Elliott v. Google: Why Verbing a Brand Does Not Kill the Trademark
The Ninth Circuit held that verb use of google does not prove genericide because primary significance to consumers, not grammar, controls under the Lanham Act.
The distinctiveness spectrum is the scale U.S. courts use to decide whether a name can function as a trademark at all. It sorts terms into five categories, from generic to descriptive to suggestive to arbitrary to fanciful, and where your name lands determines whether you own it immediately, own it only after years of proof, or can never own it.
The framework comes from Judge Henry Friendly’s opinion in Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976), which is why practitioners call it the Abercrombie spectrum. The Supreme Court adopted it in Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992), and it now governs registration at the USPTO and infringement in every circuit.
The categories are not about how clever a name is. They describe the relationship between the word and the product it sits on. The same word moves categories depending on what it sells.
Notice that Apple is arbitrary for computers and generic for fruit. The word is not strong or weak in the abstract. It is strong or weak against a specific set of goods.
This is where nearly every real dispute lives, because the consequence is enormous. A suggestive mark registers on the Principal Register today; a descriptive one waits, sometimes forever.
Courts apply two workhorse tests. The imagination test asks how much thought a consumer needs to get from the term to the product. If the connection is instant, the mark is descriptive. The competitors’ needs test asks whether other sellers in the market would legitimately need that word to describe their own goods. The Fifth Circuit’s opinion in Zatarain’s, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786 (5th Cir. 1983), ran both tests and split the difference: “Fish-Fri” was descriptive of a batter mix and survived only because Zatarain’s proved secondary meaning, while “Chick-Fri” failed for lack of it.
The practical lesson runs against most founders’ instincts. The name that explains what you do is the name you are least likely to own. Marketing pressure pushes toward clarity, and trademark law penalizes exactly that.
A descriptive term becomes protectable when the public stops hearing a description and starts hearing a brand. That shift is called secondary meaning, or acquired distinctiveness, and it is codified at 15 U.S.C. § 1052(f).
The evidence courts weigh includes consumer survey results, the amount and manner of advertising, sales volume, the length and exclusivity of use, unsolicited media coverage, and proof that a competitor deliberately copied the term. Surveys carry the most weight because they measure the thing the doctrine actually cares about: what buyers believe.
Section 1052(f) lets the USPTO accept five years of substantially exclusive and continuous use as prima facie evidence of acquired distinctiveness. That is a convenience, not a right. The examiner can demand more where the term is highly descriptive, and five years of use by a company nobody has heard of proves very little.
Two structural points are worth holding onto. First, product design trade dress can never be inherently distinctive; under Wal-Mart Stores, Inc. v. Samara Brothers, Inc., 529 U.S. 205 (2000), it always requires secondary meaning. Second, a descriptive mark that is not yet there can sit on the Supplemental Register, which gives no presumption of validity but does allow use of the ® symbol and blocks later confusingly similar applications while the mark builds recognition.
The spectrum runs both directions. A mark can climb it through secondary meaning, and it can fall off the bottom when the public adopts it as the name of the category. That process is called genericide, and the Lanham Act writes it in: under 15 U.S.C. § 1064(3), a registration can be cancelled at any time if the mark becomes generic, and 15 U.S.C. § 1127 provides that a mark is abandoned when it loses significance as an indication of source.
The graveyard is well populated. Aspirin fell in Bayer Co. v. United Drug Co., 272 F. 505 (S.D.N.Y. 1921), where Judge Learned Hand distinguished what pharmacists understood from what ordinary consumers understood. Escalator fell before the Patent Office in 1950 in Haughton Elevator Co. v. Seeberger, 85 U.S.P.Q. 80 (Comm’r Pat. 1950), partly because Otis had used the word generically in its own advertising. Thermos fell in King-Seeley Thermos Co. v. Aladdin Industries, Inc., 321 F.2d 577 (2d Cir. 1963). Cellophane, yo-yo, and trampoline are all former trademarks.
Not every widely used mark is doomed. In Elliott v. Google, Inc., 860 F.3d 1151 (9th Cir. 2017), the Ninth Circuit held that “google” survived even though people say it as a verb, because the test is who-are-you/what-are-you applied to a particular class of goods: consumers still identified Google as the source of a search engine, and verb usage alone did not prove otherwise. That is why Xerox and Adobe run advertisements begging you to say “photocopy” and “edit.”
