Trademark Failure to Function

I.     Introduction

In 2006, a company called ChaCha Search launched a new smartphone app that provided search engine access via text message. Users were invited to text queries to the number “242242,” the numeric equivalent of typing “ChaCha” on their phones’ keyboards, and await results. After the service went live, ChaCha applied to register “242242” as a service mark. The U.S. Patent & Trademark Office (“USPTO”) reviewed the company’s specimen of use—a screenshot from the Beta version of the app (shown below)—and granted registration.1

 Image 1 Roberts


Less than two years later, ChaCha found itself embroiled in a dispute with competitor Grape Technology (Grape). Grape sought to cancel ChaCha’s registration on the basis that “242242” lacked distinctiveness.2

Distinctiveness is the primary protectability hurdle for most trademarks, so it was a logical basis for Grape to select in challenging ChaCha’s registration. The USPTO and courts assess the inherent distinctiveness of a text mark by considering the mark in relation to the goods or services with which it’s used3 and then placing it in one of five categories: the mark is classified as fanciful, arbitrary, suggestive, descriptive, or generic.4 If the term or phrase is neither descriptive nor generic, and is not barred by any of the other restrictions enumerated in the Lanham Act,5 it typically receives protection from its earliest use in interstate commerce.6 Trademark law presumes consumers will perceive it as a mark.

In response to Grape’s challenge, the Trademark Trial and Appeal Board (“TTAB”) reassessed the mark’s distinctiveness and let ChaCha’s registration stand. The Board concluded that “242242” was not merely descriptive for the services in question because the number “d[id] not identify an ingredient, quality, characteristic, function, feature, purpose, or use of the specified services.”7

But there’s another argument that the Board declined to consider.8 Under the Lanham Act and common law, matter used in commerce is only protectable if it is both distinctive and used as a mark. 9 Its success as a mark depends upon how it is used not just semantically, but aesthetically. While courts pay little attention to the latter requirement, and the USPTO applies it somewhat inconsistently, use as a mark plays a crucial role—along with distinctiveness—in predicting whether or not the public will perceive matter as indicating source. To be protectable, a trademark must be not only used in commerce,10 but used in a trademark way:11 It must appear where consumers expect a trademark to appear, and it must be sufficiently set off from the surrounding text and images to attract notice.12 If it isn’t, it will fail to function as a trademark to consumers, and thus shouldn’t merit protection under federal trademark law.13 This Article explains why use as a mark is crucial to trademark protection and advocates for combining use as a mark analyses with distinctiveness assessments to better serve the goals of trademark law and avoid outcomes in which matter that fails to function—like 242242—is nonetheless granted trademark protection.

Without use as a mark, there can be no trademark, and consequently no trademark rights.14 Yet, distinctiveness has received the lion’s share of attention, generating rules and tests applied in thousands of cases and USPTO decisions and discussed in hundreds of articles, books, and practice guides.15 A term’s inherent qualities are often treated as the sole predictor of whether consumers will understand it as a mark. At the same time, the USPTO and federal courts have struggled to articulate and apply a clear standard for use as a mark and endeavored to separate it from distinctiveness. When a mark does not actually indicate source or distinguish goods or services because of the way in which it is used, the USPTO refuses to register it based on its “failure to function.”16 Use as a mark and distinctiveness are two sides of the same coin.17 Each is necessary for trademark protection, because each plays a role in determining whether consumers will perceive matter as a mark. So how did use as a mark become the red-headed stepchild of trademark law, while distinctiveness got all the glory?

Distinctiveness doctrine, made seemingly straightforward by Abercrombie18 and its progeny, is one of trademark law’s stickiest stories.19 Prior to the Lanham Act, in the late 1800s and early 1900s, courts protected against infringement only “technical trademarks”—those we now call inherently distinctive—and required that they be physically affixed to products.20 Descriptive words and phrases were deemed “trade names” and received some analogous protection if they acquired secondary meaning based on an unfair competition theory, rather than a trademark infringement theory.21 The early common law conception of “distinctiveness” was holistic—it asked whether a mark truly functioned to distinguish goods in the marketplace.22 But as the definition of “trademark” became more expansive and courts developed rubrics to evaluate distinctiveness, the concept of distinctiveness itself became impoverished; a test for distinctiveness replaced the idea of distinctiveness. Under the cases that laid the groundwork for Abercrombie, distinctiveness became merely a question of into which bucket to place a mark. This Article argues that predicting whether matter will be perceived as a mark requires considering not just what it is, but—at the same time—how it is used. One way to frame this idea is to advocate enriching the concept of distinctiveness by injecting use back into it; another is to advocate layering distinctiveness, use in commerce, and use as a mark together to evaluate protectability.

In focusing on static, inherent characteristics to determine whether something qualifies as a trademark, Abercrombie leads judges and USPTO-examining attorneys to downplay or ignore the other half of the puzzle: namely, context. While Abercrombie seeks to gauge the conceptual relationship between a mark and the goods or services with which it is used, use as a mark considers the visual relationship between them.23 Studies show that context and manner of use, in combination with inherent characteristics of the mark, determine whether or not consumers will understand matter as a trademark when they first encounter it.24 Federal courts often ignore use as a mark altogether, though, even when assessing common law rights.

The threshold use as a mark requirement has received relatively little scholarly attention,25 as has trademark registration itself.26 But use as a mark is a fundamental requirement to establishing trademark rights, without which foundation the entire edifice crumbles.27 As such, use is crucial not only in federal registrability assessments, but also in infringement,28 dilution, priority disputes,29 and every other trademark cause of action that requires a plaintiff to establish rights in a mark as a prerequisite to relief.

Failure to function arises most often at the registration stage. And despite the dearth of discussion about it, registration matters. It matters because federal registrations cast tremendous shadows, tacitly enabling owners of registered marks to silence or bully30 competitors, new entrants, and entrepreneurs who lack the resources to fight back.31 It matters because registration instantly converts non-use or borderline use into a robust presumption of rights that is incredibly difficult to unravel. It matters because the mere existence of a registration affects which marks newcomers select; registrations that overprotect descriptive or thinly suggestive matter, in particular, burden a new entrant’s ability to communicate with consumers in the marketplace.32 And registration matters because by the time a mark becomes entangled in an infringement case, questions that arise early in the mark’s life, especially use as a mark and inherent distinctiveness, appear settled—the presumption of validity and the duration of the marks’ use often lead courts to treat those questions as resolved.

In a 2017 article in the Harvard Law Review, Rebecca Tushnet recommends paying the registration process “renewed attention” as an opportunity to add necessary structure and increase consistency within and across federal trademark law.33 While the USPTO regards registration as substantive, Tushnet argues, courts often treat it as merely procedural. Her hypothesis suggests courts both undervalue the work of examining attorneys and the role of registration and, at the same time, overinflate the importance of the result of that process. By ignoring the details and treating registration as binary, courts often “round up,” assigning broad, robust protection to matter that might have gained registration based on a much narrower interpretation of the mark’s scope and the mark owner’s rights. The possibility of a registered mark achieving incontestable status further intensifies this effect.34

And the USPTO is bound to make occasional mistakes, given asymmetric incentives,35 the lack of concrete guidance examining attorneys receive on issues such as use as a mark, and those attorneys’ high-volume workload and individualized approaches. When the USPTO makes mistakes, the presumption of validity afforded a registered mark compounds courts’ reluctance to reopen basic questions like use and distinctiveness, which it presumes the USPTO investigated sufficiently at the registration stage. Agency decisions about use can also have binding consequences for litigants: The Supreme Court recently made clear that TTAB decisions may ground issue preclusion.36

Registration matters because it affects litigation outcomes and delineates the contours of rights for mark owners, existing competitors, new entrants, and noncommercial speakers. And use as a mark determinations lie at the heart of registrability. Use as a mark arguably matters even more in litigation, because when a mark owner sues for infringement, its marks have typically been in use longer, so the type of use being made has crystallized. Courts can review evidence representing the full spectrum of a plaintiff’s trademark use and should invalidate asserted trademarks if that evidence does not reflect use as a mark. Whether or not use as a mark is outcome-determinative in a particular case, how courts handle it has substantial downstream effects. That’s because the rules and guidelines that USPTO-examining attorneys and TTAB judges apply in assessing registrability draw heavily from case law. When courts ignore the use question, their indifference reinforces the idea that use doesn’t matter. The stakes are high, because granting federal protection for ineligible matter chills speech and competition.

