©2021. Published in Landslide, Vol. 13, No. 3, January/February 2021, by the American Bar Association. Reproduced with permission. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association or the copyright holder.
The U.S. Supreme Court handed down its ruling in the final trademark case of the 2019–20 term in U.S. Patent & Trademark Office v. Booking.com B.V.1 on June 30, 2020. In an 8–1 decision authored by Justice Ginsburg, the majority held that consumer perception alone should determine whether terms are registrable as trademarks or generic and free for all to use.
The issue was whether a term that is generic for the class of goods or services can be protected as a trademark when followed by “.com.” The U.S. Patent and Trademark Office (USPTO) refused to register BOOKING.COM in four related applications for travel reservation services, finding that “booking” was generic for these services and that the addition of the generic top-level domain (gTLD) “.com” failed to cure the problem as a matter of law. After unsuccessfully appealing the trademark examining attorney’s refusal in the Trademark Trial and Appeal Board, the applicant sued the Director of the USPTO in district court, likely as a means to escape unfavorable Federal Circuit precedent, where an appeal could have been made. The district court, hearing the case de novo and declaring it a case of first impression in the Fourth Circuit, ruled that a generic term plus a gTLD creates a merely descriptive mark and that Booking.com had acquired secondary meaning. The finding that the mark was descriptive with acquired distinctiveness, but not the per se rule, was affirmed by the Fourth Circuit.2
At the Supreme Court, Booking.com did not contest that “booking” is generic for the services it offers, and the USPTO did not contest the secondary meaning evidence proffered by Booking.com showing that consumers do not perceive “booking.com” to be generic. The case then was simply about the proper test for genericity in “generic.com” cases. The USPTO argued the rule in generic.com cases should follow the rule that the addition of “Inc.” or “Co.” to generic terms does not create a composite that is not generic. This rule was established in an 1888 Supreme Court precedent—Goodyear’s India Rubber Glove Manufacturing Co. v. Goodyear Rubber Co.3—that held that a generic term embellished only by the generic designation of a business entity necessarily produces a generic composite.
In its first domain name case, which was also the first case in which the Court had ever permitted a live audio recording of oral argument, the Court disagreed. Holding that “[a] term styled ‘generic.com’ is a generic name for a class of goods or services only if the term has that meaning to consumers,” the Court ruled that the USPTO must base its decision on evidence of consumer perception and not on any other rule.4 The USPTO had argued that the consumer perception inquiry was unnecessary in this case just as it had been in Goodyear.
The Court did not explicitly overrule Goodyear but chose not to apply it. This case was distinguishable, it reasoned, because only one entity can occupy a particular domain name. Thus, while there could be several entities calling themselves “Wine Co.,” there can be only one wine.com. This distinction is not a logical distinction, but a factual distinction. It presumes that consumers know that under our federal system, corporate law is governed by state law so that different companies can register the same corporate name in different states, but under the technical rules of the domain name system, only one entity may register that name in .com. This factual distinction is then the basis for an argument that consumers would perceive a “generic.com” differently than they would perceive a “Generic Corp.,” but it is not a policy rationale for evading the Court’s previous decision. The majority, however, resisted the implications of this factual distinction. If it is significant that there can be only one wine.com, presumably that is because consumers will understand the exclusivity conveyed by such a designation. It is curious then that the Court stated that it did “not embrace a rule automatically classifying such terms as nongeneric.”5
The Court supplemented this distinction with an interpretation of Goodyear that clashed with the USPTO’s. The USPTO understood Goodyear to pronounce a rule that a generic term supplemented only by a generic corporate designation could not produce a protectable mark as a matter of law. Characterizing such a rule as “unyielding,” the Court noted that it “entirely disregards consumer perception,” which the Court stated is a “bedrock principle of the Lanham Act.”6 Instead, the Court saw the Goodyear rule as “a more modest principle harmonious with Congress’ subsequent enactment” of the Lanham Act.7 The Court suggested that a rule “incompatible” with consumer confusion is therefore incompatible with the Lanham Act. Booking.com had argued that Goodyear was “repudiated” by the Lanham Act.8 The Court declined such a pronouncement, but suggested that any rule that doesn’t yield to consumer perception has been repudiated.
