In re Vox Populi Registry Ltd.: Trademarks

The Federal Circuit affirmed refusal of .SUCKS because consumers see a gTLD, not a brand: the leading appellate word on failure to function as a mark.

In re Vox Populi Registry Ltd., 25 F.4th 1348 (Fed. Cir. 2022), decided February 2, 2022, is the Federal Circuit’s leading statement on the USPTO’s most rapidly growing ground of refusal: failure to function as a mark. Vox Populi operates the registry for the .SUCKS generic top-level domain, the internet extension that lets anyone register a gripe-ready address like brandname.sucks. When Vox tried to register .SUCKS itself, in standard characters and in a retro pixelated design, the Office refused both applications, not because the term was descriptive or scandalous, but because consumers would not perceive it as identifying any source at all. Judge Timothy Dyk, writing for a panel that included Judges Lourie and Stoll, affirmed.

The decision matters well beyond domain-name registries. Failure-to-function refusals now reach hashtags, viral slogans, informational phrases, and ornamental designs, and Vox Populi supplied the appellate framework: consumer perception is the touchstone, the Board’s findings about that perception get substantial-evidence deference, and stylization saves unregistrable wording only when the design makes its own impression.

At a glance

Case: In re Vox Populi Registry Ltd., 25 F.4th 1348 (Fed. Cir. 2022), No. 2021-1496
Decided: February 2, 2022, by Judge Dyk for a panel with Judges Lourie and Stoll; the TTAB’s refusal was affirmed
Holding: Substantial evidence supported the Board’s finding that consumers perceive .SUCKS as merely one of many gTLDs used in domain names rather than as a source identifier, and the pixelated stylization did not create a separate commercial impression capable of carrying the mark to registration
Significance: The controlling Federal Circuit precedent on failure-to-function refusals, cited throughout examination practice as of July 2026 alongside TMEP § 1202.04

The statutory roots of failure to function

The Lanham Act nowhere uses the phrase “failure to function,” which is why the doctrine surprises applicants. It flows instead from the statute’s definitions. Sections 1 through 3 of the Act, 15 U.S.C. §§ 1051-1053, authorize registration of trademarks and service marks, and Section 45, 15 U.S.C. § 1127, defines a service mark as matter used to identify and distinguish the services of one person from the services of others and to indicate the source of the services. Matter that does not perform that identifying job is simply not a mark, whatever its other qualities, and so cannot be registered.

The court situated the doctrine in a longer lineage. Its predecessor observed in In re Cooper, 254 F.2d 611 (C.C.P.A. 1958), that the source-identifier and descriptiveness inquiries are complementary sides of the same coin, but Vox Populi confirmed that the source-identifier requirement is broader than the familiar categories of generic and descriptive matter. The Board’s own precedent, collected in TMEP § 1202.04, refuses matter that merely conveys information or an informational message, such as the refusal of I ♥ DC for apparel and souvenirs in D.C. One Wholesaler, Inc. v. Chien, because ubiquitous expressions of enthusiasm belong to everyone and point to no one. The analysis always centers on how the relevant public actually encounters the matter, examined through the applicant’s specimens and other real-world evidence of use.

Vox’s two applications and the road to the Federal Circuit

Vox filed two applications. The ‘941 application sought to register the standard character mark .SUCKS in Class 42 for domain registry operator services related to the gTLD and in Class 45 for domain name registration services. The ‘215 application sought the stylized form of .SUCKS, which Vox’s own brief described as a retro, pixelated font resembling how letters were displayed on early LED screens.

The examining attorney refused both on the ground that, as used with the identified services, each fails to function as a mark. The Board affirmed, concluding that the standard character mark .SUCKS “will not be perceived as a source identifier” and instead “will be perceived merely as one of many gTLDs that are used in domain names.” As to the stylized version, the Board found that the literal element fared no better and that “the stylized lettering or design element in the mark does not create a separate commercial impression and is not sufficiently distinctive to ‘carry’ the overall mark into registrability.”

Vox appealed only the ‘215 stylized application, but spent much of its brief arguing that the wording .SUCKS functions as a mark regardless of design. That litigation choice framed the two questions the Federal Circuit answered: whether substantial evidence supported the Board’s perception findings about the wording, and whether the design added enough to register anyway.

Consumer perception controls, and the evidence pointed one way

Reviewing the Board’s factual findings for substantial evidence, the court found the record comfortably sufficient. Vox’s own specimens used .SUCKS to refer to a product rather than a provider, advertising the exceptional value of Registry Premium names such as life.sucks and divorce.sucks. Online articles discussing Vox used .SUCKS the same way, as did third-party domain name registrars, who sold .SUCKS domains to the public as a product rather than pointing to Vox as a service source.

