FTC Explains Why It Thinks 1-800 Contacts’ Keyword Ad Settlements Were Anti-Competitive–FTC v. 1-800 Contacts
As you may recall, the FTC is pursuing 1-800 Contacts for antitrust violations based on 1-800 Contacts having sued and then settled with competitors who bought keyword ads on 1-800 Contacts’ trademarks. Recently, the FTC filed its “Complaint Counsel’s Corrected Pre-Trial Brief and Exhibits” in the Matter of 1-800 Contacts, Inc. This 90 page document, unfortunately swiss-cheesed by numerous redactions, lays out the FTC’s case. It’s a fascinating read, and I encourage you to read the whole thing. In this post, I’ll flag some of the highlights.
Tangling with the U.S. Government. The first page lists 14 FTC lawyers as “Counsel Supporting the Complaint,” of which 12 are listed in the signature block. This is what it’s like to bring the full weight of the US government down on a defendant. 1-800 Contacts is no slouch, with 7 attorneys listed in the case, which sounds a lot of lawyers until you compare it to the FTC’s army.
An (Expensive) Battle of the Experts. As further evidence of how expensive this litigation is, both sides have a slew of high-priced experts. 1-800 Contacts has 6 experts: Howard Hogan (Gibson Dunn), Dr. William Landes (University of Chicago Law School), Dr. Anindya Ghose (NYU Stern Business School), Dr. Kent Van Liere (NERA), Dr. Ronald Goodstein (Georgetown McDonough Business School), and Dr. Kevin Murphy (University of Chicago Booth Business School).
The FTC’s experts including Dr. Susan Athey (now Stanford Business School, but in 2009 she was helping Microsoft attack Google on antitrust matters), Dr. David Evans, Dr. Jack Jacoby (NYU Stern Business School, and one of the most experienced trademark survey experts of all time), and my long-time collaborator Rebecca Tushnet (soon to be of Harvard Law School).
Overview of the FTC’s Case. This excerpt from the introduction summarizes the FTC’s topline perspective:
The evidence will show that, through these Bidding Agreements, 1-800 Contacts effectively shut down its significant rivals’ search advertising against 1-800 Contacts’ trademarks, blocking relevant, valuable advertising that would have been displayed to consumers absent these agreements. Further, consumers suffer direct pecuniary harm, because the eliminated advertising would inform consumers that identical products are available at lower prices from 1-800 Contacts’ online competitors. Because the Bidding Agreements block this valuable advertising, consumers receive degraded search results, and at least some consumers have paid more for their contact lenses. 1-800 Contacts’ contemporaneous business documents demonstrate unequivocally that it lost sales, to the benefit of consumers, when its lower-priced rivals placed advertisements blocked by the Bidding Agreements.
The evidence will also show that 1-800 Contacts’ Bidding Agreements have artificially depressed prices in millions of auctions held by the major U.S. search engines for the display of advertising, thus depriving search engines of significant revenues they would otherwise have earned. The Bidding Agreements have also harmed the search engines by degrading the overall quality of the product they offer to consumers.
Consumers Pay for 1-800 Contacts’ Expensive Marketing. According to the FTC:
While 1-800 Contacts prices below traditional ECPs, its price is on average higher than that of other online merchants, often by a substantial amount. Consumers are generally unaware of this, and believe that 1-800 Contacts’ prices are comparable to that of other pure-play online retailers.
Later, the FTC reiterates: “1-800 Contacts is consistently the highest-priced seller on the internet, and consumers do not know it.”
If you are a 1-800 Contacts customer, did you know this? Will it influence your future buying decisions?
The FTC Now Loves Google. Remember all of the teeth-gnashing about whether the FTC would bust Google? Apparently Google is now aces in the FTC’s book:
In effect, users are continuously “voting” (with their clicks) on what is useful to them and what is not, and Google is continuously reacting to those votes, revising its SERP accordingly. Ultimately, Google is able to predict, typically with a high degree of confidence, what SERP will be relevant to any given user based on how other users have behaved in response to similar SERPs constructed in response to similar search queries.
Apparently the FTC Loves Keyword Advertising, Too. If I were Google, I’d be quoting this language in all of my marketing copy:
Search advertising is uniquely valuable to advertisers because it puts an advertisement in front of a consumer at the precise moment the consumer is signaling her interest or intent by telling the search engine what she is seeking: it is literally the right ad, for the right user, at the right time.
Later, the FTC says “search advertising is quite different from other types of advertising, and of unique value to both consumers of contact lenses and to competitors of 1-800 Contacts.”
