Wednesday, July 27, 2011

Travel company falls into the gap: confusion, but not dilution

The Gap, Inc. v. G.A.P. Adventures Inc., 2011 WL 2946384 (S.D.N.Y.)

Congratulations to Bruce Keller & Debevoise & Plimpton, my former employers, on their victory on behalf of Gap.

You know the Gap brand. G.A.P Adventures Inc. is an Ontario company in the field of travel booking and travel agency services, founded in 1990 and making its first US sales in 1993. G.A.P (the last period is in the caption, but apparently not in the logo) was intended as an acronym for “great adventure people,” which was part of the logo underneath the company name. The company has grown consistently and made over $87 million (Canadian) by 2007, when The Gap sued, and has a significant US presence in advertising.

The court found that, throughout its history, G.A.P was aware that customers associate its name with Gap’s famous marks and are likely to be confused. Evidence for this included G.A.P’s choice of as its website since 1995. Its logo redesign in late 2008 “placed even more prominence on the word ‘gap’ and indicated an intention to increase the association with plaintiffs' name and trademarks.” It used bold font and lower-case letters and eliminated the periods between the G and the A and the A and the P, minimized the word “adventures,” and left out the “great adventure people” tagline. After Gap objected, G.A.P. modified the new logo, restoring the tagline, but “gap” was still the most prominent aspect. G.A.P often did not use the tagline on promotional items and advertising material, frequently calling itself “Gap Adventures.”

In 2010, G.A.P sent an email to its employees, instructing them to use the company’s name in full at all times, employees frequently disregarded this and used G.A.P instead; the company didn’t repeat the memo and had no compliance program to discipline employees for calling it G.A.P.

G.A.P also stocks clothing, such as T-shirts printed with “G.A.P Adventures”; it sold such clothing in 1998 and 2004. Despite testimony that it no longer offered clothing to the general public, an investigator visited the Toronto office in May 2011, “saw T-shirts in various colors in the reception area with G.A.P Adventures' new logo printed on the top back of the shirt, was given a T-shirt, and brought it to the trial.”

The court found actual confusion. A former principal testified that between 1991, and 1998, customers called G.A.P’s Toronto office looking for Gap clothing, and a search of computer records (revealed after discovery closed) covering 18 months in 2001-2002 found 11 potential customers requesting brochures who said they’d been looking for Gap clothes. (The court was not complimentary to G.A.P’s principal or counsel, who didn’t supplement the company’s answers to interrogatories when it found this information.) The court inferred that “a certain number of visitors to its website and readers of its brochures responded to G.A.P Adventures in the belief that there was a connection to plaintiffs.” I’ll omit discussion of other evidence (confusion by nonconsumers) but it was there.

Gap also presented a dilution survey by Gerald Ford covering people over 18 who reported that they were likely, within the next year, to use the Internet to search for information on travel tours outside the continental United States. The court found the more general question appropriate to define the survey population; though G.A.P promotes ecotourism, it offers a wide variety of travel tours to appeal to varying and extensive population segments. In the test cell, respondents saw an abbreviated version of G.A.P’s website with the old logo. In the control cell, respondents saw a site identical except for that Great Adventure People replaced every instance of “G.A.P” and the tagline was eliminated as redundant.

Respondents were asked "What company or brand, if any, comes to mind when you see the name on this website?" and then asked follow-ups if there was a positive response. In the test cell, 60.95% of respondents reported that Gap came to mind, while none did so in the control cell. (Are that many decimal points really statistically appropriate? It’d have to be a pretty big survey! Which is not to say I distrust the basic magnitude of the findings; I just suspect the decimal points are statistical mumbo-jumbo. I recommend Jacob Jacoby, Amy H. Handlin, & Alex Simonson, Survey Evidence in Deceptive Advertising Cases Under the Lanham Act: An Historical Review of Comments From the Bench, 84 Trademark Reporter 541 (1994), for an overview of relevant issues.) The court found the survey demonstrated that exposure to the old logo caused an association with Gap’s marks.

Ford also did a survey with the new G.A.P logo. In this case, the control cell replaced “gap” with “tap” and the tagline with “travel adventure people.” This test cell produced 38.61% who reported that Gap came to mind, while the control cell found 0.99% association. 37.62% of respondents who associated the new logo with Gap did so as a result of the “gap” in the new logo.

Defendant’s survey expert opined that the control should have included uses of “gap” to which Gap had not objected as a proper baseline, but the court disagreed. The expert also criticized the question as leading; Ford responded that it was appropriate in light of the survey’s purpose and that he’d asked the same thing of the control cell. (People have successfully argued that asking the same questions of the control cell avoids any problems with leading questions. I think that depends—if the question leads consumers to draw inferences they wouldn’t have drawn in ordinary exposure, then having a control cell doesn’t help cure the distortion from the real world. If the question just leads consumers to articulate things that they did think, but weren’t at the top of their minds, then having a control cell is quite useful. Trouble is: how do you know which situation you have?) The court concluded that a control cell with an open-ended question such as "What is the first thing that comes to mind when you see the name on this website”" might have been desirable. But “two different control cells could cause skewing in other respects and, by complicating and extending the questioning, affect the willingness of respondents to be questioned.” (Not sure I get this last point. If there’d been a separate third cell, why would the questions have been longer? Or does the court mean that open-ended questions lead to extended questioning?) The court found that the surveys were appropriate and relevant to show likelihood of association between the parties’ marks.