The rule Goodyear announced does not yield to consumer confusion. Goodyear declared:
[P]arties united to produce or sell wine, or to raise cotton or grain, might style themselves Wine Company, Cotton Company, or Grain Company; but by such description they would in no respect impair the equal right of others engaged in similar business to use similar designations . . . . Names of such articles cannot be adopted as trade-marks . . . ; nor will the incorporation of a company in the name of an article of commerce, without other specification, create any exclusive right to the use of the name.9
The Court thus stated that the adoption of a “Generic Corp.” cannot be an adopted trademark as a matter of law. Had there been evidence indicating that “Goodyear Rubber Company” was not understood by consumers as generic, the result would have been the same. The Court stated that “[t]he designation Goodyear Rubber Company not being subject to exclusive appropriation” could be freely employed by others.10 It was a formulation that was not subject to appropriation regardless of the facts.
The Goodyear Court’s rule is thus incompatible with the Booking.com rule: “Whether any given ‘generic.com’ term is generic, we hold, depends on whether consumers in fact perceive that term as the name of a class or, instead, as a term capable of distinguishing among members of the class.”11 The Court thus silently overruled Goodyear.
The Customer Is Always Right
Significantly, the Court’s broad statement that following Congress’s enactment of the Lanham Act all rules must be compatible with consumer perception cites no authority within Congress’s enactment for such a changed standard. Indeed, the Lanham Act offers no citable language for this proposition. In fact, the Lanham Act makes consumer perception subservient to other doctrines, such as the absolute registration bars (deceptive marks, flags, etc.), and certain defenses, such as fair use and laches.
Booking.com convinced the Court that the Goodyear rule was odious. In its briefs and oral argument, the rule was disapprovingly characterized as a “per se” rule, a classification taken up by the majority. Booking.com also effectively denigrated the precedent as being too old. In the oral argument, Chief Justice Roberts twice noted that the Goodyear case was 130 years old and pointedly asked why we would now focus on it.12 Likewise, Justice Alito, noting that Goodyear was “from a different era,” asked for a rule “that makes sense in the Internet age.”13 Also echoing Booking.com’s argument, the Court noted its inconsistent application by the USPTO and stated that its adoption could therefore put issued registrations in peril.
None of this, however, gives the Court license to ignore its own precedent. Goodyear is far from out of date; it perfectly captures the issue of claiming trademark rights in domain names. The rule advanced by the USPTO is an extension of Goodyear for the internet age. The Court, however, gave itself license to ignore Goodyear with this statement: “we discern no support for the [Goodyear rule] in trademark law or policy.”14 Therein lies the importance of this case going forward. Goodyear is a part of current trademark law, and its rule rests entirely on trademark policy. The policy it espouses is trademark law’s protection of competition. For the majority, that policy had no purchase in this case. Harm to competition just wasn’t tangible. In the void, the Court latched on to consumer perception as the driver of trademark policy.
Goodyear was concerned with the monopolization of a generic term by a Generic Co. Meanwhile, the majority in Booking.com was convinced by Booking.com that attempts to assert exclusive rights in generic.coms were “losers.”15 Booking.com also promised that if it received a registration, it would not enforce it against similar domains. Both the majority and dissent referred to this promise and cited it in the oral argument transcript, presumably to make it binding? That plus $5 will get a competitor a cup of coffee while it reads the cease and desist letter it will receive from Booking.com for using a similar domain. Booking.com also convinced the majority not only that highly descriptive terms result in thin trademarks, but also that the doctrine of descriptive fair use adequately protects competition. If in fact the trademark offers Booking.com such weak rights, one wonders why it chose to pay to litigate a registration refusal all the way to the Supreme Court when it already claims common-law rights in the designation, holds related registrations in design marks, holds 85 registrations for Booking.com worldwide (each of which can be relied on in a domain name dispute), and has secured the domain name. If all it really wanted was a thin trademark, it could have more easily achieved that by simply disclaiming the generic term in a stylized mark.