If the FTC says it, it must be true, right?
The Lens.com Litigation Was a Big Loss for 1-800 Contacts. Experienced litigators know that bringing a lawsuit always has the potential to backfire. In this context, it appears that 1-800 Contacts’ enforcement against Lens.com has become the centerpiece in the FTC’s case against it. The FTC brief cites Lens.com dozens of times, and the 10th Circuit Lens.com opinion remains one of the most important cases against the initial interest confusion doctrine. For example, the FTC says:
in the one case 1-800 Contacts has fully litigated on its view of trademark infringement, it lost decisively, twice. Lens.com is a widely-cited precedent on the limits of trademark infringement liability with regard to keyword bidding
Lost Sales. In 2008, 1-800 Contacts estimated that competitive keyword bidding was causing it to lose over $68k/month in sales.
What Do Consumers Expect from Trademarked Search Queries? I’d love to see the detail about this assertion from the FTC: “As Dr. Evans will testify, when a consumer of contact lenses is presented with only the 1-800 Contacts advertisement, that consumer is more apt to curtail her search and to settle for whatever price is offered by 1-800 Contacts.” This gets to the heart of keyword advertising’s pro-competitive potential. If true, multiple bidders on a trademarked keyword don’t “divert” consumers from the trademark owner, but instead keep consumers from settling for what they mistakenly believe is the only option in the marketplace.
Later, the FTC says “consumers not only understand that searches will bring ads from multiple companies, but have come to expect this variety.” 1-800 Contacts tried to rebut this by citing the Hyman/Franklyn study–curious, because I cite that study for the direct opposite proposition that many consumers want and expect ads from multiple vendors for trademarked search queries. The FTC has several responses to the Hyman/Franklyn study: (1) the survey took place in the abstract, not in the context of actual search results, (2) the survey doesn’t reveal a consumer’s search intent, either at the outset or as it evolves throughout the course of a search, (3) the survey doesn’t model whether the consumers found the information helpful, even if it wasn’t what they initially sought (later, the FTC says “options do not ‘distract’ users from consummating a desired transaction – they help users to do so”), and (4) no matter what, Hyman/Franklyn showed that rival ads created very low levels of confusion. Dr. Jacoby did a report buttressing the latter point.
1-800 Contacts also offered Dr. Van Liere’s study purporting to show that consumers searching for 1-800 Contacts only wanted to find it. I’d love to see the report too. The FTC raised numerous challenges to the report.
Google as Victim. The FTC says consumers were harmed by the reduction in keyword ads, but it also says Google–yes, the company with $90B in 2016 global revenues and at least 64% of the search query market–was a victim of 1-800 Contacts’ scheme:
The Bidding Agreements also directly result in financial harm to the search engines. Dr. Evans constructed a model showing that the Bidding Agreements reduced 1-800 Contacts’ cost-per-click by 49-59 percent. 1-800 Contacts itself estimated in 2008 that competitive auction bidding cost it an additional $20,434 per month in advertising costs. 1-800 Contacts’ contemporaneous documents directly link a reduction in competition to lower advertising costs: “low competition equals low cost.” And a 2009 email between 1-800 Contacts marketing employees explains that the purpose of the trademark enforcement policy was to “remove competitors[,] which in turn drives down how much we pay per click.”
1-800 Contacts’ Bidding Agreements also reduce the quality of the SERP displayed by the search engines. Because they have fewer potentially relevant ads to choose from, the search engines are unable to display information that they believe may be relevant and useful to consumers. Dr. Evans constructed a model demonstrating that, in the absence of the Bidding Agreements, Google would have served more than a hundred million additional ads between January 2010 and June 2015.
In addition to not being able to serve up a large volume of potentially relevant advertising, these artificially-imposed restraints hamper the search engines’ ability to learn by analyzing what users are choosing to click on (or not to click on)….
Oh the irony…the FTC is now *protecting* Google from antitrust violations…!
Why Can’t 1-800 Contacts Enforce Its Trademarks? The FTC (unfortunately) concedes that 1-800 Contacts is an enforceable trademark. Personally, I think we should question the legitimacy of all “1-800 [noun]” trademarks for selling the noun, just like [noun].com is generic for a site related to the noun. Still, despite 1-800 Contact’s putatively valid trademark, the FTC sees limits on its enforcement:
A valid trademark invests the trademark owner with a far more limited right to bar only confusing uses of the trademark. On their face, the Bidding Agreements reach significantly beyond 1-800 Contacts’ property right by (i) barring non-confusing uses of the trademark; (ii) requiring negative keywords; and (iii) providing for reciprocal restraints on competition by 1-800 Contacts.