Yet despite those stakes, Abercrombie continues to dominate assessments of trademark protectability and predictions about consumer perception, even in the face of empirical research revealing the importance of context and use. As scholars have noted, judges often rely on psychological theories in precedential case law in ways that lead those theories to “remain embedded . . . long after they have been disconfirmed or superseded by advances in the empirical social sciences.”37 Distinctiveness doctrine isn’t wrong, but as an account of what makes matter protectable, it’s incomplete.38 Abercrombie has dominated trademark law because it acquired the weight of federal circuit court authority and rendered USPTO and judicial assessment of marks straightforward and simple. And while empirical data reflects the importance of use as a mark in a way that maps on to the Lanham Act and case law’s definitions of what a trademark is, the law has been reluctant to make protectability,39 validity, and priority analyses more complex by assigning that data any real weight.

Despite the role of context in consumer perception, the affirmative use as a mark requirement has not been the subject of sustained critical attention. This Article explores how judges and USPTO examining attorneys have considered (or failed to consider) use as a mark in both ex parte and adversarial contexts and how they can make those assessments more robust. Ultimately, it concludes that the relationship between use as a mark and distinctiveness is interdependent and inverse: the less distinctive a mark is, the greater indicators of trademark use are needed to ensure consumers will perceive it as a mark, and vice-versa. Use as a mark is thus particularly outcome-determinative when it comes to marks whose status falls on the border between inherently distinctive40 and merely descriptive, as well as those marks that primarily play some other non-trademark role, such as marks that serve as domain names, hashtags, celebrity nicknames, or ornamental matter. The below chart provides a visual illustration of the relationship between use as a mark and distinctiveness41:

Image 2 Roberts

Part II of this Article parses the statutory definition of “trademark” with a focus on use as a mark, making a case for the doctrine’s importance and highlighting the lack of clear and consistent guidelines for factfinders in every procedural context. It explicates what use as a mark, and conversely failure to function, actually look like in practice, underlining how trademark presentation and context can affect perception and protectability. Part III presents empirical data documenting the influence of trademark use on consumer perception. Part IV focuses on the intersection of use and distinctiveness in determining whether matter qualifies for trademark protection. While factfinders often assess one without the other, the more comprehensive approach incorporates each consideration in a way that reflects the relationship between them. Part V thus acknowledges that inherent distinctiveness and use as a mark are interrelated and must be considered together. It proposes taking a cue from nonverbal trademark and trade dress analyses to ensure that courts and the USPTO accurately gauge the answer to the question both doctrines ostensibly seek to resolve: whether consumers are likely to perceive claimed matter as a mark.

II.     Use as a Mark

This Part unpacks the derivation and current status of the threshold use as a mark requirement and illustrates its importance. It offers concrete criteria for evaluating use as a mark, along with examples in which those criteria are applied correctly and others in which they are ignored.

A.     What Use as a Mark Looks Like

The Court of Customs & Patent Appeals, the Federal Circuit’s predecessor court, seemed to state the obvious in Standard Oil and again in In re Bose :

The Trademark Act is not an act to register mere words, but rather to register trademarks. Before there can be registration, there must be a trademark, and unless words have been so used they cannot qualify. Thus, it is essential for registration that words first become a trademark.42

And as myriad other courts have affirmed, that “words have been so used” is a prerequisite not only for registration,43 but also for common law rights.44 That use requirement sets the United States’ federal trademark registration regime apart from those of almost every other country.45 And while U.S. producers can apply to register trademarks based on an intent to use them in the future (“ITU”) and thus stake out priority,46 the registration itself will not issue until an applicant demonstrates actual use.47 But factfinders lack clear-cut guidance for determining when matter is used as a mark.

What does use as a mark look like in practice? Classic use displays a word mark in large, clear, prominent lettering, often using capital letters or font that differs from that used in the surrounding text.48 It may employ color to draw attention to the mark, or frame the mark with an oval or rectangle to offset it from the rest of the packaging.49 COLD-EEZE, SUAVE, and the stylized DUNKIN’ DONUTS mark below are examples. On packaging or signage, the mark should be centrally located and may be followed by a TM, SM, or ® symbol. When consumers regard the coffee cup below, we can expect them to perceive the word mark DUNKIN’ DONUTS, the colorful logo, and the slogan AMERICA RUNS ON DUNKIN’™ as marks.50

 Image 3 Roberts


Other types of use may employ fewer indicators, but enough for consumers to perceive them as trademarks, like “Bumble and bumble” on the container below.

 Image 4 RobertsUse that does not qualify for protection typically lacks those indicia of trademark use. Instead, the matter might function primarily as a domain name, like “,” or informational or descriptive text, like “caution: this beverage is extremely hot”51 on the coffee cup, “Family Size!” on the shampoo bottle, or “All Natural Cherry Flavor” on the cold remedy package.


In modern law, the use as a mark requirement draws its power from the Lanham Act’s definition of the terms “trademark” and “service mark,”52 although the requirement predates the statute53:

The term “trademark” includes any word, name, symbol, or device, or any combination thereof . . . used by a person . . . to identify and distinguish his or her goods, including a unique product, from those manufactured or sold by others and to indicate the source of the goods, even if that source is unknown.54

In order to gain protection, then, the matter must fulfill a trademark’s most basic function—it must identify and distinguish an entity’s products from those of other entities and indicate the products’ source.55 Like the statute, the USPTO’s Trademark Manual of Examining Procedure (“TMEP”)56 emphasizes identification, differential distinctiveness, and source distinctiveness57 in delineating what a trademark must do.58

Courts, the USPTO, scholars, and treatises have articulated the use as a mark requirement in a number of ways, none of which can be said to provide a clear roadmap to factfinders or mark owners.59 Some courts frame the salient question as whether the matter is “used in such a way as to make such a visual impression that the viewer would see it as a symbol of origin separate and apart from everything else?”60 Use of a mark must be “of such nature and extent as to create an association of the goods or services and the mark with the user thereof.”61 The TTAB has asked “whether the designation in question, as used, will be recognized in and of itself as an indication of origin for this particular product.”62 Matter “must be used in a manner calculated to project to purchasers or potential purchasers a single source or origin for the goods.”63 Margreth Barrett construes the common law and Lanham Act generally to require the “application of a mark in a manner that invites consumers to associate the mark with goods or services that the user is offering for sale or distribution and to rely on it for information about the source, sponsorship, or affiliation of those goods or services.”64 Per Mark McKenna, “for ‘use’ of a term to trigger substantive trademark rights, a party must use the term in such a manner that consumers regard it as indicative of the source of that party’s goods or services.”65 McCarthy on Trademarks, a leading trademark treatise, explores the topic at length and then summarizes tautologically: “To be a trademark, a designation must do the job of a trademark.”66

The type of use that qualifies as a trademark use,67 then, is somewhat difficult to pin down—the statute, USPTO, courts, treatises,68 and experts recite requirements that are similar but not the same. The use as a mark inquiry also depends upon a determination of whether a mark is used to indicate “source,” where source itself is a murky idea. Mark McKenna has called source “ill-defined” and “an extraordinarily vague concept capable of encompassing almost any imaginable relationship between parties.”69 Barton Beebe refers to source as “a legal fiction.”70 And as the Lanham Act has acknowledged since 1984, consumers need not—and often cannot—identify the precise company that manufactures particular goods.71 It is enough for consumers to understand a trademark as indicating consistent source in a general sense, even when the mark does not literally tell consumers who makes or stands behind the product. This idea has come to be known as anonymous source doctrine.72

As the TTAB has noted, “the critical enquiry in determining whether a designation functions as a mark is how the designation would be perceived by the relevant public.”73 And in order to satisfy the general use in commerce requirement, as well as the specific use as a mark requirement, the use must be consumer-facing, i.e. public rather than private or internal.74 Use that is merely internal75 or preparatory76 will not constitute use as a mark.77 Courts focus on whether trademark use is “sufficiently public to identify or distinguish the marked goods in an appropriate segment of the public mind as those of the adopter of the mark.”78 That requirement logically follows from the Lanham Act’s definition of a trademark: identifying goods, distinguishing them from others’ goods, and indicating the source of goods all turn on whether consumers perceive the matter to perform those functions.