Hal Poret also conducted an infringement survey of people 18 or over who visited the area in which G.A.P’s NYC store is located and who were likely to shop for travel tours in the next 12 months. Defendant’s expert opined that intercepting those people on the streets near the store was inappropriate, because there were two Gap stores in the vicinity “and there was no way to control the effect of outside stimuli, e.g., passersby carrying Gap shopping bags or wearing Gap clothing.” The court disagreed because the method and locations “simulated the market conditions under which a consumer would see the G A.P Adventures storefront.” Poret initially just conducted a test, with no control, because Gap wanted to know whether it should worry about the NYC store. A year later, after Gap sued, Poret ran a control cell. Defendant’s expert argued that the time gap made it impossible to account for changes between cells, but the court found that the record didn’t show any change in context or environment that would be relevant.

The test cell used a photo of the front of the G.A.P NYC store. The control cell used a photo altered so that the storefront read “The Great Adventure People Adventures.” Defendant’s expert opined that the photo was not very clear and the nature of G.A.P’s business was not evident, but the survey was designed to test initial confusion and the photo was taken from G.A.P’s own website.

The survey asked respondents what services or products they would expect to be able to get at the store in the photograph; whether they thought the company that operates the store in the photograph also operates any other store or business, with followups if they said yes; and whether they thought that the company that operates the store in the photo got “authorization, that is, permission, from any other company to use their name, or not, or don't you know?” with followups if they said yes. The court found that the survey showed that 40.9% of respondents in the test cell were confused into thinking that (1) the G.A.P Adventures store was a Gap store, or (2) the G.A.P Adventures store was operated by the same company as Gap, or (3) the G.A.P Adventures store was authorized by Gap. None of the respondents in the control cell had the same reactions. The court credited this as evidence of likely confusion.

Defendant’s expert Yoram Wind used a marketing analysis to conclude that there were “dramatic” differences between the parties’ logos, target markets, store setup, goods and services, price range, means of distribution, number of stores, advertising, websites, value propositions, and overall messages. Thus, confusion and dilution were unlikely.

The court didn’t give this analysis significant weight because it didn’t address consumer perception “upon initially encountering G.A.P Adventures in the marketplace, as reflected by the evidence of actual confusion actually encountered by defendant throughout its history, and as measured by the survey evidence. The confusion arises from an association between plaintiffs and defendant that causes an attraction, because of the application of plaintiffs' famous marks to defendant's service offerings.”

Wind offered a very interesting survey showing that about 54% of respondents thought of Gap in reference to descriptive uses of “gap” such as "Watch the Gap" and "Bridge the Gap." (I personally remember the “Fall into the Gap” marketing campaign.) Wind argued that consumers think of Gap whenever the word “gap” is used, whether descriptively or not, and that if these association-evoking uses were benign, so was G.A.P’s. (I make a related point in my article arguing that dilution law is unconstitutional because it suppresses truthful commercial speech when there’s no evidence that commercial speech is more harmful than noncommercial speech; it’s also of some note that there is academic research using Gap as one of the test brands to determine whether dilution occurs, cited in the article.) The court found that the study merely confirmed the fame and strength of the Gap marks, and did not speak to likely confusion or dilution.

Gap knew about G.A.P for a while. It sought the domain name in 2005, and engaged in unsuccessful negotiations over its concerns with G.A.P’s use of the marks in Canada. Gap opposed G.A.P’s published application for US registration for G.A.P Adventures for travel services in 2006. G.A.P “opened a street-level concept store in New York City” in 2007 “to increase brand awareness.” Gap sued under three months after this event.

The court granted judgment in favor of Gap on its Lanham Act infringement claims, for the obvious reasons. Though some third-party registrations also use “gap,” G.A.P failed to introduce evidence that they were actually used, promoted, or recognized by consumers. The court found that proximity of goods and services weighed in Gap’s favor, “although only slightly.” There was no direct competition, but the customer classes overlapped, “as evidenced by G.A.P Adventures' 2007-2008 marketing plan, which describes G.A.P Adventures' ‘core user’ as someone who shops at Gap.” (Ouch!) “In addition. Gap and G.A.P Adventures both offer their products for sale on their websites and through retail stores. This minimal degree of proximity increases the likelihood of consumer confusion, but only slightly.” In terms of actual confusion, the court found initial interest confusion—increased credibility during the initial phases of a potential transaction. Consumer sophistication in making an expensive travel purchase doesn’t protect against this.

