark Settlements to prevent advertisements from appearing in response to searches for both parties’ trademarks. The settling parties included a negative keyword provision in response to Google’s explicit encouragement for 1-800 Contacts to resolve its trademark disputes with competitors by having them implement 1-800 Contacts’ trademarked terms as negative keywords. See, e.g., Schmidt, Tr. 2904:2-16, 2905:16-25; CX9031 at 010-011 (C. Schmidt Dep. 33:20-35:2, 35:23- 36:2, 36:13-37:3); CX9013 at 044 (Aston Dep. 172:1-3) (“They [Google] instructed us [1-800 Contacts] to have the offenders add those specific trademarked terms into their negatives for their -- for their AdWords campaigns.”); id. at 044-045 (170:8-20, 171:10-19, 173:5-20). They did so because, without negative keywords, a settling party’s advertisements could appear in response to searches for the counterparty’s trademarked terms. Almost all of the Trademark Settlements balanced these restrictions with a provision explicitly permitting a settling party to use the counterparty’s trademarks in a manner that would not constitute infringement in the non-internet context, including comparative advertising. IDF 369 (“Ten of the thirteen Settlement Agreements provide that the prohibited acts ‘shall not include (i) use of the other Party’s trademarks on the Internet in a manner that would not constitute an infringing use in an non-Internet context, e.g., the use on the Internet of comparative advertising, parodies, and similar non-Infringing, uses.’”) (citations omitted); see also IDF 305 (finding that 1-800 Contacts accepted changes to a draft settlement agreement with Vision Direct and stated that both parties should be able to engage in comparative advertising); IDF 309 (confirming that the 2004 Trademark Settlement between 1-800 Contacts and Vision Direct permitted non-infringing uses, such as comparative advertising, parodies, etc.). As a result, in my view, the Trademark Settlements were appropriately tailored to achieve their goal of preventing trademark infringement while balancing the need to permit noninfringing advertising.

III. The Majority Fail to Analyze the Luxottica Sourcing and Services Agreement. The majority do not analyze the sourcing and services agreement between Luxottica and 1-800 Contacts (the “Luxottica Agreement”) correctly. Sourcing and services agreements, like trademark settlement agreements, are typically considered procompetitive. See Fed. Trade Commu & U.S. Dept of Justice, Antitrust Guidelines for Collaborations Among Competitors, at 1 (Apr. 2000) [hereinafter “Competitor Collaboration Guidelines”]. As a result, courts typically analyze ancillary restraints accompanying sourcing and services agreements under the rule of reason. See AREEDA & HOVENKAMP, supra, ¶ 1908(c). The majority, however, treat the sourcing 359 1-800 CONTACTS, INC. Dissenting Statement and services agreement between Luxottica and 1-800 Contacts as “inherently suspect” by lumping it in with the Trademark Settlements. See Op. at 10. The only time the majority discuss the Luxottica Agreement is to note that certain procompetitive justifications that 1-800 Contacts proffered for the Trademark Settlements do not apply to the Luxottica Agreement. See, e.g., id. at 12 n.14, 37. By ignoring its plain language and considering the Luxottica Agreement to be just another Trademark Settlement, the majority lay bare the broad scope of the rule they announce and fail to address additional procompetitive justifications that typically accompany supply and sourcing agreements.57 A. The Luxottica Sourcing and Services Agreement Is a Supply Agreement, Not a Trademark Settlement.

In December 2013, 1-800 Contacts entered a sourcing and services agreement with Luxottica. IDF 393. Luxottica operates chains of brick-and-mortar retail stores—such as LensCrafters, Pearle Vision, Sears Optical, and Target Optical—that sell, among other things, contact lenses. IDF 394. The Luxottica Agreement did not end any alleged trademark infringement; instead, it provides for a mutually beneficial vertical relationship between 1-800 Contacts and Luxottica. See CX0331. In particular, under the Luxottica Agreement, 1-800 Contacts provides (1) fulfillment services by shipping contact lenses to Luxottica’s retail chain stores and (2) other services, including assistance with sourcing contact lenses from the four major contact lens manufacturers. IDF 394.

