It’s that time of year again, when people are traveling hither and yon.  If you haven’t made your reservations by now, or the French strikes have you worried, it’s likely you will find yourself at midnight scrolling through online travel sites to snag a last-minute, 20% off deal. may be just the ticket.

A Date in Washington, D.C. will soon be Booked

Even the Supreme Court has turned its attention in that direction.  In early November, the Court granted certiorari in United States Patent and Trademark Office et. al v. BV (case number 19-46). The mark at issue is the straightforward name of the website – BOOKING.COM owned by a Netherlands company BV. (Author’s disclosure: I’ve successfully used several times; their coverage in Europe is particularly complete and the British turns of phrase rather endearing to an American ear.)  The case turns on the registrability — or not — as a U.S. trademark of BOOKING.COM, comprised of the word BOOKING plus the top-level domain suffix .COM. (Let’s refer to Top Level Domains as TLDs; the Booking part of the domain address is known as a Second Level Domain or SLD.)

The case is hardly new. The appeals process has wended its way slowly from an initial examiner’s refusal to the Trademark Trial and Appeal Board to the Eastern District of Virginia and then to the Fourth Circuit Court of Appeals. The case raises questions about the registrability of commonplace e-commerce site names that achieve some degree of consumer recognition.  It puts in jeopardy the territorial division staked out in the last twenty-five years between trademark rights and registration and domain name registration. Should ordinary website names like HOTELS.COM be eligible for trademark registration if the sites have endured and achieved a measure of commercial success? The TTAB and lower courts have earlier said “no” to registration of HOTELS.COM and its ilk. But other .com sites have been registered by the USPTO:,, and, to name a few.

The Supreme Court will be taking a look at this patchwork of decisions and the split between the Fourth Circuit’s decision (915 F. 3d 171 (4th Cir. 2019), on the one hand, and Federal Circuit and Ninth Circuit decisions rejecting registration of website names like and  Along the way, the Justices will need to examine the boundaries between generic marks and merely descriptive marks. And they will have to determine the import of creating a compound mark by adding a TLD (like .com or .net) to an otherwise generic term, where no pun or “wink” to the public is intended.

An Abbreviated History of the Case

Amsterdam Canals 2017 BV sought to register BOOKING.COM in the U.S. as a “block letter” word mark and in several stylized versions, including color, for various services, including for online hotel reservations  based on the company’s use of the marks since 2006. (The USPTO website displays the stylized marks in their signature chunky font with light and dark blue highlights.) According to USPTO records, the initial applications were based on a prior international registration at World Intellectual Property Organization, an avenue to registration permitted by Section 66(a) of the Lanham Act. Upon initial examination, a USPTO examiner rejected the application for the block letter form, finding that the mark BOOKING.COM was generic as applied to relevant services, the hotel reservation services recited in the application. [The services initially included some in Class 39, but the lower court only found that the Class 43 services would be protectible, so the current case deals only with for hotel reservation services.] In the alternative, the USPTO initially concluded that the marks were merely descriptive but that the applicant had failed to establish that the marks had acquired secondary meaning by the time of the filing of the application. (Later in the process before the Eastern District, the applicant did mount persuasive evidence of secondary meaning, were the mark to be viewed as merely descriptive in the first instance.)

Generic vs. Descriptive

Until rather recently, and certainly before the .com era, a bright line divided generic marks and descriptive marks. With language borrowed from the 19th century, the traditional trademark opinions  explained that generic terms indicate the genus of a good, the very name by which it is commonly known, that cannot indicate the source of such good. By comparison, a descriptive mark, and certainly suggestive marks, indicate the species to which a certain good belongs, and can serve to indicate the source of the goods. These Linnean explanations have never much impressed my law school students. Rather, they appreciate that a generic name is one that is the ordinary name for an item in commerce. A descriptive term is one that limns out a quality or characteristic of a good, even if a rather prosaic attribute. If a descriptive term is no more than that, it is viewed as merely descriptive. As such, it may also fail the test for registration if no evidence is marshaled that the mark has become associated with the source of the goods.

Do you know a Generic mark when you see it?

In, the Fourth Circuit laid out its three-step test to determine if a mark is generic: 1) first identify the class of products or services to which use of the mark is relevant; 2) identify the relevant consuming public; 3) determine whether the primary significance of the mark to the relevant public is as an indication of the nature of the class of the product or services to which the mark relates (in which case the mark would be generic) or as an indication of source or brand (which suggests that the mark is not generic.) Due to the stringent consequence of a finding of genericness, the USPTO has the burden to establish that an applied for mark is generic.