The Initial Interest Confusion Doctrine (IIC). The IIC doctrine is an anachronistic and analytically deficient doctrine. The FTC should have rejected it. Instead, the FTC unfortunately acquiesced to the doctrine’s existence. The FTC pulls a definition from the Network Automation case:
Initial interest confusion arises where a defendant uses the plaintiff’s trademark in a manner calculated to capture initial consumer attention, and diverts the now-confused consumer to the defendant’s store (or website), “even though no actual sale is finally completed as a result of the confusion.”… however, the key question is whether a consumer is confused as to the source, sponsorship, or affiliation of the defendant’s product or service
I don’t love the first part of this definition. First, “capturing initial consumer attention” is an overbroad phrase. That’s called marketing. Second, “diversion” is a weird concept because it assumes we know where consumers intended to go in the first place, but consumer search objectives are opaque and constantly-evolving. Still, the FTC’s qualifier about source confusion is a powerful limit on the IIC doctrine. With that limit, we actually don’t need the IIC doctrine at all because normal trademark confusion doctrines would apply. As a result, most old-school IIC cases weren’t consistent with the FTC’s qualifier. I’m not sure what IIC means any more, but the FTC’s definition of it may not actually reflect the current state of the law.
Worse, the FTC quotes the insipid Brookfield billboard analogy, even though I spent several pages *in 2005* pointing out why the analogy was stupid. SIGH.
Having set up its definition, the FTC then adds further qualifications:
All cases finding liability under the initial interest confusion doctrine address conduct that misdirects consumers by creating a false association with the trademark holder. By contrast, where a competitor wins sales or diverts customers though comparative advertising, particularly where the ad is “clearly labeled” as to its source, there is no actionable confusion. In other words, a plaintiff can win an initial interest confusion case only by proving that a significant percentage of consumers are confused into thinking that the plaintiff is either the source of the defendant’s goods, or that the defendant’s goods are sponsored by, or affiliated with, the plaintiff. The fact that a consumer has merely been diverted is not actionable.
So what does the FTC think qualifies as IIC? It appears to be as confused by the doctrine as everyone else.
Keyword Advertising Isn’t IIC. The FTC says “Over the past decade, courts have consistently held that bidding for trademarked keywords alone is insufficient to establish a likelihood of confusion, and that an initial interest confusion analysis must take into account the content (i.e., text) and appearance of the ad, as viewed by the user on the SERP….courts have repeatedly affirmed the principle that clear labeling as to source or affiliation greatly eliminates confusion in the context of search advertising.” Citing the Multi-Time Machine v. Amazon case, the FTC says “in assessing the likelihood that search advertising results in initial interest confusion, a court must consider both the use of the keyword and the text, appearance, and context of the resulting advertisement.”
1-800 Contacts’ expert purportedly cited dozens of cases finding trademark infringement in keyword advertising cases. “But, as 1-800 Contacts’ expert conceded at his deposition, almost none of those cases address the legality of keyword bidding standing alone….The unassailable reality is that 1-800 Contacts can
point to ‘no case indicating that the simple purchase of advertising keywords, without more, may constitute initial interest confusion.'” When pushed, the expert cited the Australian Gold and Soilworks cases as keyword-only infringement cases. However, the Australian Gold case involved false claims of authorized product distribution, plus the injunction didn’t reach keyword ads. Re Soilworks, the defendant used the trademark on its website, plus it’s likely the case got overturned by the Ninth Circuit’s Network Automation case.
Free Riding. 1-800 Contacts argues that its keyword ad restrictions prevent competitive “free riding.” For more on why that argument is wrong, see my Brand Spillovers article.
You can track all of the case filings here.
In separate news, the civil antitrust lawsuits against 1-800 Contacts (inspired by the FTC case) are not being consolidated into a multi-district litigation. In re 1-800 Contacts Antitrust Litigation, 2017 WL 1282949 (MDL Apr. 5, 2017)
BONUS: A recent but garbled Canadian ruling on keyword advertising: Vancouver Community College v. Vancouver Career College (Burnaby) Inc., 2017 BCCA 41. Related blog post. From my perspective, the opinion’s highlight: “Merely bidding on words, by itself, is not delivery of a message. What is key is how the defendant has presented itself, and in this the fact of bidding on a keyword is not sufficient to amount to a component of passing off, in my view.”
More Posts About Keyword Advertising