Judge Chappell made explicit the benefit of the Luxottica Agreement to 1-800 Contacts: “As a result of the agreement between 1-800 Contacts and Luxottica, 1-800 Contacts is {able to obtain better wholesale pricing on contact lenses}.” IDF 395 (citations omitted). But there were also benefits to Luxottica. In particular, 1-800 Contacts managed and operated Luxottica’s contact lens business. See CX0331 at 006. In effect, Luxottica outsourced its entire contact lens business, including negotiating with contact lens suppliers, to 1-800 Contacts.58 See id. at 025 (“LUX shall use 1-800 exclusively to source Trial Lenses and Revenue Product in the Territory.”) (emphasis added); id. at 014 (defining “Revenue Products” as “contact lenses for retail sale”); id. at 018 (defining “Territory” as “the United States of America, its territories, and Canada”); id. at 026 (“1-800 shall lead all negotiations with Suppliers”). 57 The majority defend their approach by stating that Respondent did not carry its burden of establishing the procompetitive nature of the Luxottica Agreement. See Op. at 37 n.38. That does not justify ignoring the plain terms of the Luxottica Agreement, which clearly is not a settlement of any kind, much less one of the Trademark Settlements analyzed here.

58 Under the Luxottica Agreement, 1-800 Contacts maintained the inventory of contact lenses and shipped them directly to Luxottica’s retail chain stores and directly to the homes of customers of Luxottica’s retail stores. CX0331 at 029; Bethers, Tr. 3524:5-3525:6, 3694:14-3695:14. The packaging of all contact lenses shipped by 1-800 Contacts under the agreement bore Luxottica’s labels and in no way indicated that 1-800 Contacts was involved. CX0331 at 029; Bethers, Tr. 3525:7-21, 3694:14-3695:14.

VOLUME 166 Dissenting Statement B. The Majority Ignore Procompetitive Justifications for the Luxottica Agreement.

The majority assert that certain justifications for the Trademark Settlements do not apply to the Luxottica Agreement, see Op. at 12 n.14, 37, but they simultaneously ignore procompetitive justifications for sourcing and services agreements. The Commission enumerated some of those justifications in guidelines jointly published with the U.S. Department of Justice. See Competitor Collaboration Guidelines. The Luxottica Agreement falls squarely within the agencies’ definition of “competitor collaborations.” Id. § 1.1 (“Competitor collaborations involve one or more business activities, such as research and development (‘R&D’), production, marketing, distribution, sales or purchasing.”) (emphasis added). The Guidelines recognize the Commission’s view that agreements among competitors (or potential competitors)59 can benefit customers in a variety of ways. See id. § 2.1. Among the many consumer benefits that could result from the Luxottica Agreement is the fact that 1-800 Contacts has the largest inventory of contact lenses in the industry, see IDF 44, and therefore may have a comparative advantage over Luxottica in negotiating with suppliers and delivering contact lenses to customers. As a direct result of its decision to outsource much of its contact lens business to 1-800 Contacts, Luxottica customers could receive lower prices and better service (e.g., faster delivery). The majority opinion fails to analyze any of the foregoing (or any other potential) procompetitive justifications for the Luxottica Agreement.60 Instead, they summarily condemn it as part-and-parcel of the Trademark Settlements. Given the seemingly apparent procompetitive justifications, I fear this omission speaks more to the breadth of the conduct the majority condemn.

IV. The Majority’s Remedy The remedy proposed by the majority is ineffective. The Order states that the only agreements that 1-800 Contacts can enter are those that, in effect, tell the counterparty that they cannot violate the trademark laws. See Final Order at 2-3. Such agreements resolve nothing and will only lead to more litigation to determine what conduct actually violates the trademark laws in the context of paid search advertising based on trademarked keywords. Because the Order only allows agreements that do not actually resolve the dispute in trademark infringement litigation, it will reduce the incentive to settle, which, in turn, will lead to either less trademark 59 I note that it is unclear from the majority opinion whether they view Luxottica (a brick-and-mortar retailer) and 1-800 Contacts (an online contact lens retailer) as direct horizontal competitors because the majority fail to define a relevant product market. Nevertheless, the same analysis is appropriate regardless of whether the two companies directly compete. See Competitor Collaboration Guidelines, § 1.1. 60 The majority also fail to analyze the advertising restrictions in the context of the Luxottica Agreement. For example, the restrictions on paid keyword search advertising may have been necessary for the parties to enter into the Luxottica Agreement in the first place. Given the potential procompetitive benefits surrounding competitor collaborations like the Luxottica Agreement, it is likely that any anticompetitive harm caused by the advertising restrictions would be outweighed by the procompetitive benefits of the agreement as a whole. 361 1-800 CONTACTS, INC. Dissenting Statement enforcement or more costly litigation for the same reasons discussed above. See Section 0(A)(4), supra.61 * * * The Commission’s mandate is to enforce the antitrust laws, but we cannot do so in a vacuum. We need to consider competing policies, including federal trademark policy, when 61 In the section discussing the remedy, the majority repeat at least two of their earlier claims that I believe are not supported by the facts, law, or both. First, they claim that they “are not establishing a new trademark rule” and even go so far as to say that they “make no ruling on any trademark issue at all.” Op. at 56. For the reasons discussed more fully above, there is a trademark ruling implicit in the majority’s decision to truncate their rule of reason analysis. See Section 0(A), supra. Second, they assert that the Order is not novel, in part, because “[a]ntitrust has long barred rivals’ agreements regarding advertising and bidding restrictions.” Op. at 56. This does not reflect a fair reading of the case law as applied to the Trademark Settlements, as I discuss above. See Section 0, supra. VOLUME 166 Concurring Statement analyzing allegedly anticompetitive conduct. And we should recognize that unclear rules may do more to harm both to that policy and to competition than the alleged conduct here. In the case of the Trademark Settlements, precedent offers a better way: the Commission should analyze such agreements under the full rule of reason, giving appropriate weight to the trademarks at issue and the value they protect. Such a rule will decrease uncertainty in the market, encourage brand investment, and increase competition.