Justice Breyer, the sole dissenter, was also the sole link to all of the Court’s precedent protecting competition. Interestingly, Justices Ginsberg and Breyer often found themselves on opposite sides in intellectual property cases. Breyer saw the monopoly danger and issued one of his most tight and convincing opinions.
Goodyear, the USPTO, and Justice Breyer all saw a danger that the majority did not. Just as a company selling wine should not be able to own a trademark in “wine,” neither should it be able to do an end run around that prohibition by adopting the moniker “Wine Co.” To allow a trademark in such cases, according to Goodyear, would “impair the equal right of others engaged in similar business to use similar designations, for the obvious reason that all persons have a right to deal in such articles, and to publish the fact to the world.”16 In sharp contrast, the Booking.com majority concluded that the USPTO’s concern about granting “undue control over similar language, i.e., ‘booking,’ that others should remain free to use . . . attends any descriptive mark,” and “[r]esponsive to it, trademark law hems in the scope of such marks short of denying trademark protection altogether.”17 The majority seems not even to find a policy justification for the law’s distinction between generic and descriptive terms. The equivalent in patent law would be giving up § 101 of the Patent Act on the assurance that § 102 and § 103 would protect the public domain. Or perhaps Booking.com goes even further, suggesting that subject matter constraints are unnecessary when effective constraints exist on the scope of rights.
Having Its Cake and Eating the Competition Too
Although the decision was not surprising, it was nonetheless disappointing. In elevating the public policy work that consumer perception does in trademark law, the Court has opened the door to monopolizing generic terms, especially to large companies. Justice Breyer’s dissent was spot on. In deciding that a generic.com can be registered as a trademark if it has a source-identifying meaning to consumers, the Supreme Court allows Booking.com to have its cake and eat it too. In registering www.booking.com, the online booking company monopolized the generic word for its services in .com, and in adopting the domain name as its business name, it immediately conveyed to consumers the services it provides. Other advantages follow, including a conspicuous web presence and—as a result of recent domain name system rules—the sole right to own the term as its own top-level domain: .booking. The usual consequence of adopting a business name that does no more than convey the services provided is that the name can’t be registered as a trademark; a trade-off many would accept. But the Supreme Court removed that consequence and cleared the way for Booking.com to also eat its cake.
Preserving competition is the basis of the trademark rule against monopolizing generic terms. Booking.com is already one of the largest online travel agencies in the world. It is owned by Booking Holdings Inc., which also owns many of the other booking websites, including Priceline.com, Agoda.com, Kayak.com, Cheapflights, Rentalcars.com, Momondo, and OpenTable.18 The company has had antitrust issues in the European Union, U.K., and Turkey.19 In the U.S., it already owns the trademark VILLAS.COM,20 and it purchased the domain and priority rights to BOOKINGS.COM, which now simply redirects to booking.com. In addition to owning the gTLD .booking, the company also owns the gTLD .hotels.21 This is not a company that invites competition.
Survey Says: Problems Ahead
The USPTO is on a Supreme Court losing streak in trademark law. Since 2017, including the instant decision, the USPTO has lost all three of the cases it appealed to the Court.22 In Booking.com, one may wonder why the USPTO seemed to have put all of its eggs into the basket of precedent as it argued only that the Goodyear rule should apply; it had no alternative argument. In oral argument, Justices Kagan and Gorsuch both asked the USPTO for an alternative to the per se rule. By not challenging the district court’s factual findings, the USPTO seems to have chosen an all-or-nothing approach.
Anyone who has thought about how generic terms are identified by the USPTO will understand why this was a necessary litigation strategy. Goodyear was not just a preferred rule for the USPTO—it was the only practical rule to prohibit the registration of generic.coms. Under the USPTO proposed rule, survey evidence is irrelevant. As a result of Booking.com, the USPTO will now have to contend with survey evidence of consumer perception in generic.com cases. But what can the USPTO—in the form of a single examiner—do to counter such evidence? If a company has the resources to introduce survey evidence that a generic.com is not perceived by consumers as generic, the USPTO will not be in the position that a court is in when genericism is litigated between two parties. It will not have the benefit of considering counterevidence in the form of an alternate survey or a survey expert’s critique.