In articulating what use as a mark and failure to function look like, we can also turn to two fair use defenses: statutory fair use and nominative fair use. The first defense is available to defendants that simply use descriptive terms or phrases as regular words, rather than as trademarks. For example, the marketing text “Crest is the dentist’s choice for fighting cavities” was held a fair use that did not infringe the trademark DENTIST’S CHOICE for toothbrushes, because the phrase was used descriptively, fairly and in good faith, and other than as a mark. Likewise, Ocean Spray could use the phrase “sweet-tart” in connection with its juice without infringing plaintiff’s mark SWEETARTS for candy, because the phrase functioned only as a descriptor and not as a mark. The statutory fair use line of cases can help center the discussion about failure to function. According to the doctrine, even matter affixed to or used in connection with goods and services can be used in a non-trademark way—in other words, not all use is use as a mark.

Conversely, the nominative fair use defense is applicable where a defendant uses a plaintiff’s mark to refer back to the plaintiff’s goods or services, and not to the defendant’s, as when a seller of headphones describes its headphones as “compatible with all Apple products” or a jeweler advertises “Don’t pay Tiffany prices—Diamond Exchange has diamonds for less!” In such cases, the use will be allowed as long as the defendant uses only so much of the plaintiff’s mark as is reasonably necessary to identify the product and does nothing to suggest sponsorship or endorsement. In practice, this typically means using plaintiff’s mark in plain text, using the same size, color, and font as surrounding text, and using merely the word and not the stylized word mark or logo. If, in the example above, the jeweler used the registered Tiffany logo and signature Tiffany blue color—advertising 

“Don’t pay Image 5 Roberts prices—Diamond Exchange has diamonds for less!” a court would likely find the elements of the nominative fair use defense were not satisfied.

Both defenses’ limitations on use and the cases applying them therefore provide a helpful foil for assessing a putative mark owner’s use. The kind of non-use that qualifies a defendant for either fair use defense would also likely fail to qualify a mark for affirmative protection; conversely, use as a mark that qualifies for protection would likely exceed the boundaries of fair use when made by a defendant.

B.     When Use as a Mark Matters

Failure to function79 arises most often, and in the most straightforward way, in connection with registration.80 Mark owners who seek federal registration of their marks do so ex parte: unless another party opposes registration after the USPTO has preliminarily approved it, the applicant has no adversary. As a component of registrability decisions, ex parte use as a mark determinations are considered by, in order of decreasing frequency, the USPTO; the Federal Circuit; and other federal courts.

Failure to function also arises in several adversarial contexts, including opposition or cancellation proceedings or priority disputes81 before the TTAB, cancellation claims or priority disputes in federal court, and threshold validity determinations at the outset of federal infringement litigation. Courts may be more likely than the USPTO to overlook use as a mark in adversarial postures for a variety of reasons, but as this Article argues throughout, use as a mark plays a crucial role in the acquisition of rights and ought to be given due weight at each stage of a dispute.

Federal registration begins with an application to the USPTO. Once an applicant files a use-based application, a trademark-examining attorney reviews the mark, description of goods and services, specimens of use, and any other information pertinent to an initial decision.82 The examining attorney may approve the mark for publication, after which it will eventually become registered barring opposition, or she may issue an office action to preliminarily refuse the application on any of a number of different bases, including failure to function or lack of distinctiveness.83 The applicant has an opportunity to respond to the office action and address the USPTO’s concerns, including amending the specification of use, filing different specimens, or arguing persuasively that the stated bases for refusal do not apply.84

If the USPTO issues a final refusal denying registration for a mark, the applicant can appeal that decision to the TTAB and then to the Court of Appeals for the Federal Circuit (“CAFC”) or to a federal district court.85 From there, appeals follow the ordinary course.

Failure to function refusals are most common at the initial registration stage.86 Owners filing applications with the USPTO are often less experienced and more likely to commit basic errors than are those who pursue appeals or litigation. Some applicants do not understand the basic requirements for registration, and barriers to entry are minimal, including the low cost of registration compared to litigation and the option of registering a mark pro se.87 Given how often trademark-examining attorneys assess use as a mark, their expertise ought to be substantial; in addition, the TMEP offers explicit guidance at the examination stage.88

Yet mistakes are common—applications to register matter that isn’t being used as a mark are often published with no fanfare.89 The TMEP’s guidance on use as a mark is less extensive and less concrete than its guidance on other doctrines, such as distinctiveness, false association, and likelihood of confusion, for which the Manual articulates explicit, often multi-factor tests that examining attorneys can apply to determine protectability.90 As scholars have observed in the patent context,91 the asymmetric nature of the trademark application process creates incentives for the USPTO to be permissive in granting registrations92 in order to minimize appeals and reversals.93 Indeed, more than three quarters of use-based and ITU applications are published.94

Further complicating the USPTO’s assessment of use as a mark is the option, available since 1989, for entities to file applications to register trademarks based on a bona fide intent to use them in the future. ITU applications now consistently outnumber use-based applications.95 When USPTO attorneys examine a use-based application, they review the applicant’s specimen, desired trademark, and description of goods or services at the same time, which enables them to consider whether the specimen reflects use as a mark. But when they examine ITU-based applications, they receive only the trademark and the description of goods and services, devoid of context; they can assess inherent distinctiveness, but not use as a mark. At some point after the application is examined and published, the producer will begin to use the mark in commerce and then submit a statement of use and specimens to perfect the registration. In theory, the USPTO must then confirm that the mark is acceptable in light of the use shown;96 in practice, examination at that stage is cursory97 and some trademark prosecutors report that it may be performed by a paralegal rather than an attorney. The rise of ITU applications, then, exacerbates the lack of attention paid to the question of whether an applied-for mark is indeed being used in a trademark way in compliance with statutory and common law requirements.98

If the examining attorney determines that an applicant’s specimen fails to demonstrate use, she will issue a failure to function refusal “on the ground that the applied-for mark does not function as a trademark,”99 citing Lanham Act sections 1, 2, and 45.100 The matter might be said to fail to function as a trademark because it functions primarily as something else: a domain name;101 a hashtag;102 a phone number;103 a product feature;104 a description;105 a decoration.106 TMEP section 904.07(b) offers a non-exhaustive list of categories of use that tend to indicate failure to function.107 But most trademarks are capable of doing both,108 and they frequently do—a mark can be a descriptive source indicator like WEIGHT WATCHERS, or a decorative source indicator like the Nike “swoosh,” or a domain name source indicator like AMAZON.COM.109 So the emphasis for use as a mark must be on the matter’s failure to function as a source indicator—its failure to do the job of a trademark—rather than its simultaneous role as something else.

C.     Failure to Function

Without proper use, there can be no trademark rights.110 And yet despite the centrality of the use question, courts rarely devote significant attention to it, and none has articulated a clear test. Perhaps that neglect results from the expectation that whether matter functions as a trademark must be “readily apparent and recognizable without extended analysis or research, and certainly without legal opinion.”111According to McCarthy, “[u]sually, if when viewed in context, it is not immediately obvious that a certain designation is being used as an indication of origin, then it probably is not. In that case, it is not a trademark.”112 Consumers spend very little time and mental energy examining tags and labels, so if matter is to be protected “it should be obvious at first glance” that the matter is being used as a trademark.113 Courts expect a gestalt: consumers will not be “required or expected to browse through a group of words, or scan an entire page in order to decide that a particular word, separated from its context, may or may not be intended, or may or may not serve to identify the product.”114When the USPTO and courts grant protection for text-based marks that fail to function, they typically do so not because use as a mark is too difficult to evaluate, but rather because it’s ostensibly so obvious that it warrants no special attention, so they simply skip over it.