CONCURRING OPINION OF COMMISSIONER REBECCA KELLY SLAUGHTER I strongly support the Commission’s decision and order. As explained in the Commission’s Opinion, the agreements between 1-800 Contacts and its online rivals to restrict advertising on search engine result pages harmed consumers and competition. I write separately to explain why this case was a worthwhile expenditure of Commission resources. Specifically, this case merited the Commission’s attention because of the importance of competition in online search bidding for both consumers and for competitive entry by online sellers of goods and services. The Commission’s Opinion also addresses Complaint Counsel’s allegation of harm to search engines in the form of depressed prices paid for search advertising. While I agree with the conclusion that the agreements at issue in this case constituted a type of illegal bid rigging, it was important for me to connect that conduct to consumer harm rather than harm to search engines alone.

Complaint Counsel successfully demonstrated that consumers were harmed by the agreements in this case. Those agreements not only deprived consumers of information about alternative sellers of contact lenses, which is sufficient on its own to establish a violation of Section 5, but the evidence shows that consumers paid more for contact lenses as a result of 1- 800 Contacts’ efforts to protect itself from lower-priced competitors. Consumers who searched online lost a critical opportunity to explore these alternative contact lenses sellers or take advantage of 1-800 Contacts’ price match if they found such lower prices. These agreements increased the costs to consumers across the country who need contact lenses to correct their vision.

This case is important to competition and consumers – both because of the specific harm to contact lens purchasers and sellers and because of the precedent it sets as sponsored search results generated by bidding on a competitor’s brand name becomes an increasingly important avenue for businesses to break into online sales markets.1 Online search bidding restriction may be a new frontier in advertising restraints, but it is just as pernicious as traditional restraints in 1 See, e.g., Rani Molla, Amazon is Stuffing Its Search Results Pages With Ads, Recode, (Sept. 10, 2018), available at https://www.recode.net/2018/9/10/17797720/amazon-is-stuffing-its-search-results-pages-with-ads. 363 1-800 CONTACTS, INC. Concurring Statement frustrating the role that advertising plays to benefit consumers in their search for the highest value products and services as recognized by the Supreme Court.2 A competitive marketplace should ensure that consumers get the best prices, choices, quality, and innovation. This case provides a good example of how the Commission should use its resources to attack conduct that robs consumers of competition that results in lower prices, and robs competitors of the ability to challenge a dominant player. The Opinion also holds that the agreements consisted of a form of bid rigging that artificially depressed the price search engines received for online advertising. I agree with the legal conclusion, expressed in the Opinion, but I write separately to note that I would not have supported pursuing this case based on harm to search engines alone. The resources of the Commission are limited, and should generally be used to protect consumers, not large companies with substantial market share. Given the depth and precedential significance of the consumerfacing harm in this case, I support the Opinion and Order. 2 Bates v. State Bar of Arizona, 433 U.S. 350, 364 (1977) (explaining that advertising “serves to inform the public of the availability, nature, and prices of products and services, and thus performs an indispensable role in the allocation of resources in a free enterprise system.”).

VOLUME 166 Complaint