A critical issue going forward will be how to ensure that the survey evidence indicates proper consumer perception because not all consumer perception evidence is relevant. The doctrine of “de facto secondary meaning” holds that once a designation has been determined to be generic, no amount of consumer recognition can transform its status into that of a protected mark. As Judge Friendly, author of the oft-cited Abercrombie & Fitch Co. v. Hunting World, Inc. decision that made doctrinal sense of generic terms, stated:
[E]ven proof of secondary meaning, by virtue of which some “merely descriptive” marks may be registered, cannot transform a generic term into a subject for trademark. . . . [N]o matter how much money and effort the user of a generic term has poured into promoting the sale of its merchandise and what success it has achieved in securing public identification, it cannot deprive competing manufacturers of the product of the right to call an article by its name.23
Under this rule, if booking.com was generic when the domain was first used, the fact that the owner of that website has now convinced consumers that the designation is its brand name is irrelevant.
The influential McCarthy treatise notes that the danger in being too accepting of survey evidence would “allow spurious claims of trademark status for generic names based on shaky and unreliable evidence of customer perception.”24 McCarthy cautions that “[i]f courts were not scrupulous in weighing the evidence, this could constitute a large loophole in the classic rule against ownership of generic names,” and “the way for a court to handle such a case is to consider the proof of customer perception that the trademark proponent puts into evidence and to balance it against the public policy of open competition and free use of generic names.”25 The work of examiners in evaluating survey evidence in generic cases is considerable, but their tools are few.
Where a private party seeks to invalidate a mark as generic, it is better positioned than the USPTO to undermine survey evidence with conflicting evidence, which may include expert opinion on survey design and additional surveys. Because the USPTO can’t hire experts or commission surveys, it cannot effectively guard the public domain in the absence of a per se rule. What the USPTO can do in lieu of a survey is comb databases and list the generic uses of the term at issue. But in the case of a generic.com, there necessarily will be less of this evidence because competitor use may be nonexistent. For example, Booking.com’s competitors routinely use “booking,” but they don’t use “booking.com” because any reference to a domain name provides consumers with not only the address of the competition but also a nonstop free shuttle ride to their place of business. Thus, in evaluating the evidence in the case of a generic.com, the USPTO must guard against the public domain with both of its arms tied behind its back.
Defeating De Facto Secondary Meaning
A straightforward method for the USPTO to deal with survey evidence and weed out de facto secondary meaning is to sort it by the date it was created. Booking.com’s survey showing 74.8 percent of respondents recognized the domain name as a brand name was not conducted until 2016, after it had long used and heavily advertised the mark.26 Indeed, the company’s briefs and expert reports recount an extraordinary effort—especially in the five years prior to the survey—to create brand awareness.27 So did the 2016 survey measure the original understanding of booking.com or de facto secondary meaning? If the latter, the USPTO needs a workable means to discount—or even ignore—this evidence, such as a rebuttable presumption that it is evidence of de facto secondary meaning based on its date of creation. For the USPTO to otherwise simply accept survey evidence would create, in the words of McCarthy, “a large loophole in the classic rule against ownership of generic names.”
The hypothetical “washingmachine.com” exposes the vulnerability in the majority’s ruling. Invented as a control in Booking.com’s survey—no washingmachine.com website or company exists—30 percent of respondents nonetheless identified it as a brand.28 But the percent indicates that only a minority view a domain name as a brand; the majority—60.8 percent—believed washingmachine.com was generic.29 Under the majority’s consumer perception rule, these facts indicate that washingmachine.com is generic.
The key question then is what accounts for the 44.8 percent difference in brand recognition between washingmachine.com and booking.com. What other than de facto secondary meaning can explain these results? If washingmachine.com were operated by a company that invested an advertising budget comparable to Booking.com and created a comparable user base, it would also likely see a 44.8 percent boost in brand recognition. If said washingmachine.com were allowed to introduce such survey evidence, it would thereby be enabled to overcome its generic designation with de facto secondary meaning.