The cases and Board decisions assessing use as a mark demonstrate that context is king.115 As McCarthy highlights, common indicators for word marks include:

  • large font relative to surrounding text;
  • all capital letters or initial capitals;
  • distinctive print style;
  • use of color; and
  • prominent position on the label or advertisement.116

Implicit in McCarthy’s list is the idea that protectable matter must stand out from the text around it, rather than blending into the background as part of a longer phrase or sentence. For example, the TTAB held CONDOMTOY CONDOM failed to function as a mark for condoms where it appeared only as part of the sentence “That’s why Inspiral is also called a CondomToy® condom.”117 SPECTRUM failed to function as a mark for dimmable switches where it appeared as part of the longer sentence, “Now with Voltage-Controlled Dimming and a Spectrum of Sunlight Readable Colors.”118 And WHERE CONTENT REIGNS failed to function as a service mark for audio books when the phrase appeared in the specimen shown below merely “as part of a sentence” and therefore “would not be perceived by purchasers as a source indicator.”119

 Image 6 Roberts

In assessing word marks as well as nonverbal marks, factfinders often consider whether the matter is set off by itself.120 A district court held that plaintiffs who had expressed their intent to make a movie about a famous racehorse named “Ruffian” had not used RUFFIAN as a mark for entertainment services where the term was “not distinguishable or set apart in any way from the rest of” a document plaintiffs had submitted as evidence of trademark use.121 The Third Circuit declined to extend protection to plaintiff’s claimed trademark, COCOA BUTTER FORMULA, because it had not been used on its own as a mark, but only as part of the longer phrase “Palmer’s Cocoa Butter Formula.”122 And a district court found SAM’S RIVERSIDE was not used as a mark despite plaintiff alleging forty years of continuous use,123 because the phrase was never set apart from the longer phrase “Sam’s Riverside Truck Parts” using any “distinctive font, color, typeset or any other method.”124

In their empirical study of distinctiveness, Lee, DeRosia, and Christensen highlight the role of context in assessing whether matter functions as a mark.125 They emphasize location, arguing that consumers look to what they call “the trademark spot”126 and expect to see matter used as a mark there.127 The authors draw support from consumer psychology’s theories of schema-directed perception to ground trademark law’s concepts of use as a mark and distinctiveness:

After observing whatever word/symbol is found in the “trademark spot,” the consumer will compare it with his memory schema for trademarks (brands). If what he observes is congruent with his schema for brands, then he will esteem the word/symbol to be a source-identifier. If, on the other hand, the word/symbol is incongruent with the consumer’s schema for brands, then he will reject the word/symbol as the source-identifier and look elsewhere on the package for information to accomplish his perceptual goal. Thus, schema-directed perception suggests that whatever word/symbol has been placed on the label in the “trademark spot” will be seen as a source indicator, with the exception of words/symbols that the consumer must reject as incongruent and inappropriate for trademarks.128

The TTAB seems implicitly to understand and apply the “trademark spot” requirement, as for example in its discussion of placement in its assessment of use as a mark of 1-800-TIRE-911.129 But outside of its guidelines regarding ornamental use,130 the USPTO does not articulate it, nor do federal courts typically acknowledge it.

The “trademark spot” concept is not limited to word marks; it applies to images and other nontraditional trademarks as well. In West-com, for example, the TTAB held a caricature of a toilet used on a button on an electronic device that enabled patients to communicate with hospital personnel failed to function as a trademark for the device itself:

 Image 7 Roberts

The Board cited the examining attorney’s assertion that “given the industry-wide use of small, highly intuitive decorative icons in this field, consumers would be hard-pressed to realize that the pictorial icon on any particular button of a nurse paging system represents the underlying manufacturer.”131 Its reference to industry standards in adjudging the adequacy of use is consistent with case law132 and legislative history.133

The use of a TM symbol134 often weighs toward a finding of use as a mark135 because it communicates to consumers the producer’s intent to use matter as a mark, rather than merely signaling it to the USPTO. Conversely, courts have held that the failure to use a TM symbol, or the use of the TM symbol with matter other than the matter being claimed as a mark, weighs against a finding of trademark use.136 But courts, the USPTO, and practice guides are quick to point out that an applicant’s mere assertion that he is using matter as a mark doesn’t make it so,137 and the use of “TM” or “SM” is not enough to establish use as a mark without more. The TTAB has held that where matter is not in use as a mark and would not be perceived as a mark by the public, “the mere addition of the trademark symbol is not sufficient to transform the non-trademark use into trademark use.”138 

And while distinguishing a mark from surrounding text by using different color and font often suffices to show use, the TTAB affirmed the USPTO’s refusal to register SPECTRUM, even though applicant’s use showed each letter of the word in a different color139:

 Image 8 RobertsThe TTAB iterated the language of the Office Action:

[I]t would be difficult for consumers to determine if the word is intended to be used as a mark, or displayed in a variety of colors as an attention getting means of conveying the multiple color capability of such switches . . . no different from the other informational statements among which it is used.140

The oft-referenced indicators of trademark use, then, are proxies, not requirements.141 Use as a mark does not actually mandate any particular kind of display—font size, stylization, color, capital letters, and prominence are typically treated as signals that matter will likely be perceived as a mark, but they are not prerequisites to a finding of use. Nevertheless, they provide a concrete set of considerations that can guide factfinders to assess use as a mark objectively.

Distinctiveness also interacts with use analyses in ways that can be outcome-determinative. The more distinctive a mark—and that includes both inherent and acquired distinctiveness—the fewer trappings of trademark use above a minimum baseline are necessary for consumers to perceive the matter as a mark. For example, the USPTO initially refused an application to register TRULICITY for pharmaceuticals based on the applicant’s ostensible failure to show use as a mark.142 The Office Action stated, “[t]he proposed mark TRULICITY is in the same size and stylized font as the surrounding wording. Additionally, the term TRULICITY is part of a sentence . . . Furthermore, TRULICITY is not set out from the surrounding text.”143

But the TTAB reversed the USPTO’s refusal, finding that the company’s use on the specimen—a label applied to the pharmaceutical for distribution to clinical trial sites—reflected use as a mark, despite the lack of differentiation of size or stylization or color from surrounding text144:

 Image 9 Roberts

The TTAB decision rested heavily on the finding that the mark was fanciful and thus inherently distinctive for the goods in question. Similar use of a mark deemed merely descriptive or thinly suggestive would have been far less likely to garner protection, indicating that the interplay between distinctiveness and use as a mark plays a crucial role in predicting consumer perception.

D.     Bad Outcomes

The cases and USPTO decisions discussed above introduce use as a mark through examples in which the doctrine is properly applied. But just as often, use as a mark is ignored, leading the USPTO to grant registration for matter that consumers are not likely to understand as source indicators because the matter is not used in a trademark way.145 Indicators of use as a mark include size, font, placement, color, capitals, prominence (including a circle around the text or other basic graphics accompanying it), and use of trademark symbols ™ or ®.146 When matter appears on specimens in very small print; in out-of-the-way places; in the same size, color, and font as the surrounding text; in lowercase letters; and without any background shapes or other features that call attention to it, consumers are unlikely to perceive it as a mark. In those cases, matter should be deemed to fail to function. When the matter also serves as something else—a hashtag, domain name, item number, album title, decoration—indicators of trademark use are especially important. And when the matter is borderline descriptive, they are crucial.

Over the last few years, the USPTO has seen a boom in applications to register hashtags as trademarks (“tagmarks”).147 But the specimens submitted in support of those marks often reflect only use as hashtags, rather than the type of use likely to lead consumers to view those hashtags as trademarks.148 The TMEP includes a section on tagmarks, which explicitly reminds examining attorneys that applications for “such marks must still be evaluated to confirm that they function as source indicators for the goods or services,”149 but many applications seem to slip through the cracks.150

For example, a mezcal maker successfully registered #ShareTheSilence for alcoholic beverages using the following specimen, in which the tagmark appears in tiny font at the bottom of the bottle below the bar code and alongside the QR code151:

 Image 10 Roberts

 Likewise, LexisNexis registered #BeUnprecedented as a trademark for legal research services152 with specimens that reflect none of the usual indicia of trademark use and no reason to expect that consumers would view the matter as anything more than a hashtag.153

 Image 11 Roberts

Similar examples abound, because tagmarks as a category are particularly prone to fail to function as trademarks.154 But examples of specimens that should have prompted refusals by examining attorneys and did not are plentiful across all categories of marks.