The appellate court cited to Abercrombie & Fitch Co. v. Hunting World, Inc. (537 F. 2d 4, 2d Cir. 1976) involving the apparel maker’s attempt to secure SAFARI for its exclusive use on multi-pocketed hunting-style jackets based on extensive advertising and promotion. Remember the look?  However, “no matter how much money and effort the user of a generic terms has poured into promoting the sale of its merchandise and what success it has achieved in securing public identification,” that user cannot claim the exclusive right to obtain a registration for the common name used for the product or service. Nor could others be excluded from using the same term. When it comes to hunting jackets and gear, it’s SAFARI for all to use.

Measuring What the Public Understands

In the case before the Supreme Court, if the Fourth Circuit’s test is correct, the key question is: does the consuming public understand BOOKING.COM to refer primarily to the actual service (of booking hotel and other reservations online) or to the source of that service (a particular site owned by a company that commands certain hotel listings)? In assessing the primary significance to the public, the Fourth District found that the lower court did not err in ruling that the USPTO had not satisfied its burden. The USPTO failed to prove that the relevant public understands BOOKING.COM, taken as a whole, to refer to general online reservation services rather than to the services of a particular Dutch concern. Because the district court considers evidence de novo and was not bound by the findings of the TTAB, the lower court was entitled to make its own determination of the weight of consumer evidence over other evidence like dictionary definitions or trade journal mentions.

But how does a court determine what the public understands by a term like BOOKING.COM? Usage can be probative of public understanding, said the Fourth Circuit. (The dissent complained that legal error should have been found. In its view, the lower court overemphasized evidence of how the public used marks rather than how the public understands marks. But the majority gave the district court a pass in this regard.) Consumer survey evidence is the other way to get at understanding.  In the district court appeal from the TTAB decision denying registration, BV submitted the results of a Teflon survey which demonstrated that almost 75% of respondents identified BOOKING.COM as a brand name rather than a general reference to hotel reservation sites. The survey respondents seem to understand that was a website to reserve a hotel in Berlin for New Year’s Eve, for example, not just any reservation site in general.

Can a term that is “born generic” be recaptured as a Trademark?

Trademark doctrine has long recognized two types of generic marks: the mark that is “born” generic (a rose is a rose for the sale of roses) or a mark that was at one time invented or coined but becomes indistinguishable from the name of the thing itself and therefore generic. (Remember the awkward attempts to recapture THERMOS as a brand name for an” insulated heat-preserving beverage container”?). A mark that suffers this fate no longer points to a source of the product. It merely “is” the product in the eyes of consumers. In either case, a generic mark cannot escape its essence. It cannot be converted to a protectible trademark under the law.  Further, words like Company or Incorporated or Store, are likewise generic. Their addition to another term like Hotels or Mattresses cannot save a compound mark from classification as generic.

But in the internet era, does adding a TLD to a word that itself lacks distinctive qualities have the same result as affixing Company or Store, resulting in generic treatment? After all, under the schema created for TLDs by ICANN in the mid 1990s, a .com ending was to signify a commercial site, whereas a .org ending was reserved for not-for-profit organizations and .net for networking entities.  TLDs serve as directional signposts, telling the public (and computers) where to find certain sites on the Internet.  Each domain address comprised of an SLD name and a TLD is in fact a unique address in computer language.

The Fourth Circuit recognized that a TLD like .com by itself lacks source-identifying significance.  So merely adding .com does not wave a magic wand over a generic term to convert it into a source signifier.  But the Fourth Circuit also declined to adopt a per se rule against protecting as marks domain names that are formed by adding a TLD to a generic term. Rather, the court found comfort in a Fifth Circuit decision that held open the possibility that in “rare circumstances” a TLD may render a term sufficiently distinctive to be protectible as a mark.  Under this thinking, would for a group of tennis training academies be registrable as a “witty double entendre”? Should the correct inquiry be: what does the public understand by a compound mark that includes a .com, .org or .net ending?

What’s really at stake?

The USPTO and the dissent in the Fourth Circuit have argued that granting registration will provide BV with a “weapon to freeze out potential competitors” from using “booking” for similar or related services in their own marks. Having launched its site with a broad generic name to attract a wider user base, says the dissent, should now have to live with the consequences of such choice, at least as far as US trademark registration goes.

Because the basis of trademark law is to protect competition and to prevent consumer confusion about the source of goods and services, this case raises the ante. Can companies gamble that the choice of a commonplace name for their website can later be converted to a registrable trademark if the sites prove successful?  Is such an outcome fair? I’m putting my chips with the dissent.

The case is attracting amici briefs. We will report again on the case in the spring term of the Court.