For years the USPTO applied something close to a rule that a generic term plus a top-level domain equals a generic composite. The Supreme Court rejected it in United States Patent and Trademark Office v. Booking.com B.V., 591 U.S. 549 (2020), by an 8-1 vote.
The reasoning was straightforward and consumer-facing. Whether a term is generic depends on its meaning to consumers, not on a mechanical rule, and the record included survey evidence that a large share of consumers understood “Booking.com” as a brand rather than as a category of online hotel reservation services. Because only one entity can hold a given domain at a time, a generic.com conveys source in a way that a bare generic word cannot.
Justice Ginsburg’s opinion carried a warning that gets quoted less than the holding. Such marks are weak. Their scope is narrow, competitors keep the descriptive fair use defense under 15 U.S.C. § 1115(b)(4), and consumers are unlikely to be confused by similar names. Booking.com won registration and won very little exclusion.
What is the trademark distinctiveness spectrum? It is a five-category scale from Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976), that sorts marks by how well they identify a single source. From weakest to strongest the categories are generic, descriptive, suggestive, arbitrary, and fanciful. Generic terms can never be trademarks, descriptive terms qualify only after acquiring secondary meaning, and suggestive, arbitrary, and fanciful marks are inherently distinctive and protectable from first use.
What is the difference between a descriptive and a suggestive mark? A descriptive mark conveys an ingredient, quality, or characteristic of the goods immediately, with no mental step. A suggestive mark requires imagination to connect the word to the product. Courts often use the imagination test and the competitors’ needs test. Under 15 U.S.C. § 1052(e) a descriptive mark is refused registration unless it has acquired secondary meaning; a suggestive mark registers on its own.
What is secondary meaning in trademark law? Secondary meaning, also called acquired distinctiveness, exists when the buying public has come to associate a descriptive term with a single source rather than with the product category. Evidence includes consumer surveys, advertising spend, sales volume, length of exclusive use, and intentional copying. Under 15 U.S.C. § 1052(f) five years of substantially exclusive and continuous use can serve as prima facie evidence, though it is not automatic.
Can a generic.com domain be a trademark? Sometimes. In USPTO v. Booking.com B.V., 591 U.S. 549 (2020), the Supreme Court rejected a per se rule that adding .com to a generic term always yields a generic composite. The question is whether consumers actually perceive the term as a brand, which is proven with evidence such as surveys. The Court cautioned that such marks are weak and narrow in scope.
Going further: How to Trademark Your Business, step by step .
This page is general legal information, not legal advice, and it does not create an attorney-client relationship.
The Ninth Circuit held that verb use of google does not prove genericide because primary significance to consumers, not grammar, controls under the Lanham Act.
Zatarain's FISH-FRI appeal gave trademark law its four descriptiveness tests and confirmed that competitors' fair use survives secondary meaning.
The Supreme Court rejected a categorical rule that 'generic.com' terms are unregistrable, holding that consumer perception alone determines whether such a composite is generic, reshaping distinctiveness analysis for the domain-name economy.
How the Supreme Court let Kellogg use the generic name and functional pillow shape of shredded wheat, anchoring the rule that expired patents and generic terms pass into the public domain.
The Supreme Court's unanimous 2000 decision held that a product's design can qualify as protectable trade dress only on proof of secondary meaning, and told courts to classify ambiguous cases as design, drawing the line that *Two Pesos* had left open.
The Supreme Court held that Coca-Cola's name had come to mean a single product from a single source, and rejected the claim that its history barred trademark relief.
Learned Hand held that a trademark's validity turns on what buyers understand the word to mean. To consumers, "Aspirin" meant the drug itself, not Bayer.
The Supreme Court's 1992 decision held that inherently distinctive trade dress is protectable under Section 43(a) without proof of secondary meaning, extending to a restaurant's look the same first-day protection long given to coined word marks.
The 1976 Second Circuit decision that sorted every word mark into fanciful, arbitrary, suggestive, descriptive, or generic (and cost Abercrombie its oldest 'Safari' registration and its whole infringement case) still governs how courts measure distinctiveness today.