For example, the USPTO accepted an application to register CHIPSET-FREE CHARGING accompanied by this specimen, which appears both merely descriptive and unlikely to attract the attention necessary for consumers to perceive it as a mark155:

 Image 12 Roberts

And it accepted the below specimen to support registration of BIG IS THE NEW SMALL for alcohol distribution services, notwithstanding the fact that the phrase is used as part of a sentence in what appears to be an internal slide deck (the specimen is explicitly labeled “Confidential – Do Not Copy or Distribute”)156:

 Image 13 Roberts


In addition to hashtags, SMS numbers, descriptive text, and slogans, a 99-word “manifesto” was recently published as a service mark for educational and vocational services. After the USPTO issued an initial failure to function refusal, it accepted revised specimens, one of which comprised simply the typed text of the mark and the other of which, shown below, featured the mark inside the outline of a vehicle157:

 Image 14 Roberts

And the USPTO accepted specimens showing classic ornamental use of “Honey Badger Don’t Care” in support of five different registrations for the phrase, including this specimen for use in connection with the sale of clothing and caps158:

 Image 15. Roberts

Those registrations then provided the foundation for the mark owner’s infringement suit against a greeting card company for the latter’s non-trademark use; a Ninth Circuit panel relied on the presumption of the marks’ validity that follow from their registration without giving the issue of trademark use any further attention when it reversed the district court’s grant of summary judgment for defendant.159

Bad outcomes like these abound because failure to function issues are so often overlooked.160 The USPTO has little incentive to scrutinize use as a mark and refuse those applications that fail to demonstrate it.161 Examples are hard to track or quantify, because such registrations almost always go unquestioned, so no search of TTAB decisions or case law will round them up in any organized or remotely exhaustive way.162 Only a small percentage of published or registered trademarks are challenged, and almost all of those challenges are based on likelihood of confusion, dilution, false association, disparagement, or other grounds for cancellation that reflect the mark’s potential to harm a challenger. A competitor may occasionally oppose a mark or petition to cancel it based on mere descriptiveness, but will rarely if ever do so based on failure to function. It’s difficult for a challenger to show how granting registration of a mark that fails to function injures him. More importantly, competitors are extremely unlikely to discover such an error unless they are led to seek out specimens of use submitted to the USPTO because they already object to the registration of a mark on other grounds.

III.     Sticky Stories & Empirical Data

This Part explains why Abercrombie (a metonymy for distinctiveness doctrine in its current incarnation) has become trademark law’s stickiest story and why protectability analysis has not evolved to routinely consider use as a mark even in the face of evidence that use has a substantial effect on consumer perception. It reviews marketing and legal literature that demonstrates the role of context in predicting consumer perception and advocates for trademark law to heed the findings of that empirical data by paying greater attention to context.

A.     Abercrombie: Trademark Law’s Sticky Story

The crux of trademark law is consumer perception.163 Does COCA-COLA serve as a trademark for soft drinks? It does if consumers understand it as one. Would a competitor’s use of KOKE-UP infringe The Coca-Cola Company’s trademark rights? Only if consumers are likely to be confused by it. 164 Did the sale of posters featuring the slogan “Enjoy Cocaine” in the iconic font associated with Coca-Cola tarnish the COCA-COLA mark?165 It all depends on how consumers perceived that use. What if the company updates its cursive logo to keep pace with trends—do its trademark rights date back to its first use of the original logo? The Supreme Court recently ruled that question, too, a question of fact that depends on the impression the marks make on consumers.166

When an entire field of law revolves around consumer behavior and public perception, we might expect it to rely heavily on research methods and findings from psychology and social science. But trademark law rarely does.167 Unlike patent law, which considers the specialized expertise of persons skilled in the area in question, trademark inquiries usually purport to gauge or predict the perceptions of the average consumer.168 Examples abound of trademark cases in which federal judges make general statements about consumer behavior without referencing or consulting empirical evidence.169

Linda Hamilton Krieger & Susan Fiske have highlighted how theories about human behavior undergirding legal doctrines often become fossilized rather than evolving: “[J]udges are understandably hesitant to endorse proposed changes to the unstated psychological models underpinning legal doctrine, particularly if this would require modifying the doctrine itself.”170 As a consequence, theories about human behavior “remain embedded in legal doctrine long after they have been disconfirmed or superseded by advances in the empirical social sciences.”171 While many other fields are quick to revise behavioral theories, absorbing and benefiting from new research in psychology and social science, law is reluctant to do so, prizing judicial precedent and stare decisis above hard data about human behavior. Behavioral theories sneak in untested, unexamined, and unacknowledged, with a staying power that belies their casual introduction.

Consumer perception undergirds every major trademark doctrine. How consumers respond to products, trademarks, and branding efforts are topics about which empirical data is readily ascertainable. Marketing is the lifeblood of most major corporations, so those corporations invest millions of dollars in market research and study extensively how best to attract and manipulate consumers’ attention and affections. Two marketing professors note the existence of such research in their empirical study of attention capture in advertising: “Eye-movement data of large samples of consumers attending to large samples of advertisements gathered using infrared eye tracking are currently being produced on an industrial scale and used by companies to optimize decisions on the design of advertisements, packages, Web pages, and other carriers of their visual brand-equity symbols”—i.e., their trademarks.172

The Lanham Act defines a trademark as any matter used “to identify and distinguish [someone’s] goods . . . from those manufactured or sold by others and to indicate the[ir] source.”173 The USPTO and courts have interpreted that definition explicitly to require both distinctiveness (inherent or acquired) and use as a mark. Yet distinctiveness doctrine dominates protectability analyses, while use as a mark often goes unscrutinized. Krieger and Fiske note that “once embedded in published decisions, a behavioral theory can develop precedential legitimacy, and for that reason be difficult to modify, even if it is empirically unsound.”174 Distinctiveness doctrine is not empirically unsound—it’s simply incomplete. As a model for ascertaining whether consumers will perceive matter as a mark, it focuses too much on what the matter is and not enough on what the matter does.175

The ossification of Abercrombie and the sticky stories case law tells about what makes matter a mark are particularly problematic given changes in both how companies are using trademarks, and what types of matter they’re seeking to protect. Use, registration, and enforcement of trade dress, product design, and other nontraditional mark types are on the rise.176 When protection for word marks is sought, those word marks often double as slogans, hashtags, celebrity nicknames,177 or catchphrases. Producers are using trademarks on social media, on display windows, and on 3d-printed products.178 Trademarks are imprinted on everything from eggs179 to bananas180 to football fields.181 They are increasingly used online rather than imprinted on goods—some marks exclusively so. Cross-promotions and co-branding are ubiquitous. The traditional model of trademark and service mark use—a word mark placed prominently above a restaurant or stamped on a bar of soap or written across its paper wrapper—no longer represents the norm.182 That evolution makes trademark use issues more relevant than ever.183 And producer practices and consumer perception create a feedback loop: each is constantly shaping and responding to the other.184

Trademark law’s assumptions about what makes consumers perceive matter as a mark need to keep pace with social science research. While private parties often generate and rely upon empirical data in the form of consumer surveys to bolster claims and defenses in litigation, those surveys are case- and mark-specific. Trademark law more broadly has not evolved to incorporate what research has discovered about consumer perception. The breakneck pace of change in marketing strategy and trademark use by firms requires trademark law to evolve; objective empirical data can help guide its evolution. When it comes to trademark protectability, that empirical data reveals that the current regime’s over-emphasis on inherent distinctiveness and under-emphasis on use as a mark does not a­dequately predict or reflect the perceptions of real consumers.

B.     Empirical Data

A recent empirical study demonstrates the importance of context to consumer perception. In their investigation, Lee, DeRosia, and Christensen sought to confirm the basic precept of the Abercrombie taxonomy: that consumers automatically perceive fanciful, arbitrary, and suggestive marks as source indicators, while they do not perceive descriptive terms and phrases as source indicators until those terms have acquired distinctiveness and become familiar.185 The authors devised a survey that tested consumer responses to hypothetical, computer-generated marks in each category. Instead of confirming the validity of the Abercrombie spectrum, their study found that context—the common indicators of trademark use discussed above, such as large, stylized font and prominent placement—had a substantial effect on consumer perception. The presence of those indicators could lead consumers to perceive even merely descriptive terms as trademarks, while their absence fostered the opposite perception.