The majority didn’t attend to this vulnerability likely because it assumed that booking.com was not generic when first used based on its belief that consumers think domains necessarily indicate single entities. That assumption, however, is undermined by the washingmachine.com survey results that 60.8 percent thought it was generic. The majority’s assumption was supported to some extent—30 percent thought it was a brand for no apparent reason. These respondents were presumably reacting to the .com suffix since 100 percent of the survey respondents identified “supermarket” as generic. Still, the majority’s assumption is belied by the majority of respondents.
The majority’s assumption that generic.coms are understood as indicating a single entity may also be based on the belief that consumers would never utter a generic.com except to refer to a single entity. In oral argument, Chief Justice Roberts observed that “nobody refers to [travel booking sites] as Booking.coms,”30 and the majority similarly observed that a consumer would never “ask a frequent traveler to name her favorite ‘Booking.com.’”31 And yet, Booking.com’s own lawyer demonstrated in oral arguments just how a consumer would do this unthinkable act when stating, “I have searched every grocerystore.com looking for toilet paper. I have now started looking at every hardware.com.”32
This reliance on assumed consumer understanding of generic.coms is yet another example of the contrasting viewpoints between the Booking.com Court and the Goodyear Court. Under Booking.com’s reasoning, “Wine Corp.” can be registered because there is no genus of goods that consumers commonly refer to as “wine corp.” The Goodyear Court was not focused on consumer use, but instead loopholes in the generic prohibition. Its rule effectively prevented a wine seller who couldn’t register “wine” from registering “Wine Co.” because of its interest in protecting competition regardless of consumer understanding.
When You Know Who’s Going to Win at Monopoly but You Have to Keep Playing
The policy question at the heart of this case is whether the prohibition on registering generic terms is worth it. If it is, it should be worth not creating a loophole that you could sail a cruise ship through. If it is, it should be worth tolerating the inflexibility of a per se rule. If it is, it should even be worth ignoring current consumer perception.
With its Booking.com decision, not only did the Supreme Court effectively overrule one of its own precedents, but it also reversed the established rule that generic.coms are unprotectable. Both the dissenting judge in the Fourth Circuit decision and the McCarthy treatise state that the USPTO’s per se rule was the accepted position.33 These reversals have now made generic.coms much more valuable than they already were.34 The heretofore unexpected trademark rights that can now flow from a generic.com will create a renewed market interest in these domains. The irony is that the newly minted top-level domains, such as .car, .travel, and .holiday, have been made available as a means of escaping the hold of .com.35 This decision will enhance the magnetic force of .com.
Generic.coms present an even greater risk of monopoly powers than a Generic Corp. does. As the dissent in the Fourth Circuit correctly stated, “[t]his case addresses a problem that Booking.com chose to bring upon itself.”36 The adoption of a generic term as a trademark always involves a problem of the applicant’s own creation. Now, however, in the case of a generic.com, the applicant gets to have it both ways without having to make the usual trade-off between trademark rights and instant communication of the business offerings. Whereas the doctrine of trademark genericity follows the saying that you can’t have your cake and eat it too, the Booking.com rule allows a generic.com that purchases a pricey cake to grab the key to the bakery.
1. No. 19-46 (U.S. June 30, 2020).
2. See Booking.com B.V. v. U.S. Patent & Trademark Office, 915 F.3d 171 (4th Cir. 2019).
3. 128 U.S. 598 (1888).
4. Booking.com, No. 19-46, slip op. at 2.
5. Id. at 7.
6. Id. at 6.
7. Id.; see 15 U.S.C. §§ 1051 et seq.
8. See Transcript of Oral Argument at 39, Booking.com, No. 19-46 (U.S. May 4, 2020).
9. Goodyear’s India Rubber Glove Mfg. Co. v. Goodyear Rubber Co., 128 U.S. 598, 602–03 (1888).
10. Id. at 604.
11. Booking.com, No. 19-46, slip op. at 7.
12. Transcript of Oral Argument, supra note 8, at 6–7.
13. Id. at 18–19.
14. Booking.com, No. 19-46, slip op. at 8.
15. Transcript of Oral Argument, supra note 8, at 49.
16. Goodyear’s India Rubber Glove Mfg. Co. v. Goodyear Rubber Co., 128 U.S. 598, 603 (1888); see also Booking.com, No. 19-46, slip op. at 11 (Breyer, J., dissenting) (“Multiple brick-and-mortar companies could style themselves ‘The Wine Company,’ but there can be only one ‘wine.com.’ And unlike the trademark system, that exclusivity is worldwide.”).