Vox countered with a declaration from its chief operating officer describing substantial advertising spending under the .SUCKS brand and double-digit year-over-year growth in registrations since the domain debuted in 2015, plus declarations from two registrar customers who said they perceive .SUCKS as Vox’s service mark. The court held the Board reasonably discounted all of it. Advertising and sales volumes, while relevant, are not by themselves dispositive of how consumers perceive a mark; intending to present matter in the manner customary for service marks does not mean consumers receive it that way. And the registrar declarants were likely more knowledgeable than the average consumer, a class that Vox’s own website defined broadly by targeting the general public for domain sales. The court closed the point with a line that now anchors many an examiner’s brief: where the Board has reasonably weighed the evidence, it is not the role of the Federal Circuit to reweigh it and reach a different conclusion.

Notably, the Supreme Court’s decision in United States Patent & Trademark Office v. Booking.com B.V., 140 S. Ct. 2298 (2020), which rejected any per se rule that generic.com names are generic, framed rather than rescued Vox’s case. Both decisions make consumer perception the touchstone; in Booking.com the perception evidence favored the applicant, and in Vox Populi it did not.

Pixels are not enough

The stylization holding is the opinion’s most practical contribution. Design or stylization may make otherwise unregistrable matter registrable if the features create an impression on purchasers separate and apart from the impression made by the words themselves, a standard the court drew from In re Cordua Restaurants, Inc., 823 F.3d 594 (Fed. Cir. 2016), and the Board’s Sadoru line of cases. That evaluation is necessarily subjective and turns on a viewer’s first impression. The case law illustrates both sides of the line: the Board found a tubelike rendition of the letter C gave CONSTRUCT-A-CLOSET a striking commercial impression in In re Clutter Control, 231 U.S.P.Q. 588 (T.T.A.B. 1986), while the Federal Circuit itself held insufficiently distinctive lettering unregistrable for the Bundt logotype in In re Northland Aluminum Products, Inc., 777 F.2d 1556 (Fed. Cir. 1985).

The .SUCKS pixel font fell on the Bundt side. Every character in the mark was the same height and width, displayed in a font style once mandated by the technological limitations of computer screens. Given the ubiquity of that design in the early days of computing, the Board found consumers would view pixelated lettering as ordinary, and Vox’s customer declarations did not even mention the stylization, which was the only relevant inquiry left. The court also flagged the road not taken: because .SUCKS was not held generic, a showing that the particular display had acquired distinctiveness could in principle support registration, but Vox made no such claim on this record.

Open questions

Vox Populi left the outer boundaries of failure to function unsettled. The court expressly noted it has had limited occasion to address the source-identifier requirement, and it did not define how the doctrine applies outside the domain-name setting, leaving the Board’s expansive informational-matter jurisprudence, from viral slogans to hashtags, without direct appellate ratification. It also declined to decide whether administrative preclusion barred Vox’s arguments about the unappealed standard character refusal, flagging but not resolving how preclusion doctrines operate across serial ex parte applications. Finally, the opinion gestures at, without quantifying, what evidence would suffice for a registry to prove that consumers see its gTLD as a brand; a survey-supported record could present the question Vox Populi never had to answer.

Implications for brands and businesses

Audit your specimens before you file. Vox lost in large part on its own materials, which sold .SUCKS domains as a product. If your website, packaging, and advertising use the applied-for matter as a thing you sell rather than a sign of who you are, the examining attorney will read them exactly that way.
Advertising spend is not perception evidence. Investment figures and growth metrics show intent and success, not how consumers interpret the matter. Declarations from sophisticated trade partners carry limited weight when the relevant public includes ordinary consumers; surveys and consumer-facing evidence matter more.
Do not count on stylization to rescue weak wording. A font, even a distinctive-feeling one, must create its own commercial impression apart from the words. Commonplace styles, and especially styles with nostalgic ubiquity, will be treated as ordinary. Where a display is genuinely distinctive, consider pleading acquired distinctiveness in the display itself.
Expect failure-to-function scrutiny in every nontraditional filing. Hashtags, gTLDs, catchphrases, and informational slogans all draw the refusal routinely, and Vox Populi gives examiners a deferential standard of review. Building consumer-perception evidence from the start is cheaper than an appeal.

Frequently asked questions

What does failure to function as a mark actually mean? It means the applied-for matter, as consumers actually encounter it, does not do the one job the Lanham Act requires: identifying and distinguishing a single source. The refusal is distinct from descriptiveness or genericness. Informational slogans, widely used phrases, and, as in Vox Populi, matter perceived as a mere component of a domain name can all fail to function even if they are not the generic name of anything.

Can a top-level domain ever be registered as a trademark? The Federal Circuit announced no per se bar. Consumer perception controls, the same principle the Supreme Court applied to generic.com names in USPTO v. Booking.com. A registry operator that can show consumers perceive its gTLD as identifying the source of registry services, or that a particular stylized display has acquired distinctiveness, could still succeed. Vox simply had no such evidence, and its own specimens used .SUCKS as a product, not a brand.

Why did the pixelated font not save the application? Stylization rescues unregistrable wording only if the design creates a commercial impression separate and apart from the words themselves. The Board found, and the court agreed, that every character in the mark was the same height and width in a font once dictated by the technological limits of early computer screens, so consumers would view the pixelated lettering as ordinary rather than source-identifying, and Vox never claimed acquired distinctiveness in the display.