In seeking to verify their results, the team designed an additional study in which they tested consumer responses to a single descriptive term, “wonderful,” on a package of cookies, and varied the type and number of indicators of use. Distinctiveness doctrine predicts that consumers would not view the laudatory “wonderful” as a trademark because it is a descriptive term without secondary meaning for cookies.186 Surprisingly, they found that when “wonderful” appeared with a handful of classic indicators of trademark use, most consumers identified it as a trademark, controlling for prior exposure. Even when several of those indicators were eliminated, the effect persisted. Shown the packages below,187 for which distinctiveness doctrine alone predicts consumers would not perceive “wonderful” to serve a trademark function,188 80% of consumers judged the term to be a trademark in the first image and 70% of those shown the second image reached the same conclusion.189

 Image 16 Roberts

But when the study eliminated most or all indicia of trademark use, the descriptive term was no longer perceived as a mark by most consumers. Only 33% of those shown the third image, and 26.7% of those shown the fourth, understood “wonderful” to serve as a trademark for the cookies.

 Image 17 Roberts

Legal doctrines based on behavioral theories often remain static even in the face of hard data that contradicts them. The Lee, et al. study provides empirical evidence highlighting Abercrombie’s limitations and undermining its premise—that inherent qualities of a term or phrase are the sole predictor of whether consumers will understand it as a mark. At the same time, the study bolsters the importance of use as a mark, demonstrating its ability to shape perception about whether matter indicates source.

While the Lee, et al. surveys are the most directly on point, numerous other empirical studies from the fields of marketing, psychology, and linguistics explore how packaging, font, color, and other elements of trademark use affect consumer response. Researchers who looked at the effects of left-or right-side logo placement on consumer perception noted that “understanding how packaging variables, such as shape, color, and graphics, affect consumer perception, evaluation, and behavior is of theoretical and managerial importance.”190 An article in the Journal of Experimental Psychology laments that trademark registration often involves assessment of words “in relative isolation” and “without considering other factors such as colour, font, packaging, or marketing of products.”191 Attention scholars have found “that a prominent brand element, reflected among [other things] in its size, captures more attention to the brand, which is a necessary condition for obtaining the desired brand-communication effects.”192 A study conducted by Pieters & Wedel found that an increase in text size as small as 1% increased consumer attention significantly,193 and prominent placement of a brand’s “visual identity symbols” also increased attention to the brand.194 Other studies have shown “that varying the ease with which people can read a font can create a fluency effect that alters choice.”195

Moreover, marketers are acutely aware of the role that use and context play in attracting consumer attention. Researchers have studied the effects of low versus high logo placement:

On average, consumers take five to seven seconds to evaluate a package on a shelf. According to Procter & Gamble, shoppers decide whether they want to purchase a product in three to seven seconds, the same time it takes to note the product’s packaging. As such, companies such as Procter & Gamble spend considerable resources on perceptual cues in package design that can evoke almost an instant, gestalt response in consumers.196

That idea of the gestalt response echoes the definitions of use as a mark articulated by case law, treatises, and scholars alike. Marketing researchers describe consumers’ quick “read” of products based on visual evaluation and note that “the overall effect of the package comes not from any individual element but rather from the gestalt of all elements working together as a holistic design.”197 Their language is similar to that of McCarthy, who maintains that it should be clear “at first glance” that matter is being used as a mark and, “if when viewed in context, it is not immediately obvious”198 that something is a trademark, it probably fails to merit protection as one.199

The Lee, et al. studies, along with those regarding consumer attention and perception, impart clear lessons about the inadequacy of considering only inherent distinctiveness when assessing protectability. Instead of making predictions about consumer perception based on distinctiveness alone, courts should heed the findings of such studies by increasing their attention to context and indicia of use.200 While federal courts adjudicate whether matter qualifies as a mark far less often than does the USPTO, and use is at issue in only a subset of those cases, how courts handle use shapes the precedents and rules that the USPTO applies in assessing registrability. So, when courts ignore the use question, their treatment reinforces the idea that use doesn’t matter.

IV.     At the Intersection of Use & Distinctiveness

This Part reviews federal case law and USPTO decisions in which factfinders (1) focused on distinctiveness and ignored use as a mark, (2) focused on use as a mark and ignored distinctiveness, or (3) considered both seriatim even though combining them would have provided greater insight and produced more accurate and consistent decisions. It discusses tests used in various jurisdictions to assess the distinctiveness of nonverbal marks and considers how those tests might prove helpful models for assessing consumer perception of word marks.

Factfinders addressing the protectability or registrability of a mark consider use as a mark and distinctiveness in a variety of ways. Most often, when they consider use as a mark at all, courts and the USPTO treat the two separately, as distinct thresholds for protection: matter must be used as a mark, and it must possess either inherent or acquired distinctiveness, to warrant protection under the Lanham Act.201 Because the failure to satisfy either requirement is theoretically disqualifying, courts and the TTAB may find a mark fails to qualify for protection based only on one or the other determination. An adversary attempting to demonstrate that an applicant or plaintiff lacks rights may do so on either basis independently, and the USPTO can decline to grant protection for a purported trademark on one basis or the other. But treating use as a mark and distinctiveness as entirely separate inquiries renders each analysis incomplete. The below discussion demonstrates how use as a mark analyses that do not consider distinctiveness and distinctiveness analyses that fail to pay attention to use rest on shakier ground. And in many cases, that separation is awkward and artificial. Decisions often read like factfinders are struggling to split a single question—will consumers perceive this matter as a mark?—into two separate inquiries against their better judgment.

In some cases, factfinders do integrate the two inquiries. That approach is more common in assessments of trade dress and other nontraditional marks,202 for which courts and USPTO attorneys often consider the question of whether matter makes a separate commercial impression in the course of their distinctiveness inquiry. But, factfinders would do well to incorporate the “separate impression” question into their protectability analyses for word marks too, since it brings use as a mark into the distinctiveness equation, leading to more accurate predictions about whether consumers are likely to perceive matter as a mark.

A.     Use as a Mark Without Distinctiveness

In West-com, the TTAB affirmed the USPTO’s refusal to register the image of a toilet (below left) as a trademark for an electronic device for sending communication transmissions between patients and health care personnel when patients require a bed pan or other assistance.203 The specimen included images of the call device showing use of the toilet symbol on one of the device’s six buttons (below right):

 Image 18 RobertsThe TTAB focused its analysis entirely on failure to function, finding that based on the specimens submitted, consumers would not view the matter as a trademark for the goods described.204 Its thirteen-page analysis declined to classify the mark as generic, descriptive, or inherently distinctive for the goods in question.205 But a button bearing the basic image of a toilet, used by a patient to request a bedpan or assistance visiting the restroom, is at best descriptive, if not generic. In treating distinctiveness and failure to function separately, the TTAB rendered a decision with less force than one that incorporated both bases for refusal together.

That’s true not simply because two is more than one. Rather, failure to function is circumstantial and depends on the specimens submitted to reflect the applicant’s current use.206 Savvy trademark attorneys boast that, given the freedom to help their clients manipulate their use of symbols or phrases, they could surmount just about any failure to function refusal.207 As long as they counsel their client to use matter in a way that is sufficiently “trademark-y,” relying on indicators like size, color, and prominence, they can provide the USPTO with a specimen that will pass muster.208 A failure to function refusal is thus often read as a directive to the applicant to manipulate the use and try again. A refusal based on lack of inherent distinctiveness, on the other hand, is a message to the applicant that the matter itself can only become registrable after it has acquired distinctiveness based on consumers’ exposure to it. And a refusal based on genericness is a message to abandon all hope of acquiring trademark rights in the matter as a mark.

Examples of factfinders assessing use without distinctiveness are more common for marks straddling the border between descriptive and suggestive. The Ninth Circuit affirmed a lower court’s assessment that PARAMAHANSA YOGANANDA, the name of a famous yogi and guru, failed to function as a mark for religious services and book publications where the specimens used the term inconsistently and “with[out] any of the traditional trademark indicia (e.g., use of the term with a “TM” sign next to it).”209 The court did not address the mark’s lack of inherent distinctiveness.