17. Booking.com, No. 19-46, slip op. at 12 (majority opinion).
18. See Booking Holdings, Inc., Annual Report (Form 10-K) 1 (Feb. 26, 2020).
19. See Press Release, Competition & Mkts. Auth. [U.K.], CMA Closes Hotel Online Booking Investigation (Sept. 16, 2015), https://www.gov.uk/government/news/cma-closes-hotel-online-booking-investigation; Sam Schechner, Booking.com Settles Antitrust Complaints [EU], Wall St. J. (Apr. 21, 2015), https://www.wsj.com/articles/booking-com-settles-antitrust-complaints-1429623279; Gulsen Solaker & Tuvan Gumrukcu, Turkish Court Halts Activities of Booking.com over Breach of Competition Law, Reuters (Mar. 29, 2017), https://www.reuters.com/article/us-turkey-tourism-booking/turkish-court-halts-activities-of-booking-com-over-breach-of-competition-law-association-idUSKBN1701XC.
20. VILLAS.COM, Registration No. 4,883,510.
21. See Application Details for “.hotels,” ICANN New Generic Top-Level Domains, https://gtldresult.icann.org/applicationstatus/applicationdetails/1589 (last visited Dec. 11, 2020). ICANN is a private, nonprofit entity that, among other things, administers the policy of the global Domain Name System (DNS).
22. See Iancu v. Brunetti, 139 S. Ct. 2294 (2019) (holding that the Lanham Act’s bar on the registration of “immoral” or “scandalous” trademarks violates the First Amendment by discriminating on the basis of viewpoint); Matal v. Tam, 137 S. Ct. 1744 (2017) (holding that the disparagement clause of the Lanham Act was facially invalid under First Amendment protection of speech, as speech may not be banned solely because it expresses ideas that offend).
23. 537 F.2d 4, 9 (2d Cir. 1976).
24. 2 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 12:47 (5th ed. 2020).
26. See Joint Appendix at 62, 65, U.S. Patent & Trademark Office v. Booking.com B.V., No. 19-46 (U.S. filed Jan. 6, 2020).
27. “Applicant has advertised its BOOKING.COM mark extensively in the United States through web advertisements, television commercials, movie theatre previews, cross-linking partnerships with other major websites, social media participation, direct mailings, and other channels of advertising.” Id. at 23. Apparently, Booking.com’s television ads received 1.96 billion visual impressions from U.S. consumers in 2016 alone, its internet ads received 1.34 billion visual impressions from U.S. consumers in 2016 alone, an average of 10.3 million unique visitors from the U.S. per month, and it has over 2.2 million U.S.-based subscribers to newsletters. See id. at 17–18, 146–47.
28. Id. at 165.
30. Transcript of Oral Argument, supra note 8, at 8.
31. U.S. Patent & Trademark Office v. Booking.com B.V., No. 19-46, slip op. at 5 (U.S. June 30, 2020).
32. Transcript of Oral Argument, supra note 8, at 49.
33. See Booking.com B.V. v. U.S. Patent & Trademark Office, 915 F.3d 171, 193 (4th Cir. 2019) (Wynn, J., concurring in part and dissenting in part) (“[The majority] conflicts with the determination that every other court has reached in cases, like the instant case, involving the registration or enforcement of a proposed mark composed of a generic Secondary Level Domain and a Top Level Domain.”); 2 McCarthy, supra note 24, § 12:39.50 (“The general rule is that the addition of a top level domain indicator (such as ‘.com’) to a generic name does not turn the composite domain name into a protectable trademark for the goods or services named by the generic name.”).
34. For example, in 2015 the domain cars.com was purchased for $872 million. See Cars.com Domain Name Sale Valued at $872 Million Dollars!, DomainGang (Feb. 26, 2015), https://domaingang.com/domain-news/cars-com-domain-was-valued-at-872-million-dollars.
36. Booking.com, 915 F.3d at 193 (Wynn, J., concurring in part and dissenting in part).