The court granted judgment in favor of G.A.P on the state and federal dilution claims. Gap failed to prove that the use of “gap” impaired the distinctiveness of Gap’s marks, “nor is defendant likely to do so, particularly in light of the injunction that will be issued.” Though both intent to create an association and actual association were present, association alone isn’t enough: the association has to be of the kind that can impair the distinctiveness of the famous mark. The court cited McCarthy and the Charbucks case. I really don’t know what to say about this: on the one hand, McCarthy is right; on the other, I don’t believe that junior uses can impair the distinctiveness of famous marks, for reasons given in the article above, and therefore anyone asked to prove something more than association is going to fail. I don’t know what evidence Starbucks had about Charbucks over and above what Gap proved; all I can think is that, similar to what happened in the Deere state-law case, the court thinks that Charbucks was actually tarnishing and this case is only about blurring, even though the Second Circuit explicitly rejected tarnishment in the Charbucks case.

The court then rejected G.A.P’s laches defense. Laches requires plaintiff’s knowledge of defendant’s use of the marks, inexcusable delay, and prejudice to the defendant. The court found that Gap acted in a commercially reasonable manner. It monitored G.A.P’s activities over time, opposed its US trademark application, and filed suit under three months after G.A.P opened its NYC store. Moreover, G.A.P clearly suffered no prejudice: it knew of Gap’s marks from the outset, and continued using its name “ever more intensively and provocatively, Whenever plaintiffs complained, defendant intensified its infringing conduct.”

The court granted an injunction. Initial interest confusion constitutes irreparable injury. Watch how free riding turns into injury: “Remedies at law will not compensate Gap for the boost in credibility that G.A.P Adventures receives as a result of this initial confusion as to its affiliation” (emphases added).

G.A.P was permanently enjoined from using “gap,” or any variation of the word “gap,” whether as an acronym, abbreviation, or otherwise. The court granted such a broad injunction because allowing something like G-A-P Adventures “could easily give way to an infringing use, as shown by the history of this case.” G.A.P has to transition to a new name in the US, and must instruct its employees not to use the word "gap" when referring to the company. It may use "formerly known as G-A-P Adventures," but only in connection with a new, non-infringing name. Further, G.A.P had to remove all articles of clothing from publicly accessible areas of its stores. (Presumably, that can’t cover clothing that doesn’t use “gap” or a variant. Or, you know, clothing that employees and customers are wearing. Somehow I don’t think that was worded the best way possible.)


Mark said...

Interesting case. Haven't read the decision yet, but your description suggests that the court essentially found initial interest confusion. ("upon initially encountering G.A.P Adventures in the marketplace, as reflected by the evidence of actual confusion actually encountered by defendant throughout its history, and as measured by the survey evidence. The confusion arises from an association between plaintiffs and defendant that causes an attraction, because of the application of plaintiffs' famous marks to defendant's service offerings.”). Any finding about what that initial attraction harms the plaintiff?

As to dilution, I agree with you that "anyone asked to prove something more than association is going to fail." But this just points up the bankruptcy of the whole concept: no one knows what it means for the association to impair the distinctiveness of a mark or how one would measure it if it happened. So all of these cases are just going to come down to some other factor like intent, and the court will then either cite Starbucks or cases like this, depending on the result they seek.

Probably also worth noting just how much money appears to have been spent on experts here.

Rebecca Tushnet said...

The harm finding is just what the court says about the injunction, nothing more. Interesting that we have a finding of bad intent but not dilution here. On the rest we are in complete agreement.

Mark said...

(a) someone should write a paper about trademark harm; (b) I see that about the intent finding. I guess maybe the Wind survey is doing some work here - maybe the fact that lots of other things also bring GAP to mind suggests this is just a drop in the bucket (and therefore isn't likely to impair the distinctiveness of the mark).

Rebecca Tushnet said...

(a) ;) (b) another case in which the court disavows reliance and then relies? I think our judicial system might need an analyst.

Anonymous said...

This case is absolutely ludicrous. If G.A.P. Adventures was specifically trying to enter the retail clothing industry then that would be one thing... however, they are a travel company! Shame on Gap for trying to monopolize the use of the letter g, the letter, and the letter p in an industry that they are not even a part of!

John McPhee said...

This smells of American protectionism. The "big" guy going after the little one. Where is the confusion? One offers a high end, eco friendly, life changing experiences. The other offers cheap, low quality, sweatshop products meant to boost the bottom line and please shareholders. G.A.P Adventures should appeal, appeal, appeal. One is a world leader in supporting local communities with projects like building schools and ophanages, teaching locals skills to help them lead better and more productive lives & eye clinics to restore vision. A company that really does give back to the communities it works within. None of which can be said for Gap clothing where it's only $$$ and big bonuses to suits. Funny how it only seems to be Americans who are confused by this matter. The rest of the world seems to have it's head screwed on straight!