Likewise, the TTAB in Mentor Graphics210 took up the question of whether VIRTUALWIRES, as used in the below specimen, functioned as a mark for computer hardware and software.211

 Image 19 Roberts2

The USPTO had refused registration in part based on failure to show proper use, because the specimen showed the mark “embedded in a descriptive sentence . . . in small text in the middle of a specimen containing other more prominently placed text and graphic elements.”212 The Examining Attorney asserted that “[t]he descriptive nature of applicant’s mark [was] reinforced by” applicant’s use.213 He further noted “the terms VIRTUAL and WIRES are both highly descriptive terms that, when used together, create a phrase that consumers are likely to mistake for a generic technology when not displayed in a separated and prominent manner.”214 And yet, the USPTO did not list mere descriptiveness as a basis for refusal.

Without an assertion of descriptiveness, the TTAB claimed it was left to “presume that, if properly used, VIRTUALWIRES is a distinctive mark.”215 It reversed the refusal to register, finding that VIRTUALWIRES functioned as a mark in the specimen submitted, as “evidenced by the nature of the sentence itself; by the representation of the term VIRTUALWIRES in all capital letters followed by the ‘TM’ symbol in all bolded capital letters; and by the mark’s distinctive use to modify the descriptive terminology for the goods as broadly characterized”216 (i.e., its use as an adjective). Implicit in the language of the Office Action is the idea that the examining attorney believed the mark was, in fact, descriptive for the services specified, or perhaps that the mark fell on the border between descriptive and suggestive and thus required more robust use to lead consumers to perceive it as a mark. But because the USPTO rested on a single basis for denial, the TTAB addressed only trademark use in isolation and not distinctiveness.

Factfinders in a number of domain name cases have taken a similar approach, focusing only on use and neglecting to account for the role the proposed mark’s distinctiveness (or lack thereof) would play in consumer perception.217 In Eilberg, the TTAB affirmed the refusal to register WWW.EILBERG.COM as a service mark for legal services because the use demonstrated in the specimen failed to function.218 The applicant had submitted letterhead that bore the domain name; the TTAB noted that the desired mark was “used inconspicuously in a very small and subdued typeface.”219 Further, the domain name “merely impart[ed] information, in the same manner as an address, phone number or other informational statement, about how to reach the applicant,” supporting the presumption that it did not function as a service mark.220 The opinion nowhere mentions distinctiveness, despite the high likelihood that “Eilberg” would be deemed primarily merely a surname and thus merely descriptive.221 Likewise, the TTAB in Industrial Risk affirmed a refusal to register “” for insurance underwriting services for industries because the specimen reflected use as merely a domain name; its opinion failed to acknowledge that the mark’s mere descriptiveness for the services also provided grounds for refusal.222

B.     Distinctiveness Without Use as a Mark

Conversely, in many cases, courts and the USPTO have focused on distinctiveness in assessing protectability or validity without accounting for use or context at all.

The Federal Circuit’s widely-cited opinion in Oppedahl & Larson is a prime example.223 In that case, a law firm sought to register PATENTS.COM as a trademark for computer software for managing records and tracking their status online.224 The Federal Circuit affirmed the TTAB, which had affirmed the USPTO: all agreed that registration for PATENTS.COM must be refused because the mark was merely descriptive of the goods for which registration was sought.225 At each step, factfinders addressed and dismissed the argument that the “.com” top-level domain (TLD) somehow added distinctiveness to the mark; instead, they likened common TLDs like “.com,” “.net,” and “biz” to business entity designations like “Corp.” and “LLC.”226 And yet, a look at the specimen that the law firm submitted to support its application reveals a glaring omission:

Image 20 Roberts

The use reflected in the specimen predicts consumers will perceive as a mere domain name, and not necessarily as a trademark. But neither USPTO Office Action addresses the trademark’s failure to function, nor do the TTAB or CAFC opinions.227 An approach that integrated use as a mark and distinctiveness in assessing registrability would have provided a more robust basis for refusal. It also would have better served the applicant by signaling that not only was the desired mark unregistrable without a showing of secondary meaning, but also that—as presented in the specimen of use—the desired mark was not being used as a mark.

ESPN v. Quiksilver provides an example in which use as a mark ought to have affected a court’s validity assessment in the infringement context. In that case, a federal district court treated Quiksilver’s stylized letter “X” mark as sufficiently distinctive to warrant protection and declined to consider failure to function, despite ESPN’s compelling argument that Quiksilver had not made a threshold showing that it actually used the stylized X as a mark.228 Quiksilver asserted in counterclaims that ESPN had infringed its valid trademark rights, alleging that since 1986 it had “incorporat[ed] an ‘X’ symbol into its boardshorts and other clothing designs and marketing materials[.]”229 To support its assertion of trademark rights, it appended images of 22 different logos230 that included a stylized “X.” In defending itself against ESPN’s infringement claim, however, Quiksilver argued in the alternative that it had made only “ornamenta[l] . . . non-trademark use” of the matter in question, and so could not be found to infringe ESPN’s trademarks.231 In its motion to dismiss Quiksilver’s counterclaims, ESPN explained:

Visual inspection reveals that Quiksilver’s grouping of the images under the moniker “Gen X Brand” is an artifice. The “designs” are not uniform by any means – some include human or animal figures; some include the word “Quiksilver”; others include different words, such as “Extreme”, [sic] “Factor,” “Quix,” or “America’s Future” . . . .

Independent of its failure to identify in any decipherable way the so called “Gen X Mark”, [sic] Quiksilver presents the images without context, failing to show how any of the cited “designs” were actually used on clothing or other items.232

The court declined to seriously consider whether Quiksilver had used its X as a mark. It declared the failure to function argument “simply wrong” without further analysis233 and denied ESPN’s motion to dismiss. While Quiksilver would still need to establish trademark rights at trial, the court’s refusal to acknowledge at the pleading stage that a party alleging infringement and dilution of a common law mark must show use as a mark undergirding its assertion of valid trademark rights is surprising.234

And the Quiksilver case may not be an outlier. Many courts simply decline to take use as a mark seriously. Too often, courts recite the use requirement and then summarily conclude that it is met.235 In a line of cases regarding trademarks in the worlds of television shows or comic books, courts held marks protectable based on the role they played in their respective fictional universes. Rather than ask whether marks like KRUSTY KRAB, KRYPTONITE, or DAILY PLANET were used in a trademark way in connection with actual goods or services, the courts granted the marks protection as “elements” and “ingredients” of successful television shows and comics.236 As the defendant in the KRUSTY KRAB case pointed out, the plaintiff had not demonstrated or even adequately pleaded use as a trademark or use in commerce to support its assertion of rights: “Since plaintiff does not use KRUSTY KRAB as a trademark, it is not certain to what plaintiff is alleging its ‘trademark’ applies,” making a true validity analysis impossible.237 A review of federal case law and underlying briefs reveals that courts often fail to identify use as a mark as an issue or decline to make a finding on it even when litigants raise it.

In some cases, courts doggedly focus on distinctiveness and ignore use as a mark even in the face of contrary guidance from the USPTO. In Tenneco v. Kingdom Auto, an infringement suit based on defendant’s copying of plaintiff’s six-digit automotive replacement part numbers,238 the Sixth Circuit considered whether a district court erred in holding model numbers ineligible for trademark protection. Unlike the TTAB in ChaCha, the court treated the six-digit numbers as merely descriptive and focused its lengthy opinion entirely on assessing whether Tenneco had made a showing of secondary meaning sufficient to support its claim. Ultimately, the Sixth Circuit majority affirmed the district court in holding that the product numbers possessed neither inherent nor acquired distinctiveness, granting the defendant’s counterclaim for declaratory judgment that the numbers were not protectable trademarks. The Sixth Circuit several times referenced the USPTO’s refusal to register the number marks because they failed to function, but the circuit court itself never explicitly addressed the issue. The district court’s approach was similar, although it made passing reference to the way Tenneco used the model numbers “in a large, bold font” and consistently paired them with other marks.239

Had the Sixth Circuit foregrounded use as a mark, it could have dispensed with the protectability issue far more easily. Tenneco had submitted the following specimen with its application to register 171994 as a trademark for automobile parts240:

 Image 21 Roberts

The examining attorney issued several office actions, asserting that the number would not be perceived as a trademark and citing the TMEP’s straightforward guidance indicating that model numbers are likely to fail to function as marks.241 Per the TMEP, “matter used merely as a model or grade designation serves only to differentiate between different products within a product line or delineate levels of quality, and does not indicate source. . . . Even though a model or grade designation seems ‘arbitrary’ in the sense that the combination of letters, numbers, or both does not immediately describe the goods, it often does not function as a trademark.”242 While some model numbers may be capable of serving as trademarks, the TMEP presumes they do not, a presumption applicants can only overcome with evidence of secondary meaning.

Cases like Tenneco, in which federal courts simply ignore the wisdom of the USPTO, perhaps best embody factfinders’ inconsistent approaches and the pitfalls of segregating distinctiveness and use considerations in the first place.

C.     Integrating Use & Distinctiveness

This Article focuses on the importance of use as a mark to the protectability of word marks and other two-dimensional visual marks that feature text. That’s because those represent the largest proportion of marks that are registered or enforced in federal actions. It’s also because courts and the USPTO pay the least attention to use as a mark in assessing protectability for marks in those categories, instead allowing the Abercrombie version of distinctiveness to do the heavy lifting. But when factfinders assess rights in trade dress and other nonverbal marks, use as a mark is harder to ignore. In a sense, trade dress doctrine seems to have stumbled upon a more comprehensive approach that can serve as inspiration in reshaping assessments of word mark protectability. Because trade dress doctrine is not particularly explicit or purposeful in integrating use as a mark in its approach to assessing distinctiveness, though, it cannot by itself provide a clear roadmap to guide trademark doctrine.

In 2004, the gas station chain Chevron filed a use-based application to register the design of its “stylized pole spanner sign” for automobile service station services.243 The first image below depicts the trade dress in isolation; the second shows how consumers encounter it:

 Image 22 Roberts

The TTAB affirmed the Examining Attorney’s refusal to register. It assessed the design’s inherent distinctiveness under the Seabrook test,244 which is frequently invoked for nonverbal marks. The Seabrook factors are often described as a substitute for Abercrombie, more appropriate for nonverbal marks; in addition to considering the relationship between the mark and the goods or services, Seabrook also places the mark in the context of the relevant industry and its norms, and foregrounds competitors’ uses of similar matter as relevant to the distinctiveness inquiry. Courts consider four factors:

[1] whether [the matter] was a “common” basic shape or design, [2] whether it was unique or unusual in a particular field, [3] whether it was a mere refinement of a commonly-adopted and well-known form of ornamentation for a particular class of goods viewed by the public as a dress or ornamentation for the goods, or [4] whether it was capable of creating a commercial impression distinct from the accompanying words.245

The design mark in Chevron fell short on every front—the Board found that the pole spanner shape was indeed a mere refinement of a common design for the particular class of services. It further noted that “under Seabrook factor four, the pole spanner shape is not such that it will create a commercial impression distinct from the CHEVRON word mark and logo.”246 Upon initial review, “[t]he Examining Attorney [had] refused registration on the grounds that the subject matter sought to be registered is nondistinctive trade dress that does not function as a service mark . . . and . . . has not acquired distinctiveness[;]” the TTAB affirmed on the same grounds.247

In analyzing whether the pole spanner design would be perceived as an inherently distinctive source indicator by consumers, the Board in Chevron simultaneously contemplated whether the trade dress failed to function.248

The fourth Seabrook factor invites factfinders to combine the two inquiries: it asks, as part of a distinctiveness analysis, whether matter creates a distinct commercial impression.249 That “distinct commercial impression” language is echoed in some cases assessing whether word marks fail to function.250

While a number of courts apply Seabrook in assessing the protectability of nonverbal marks, many dispense with the final factor precisely because it measures something different from the first three.251 Tests from Duraco252 and Knitwaves,253 used in several circuits to assess the inherent distinctiveness of nonverbal marks, also straddle the intersection of distinctiveness and use as a mark, asking whether the matter is “conceptually separable from the product” and whether it is “likely to serve primarily as a designator of origin of the product.”254 And occasionally the court or Board assessing the distinctiveness of trade dress considers whether a mark makes a “separate commercial impression” from the surrounding matter without reciting the tests from any of those cases.255

Use as a mark and inherent distinctiveness are so often intertwined in trade dress analyses that they can be impossible to separate, regardless of the specific language used.256 The same intermingling of use and distinctiveness is discernible in the TTAB’s treatment of other nontraditional trademarks, as in Nextel,257 assessing the registrability of a chirp sound for mobile phones; Organon, considering an application to register an orange flavor as a trademark for a fast-dissolving antidepressant;258 and Fantasia, affirming the USPTO’s refusal to register a repeating diamond pattern as a trademark for hookahs.259 In those and similar cases, the analyses take as a given the interdependent nature of use as a mark and inherent distinctiveness when it comes to nonverbal marks.

Perhaps that’s because most forms of trade dress can only be evaluated in context. Whether consumers will perceive the pole spanners at issue in Chevron or the diamond pattern on the hookahs in Fantasia as source indicators are questions that take into account both what the design features are and how consumers encounter them.260 Asking whether “a beveled pole spanner sign in the shape of [a] tri-dimensional beam on the top and . . . a tri-dimensional hexagon on the bottom” is inherently distinctive for gas station services,261 or whether “repeating rows of diamonds . . . on the lower third of the cylinder of the hookah device” are inherently distinctive for hookahs is nonsensical.262 Those questions are only cognizable at the intersection of inherent distinctiveness and use.263 While courts and the USPTO can assess the Abercrombie distinctiveness of potential word marks by considering them in isolation, trademark law is not well-served by that approach. Too often, it leads factfinders to ignore how a term or phrase is actually used and whether consumers will understand the matter as a mark. Instead, they should adopt an approach similar to that seen in trade dress cases and integrate distinctiveness and use analyses in assessing the protectability of word marks. Whether consumers are likely to perceive matter as a mark, and thus whether that matter warrants trademark protection, depend on what it is and whether it creates a distinct commercial impression. If different kinds of marks require different levels of use, the fourth Seabrook factor helps factfinders gauge whether even minimal use as a mark is made; the McCarthy markers (font, color, size, prominence, trademark spot, TM and ® symbols, use that isn’t embedded in a sentence, etc.) can aid further in determining whether the use is strong enough to qualify the matter for protection given the mark type. And factfinders should keep in mind the overall relationship between word marks’ inherent distinctiveness and their use:

 Image 23 Roberts

V.     A Use-Based Approach to Source Distinctiveness

In building and understanding trademark doctrines, we usually begin with words. Most trademarks are word marks. Federal protection initially extended solely to words, and only later to nonverbal marks. Words seem to offer the simplest foundation upon which to construct doctrines based on use, distinctiveness, strength, fame, confusion, and dilution. Doctrines based on word marks are then extrapolated to nonverbal marks and adjusted where ill-fitting or incomplete. But courts and the USPTO could benefit from modeling their approach to assessing word marks’ use and distinctiveness after the approach typically used to assess the protectability of nonverbal marks. Judges and examining attorneys assessing word marks must do more than simply place a term or phrase into one of five Abercrombie buckets. As the empirical evidence demonstrates, analyses that fail to take into account how matter is used will lack sufficient information to predict consumer perception. An approach to protectability that incorporates both use and distinctiveness accepts and acknowledges that the question of whether consumers are likely to perceive matter as a mark depends in no small part on how they encounter that matter—its prominence, placement, appearance, context, and separability from other surrounding matter.

This Article advocates that courts and the USPTO incorporate attention to trademark use into their protectability analyses along with Abercrombie distinctiveness,264 because each assessment shapes the other and both doctrines attempt to predict whether consumers will perceive matter as a mark. Doing so would not require any change to the statute, because the use as a mark requirement is already embedded in the Lanham Act and case law interpreting it. It would only require a change in approach, perhaps along with more robust TMEP guidelines that emphasize the indicia of use discussed above and the requirement that matter create a separate commercial impression of the type that will lead consumers to regard it as a mark.

What would such a merged analysis look like in practice? For nonverbal marks, Seabrook and its brethren tests for inherent distinctiveness often already incorporate use as a mark. In asking whether a mark creates a separate or distinct commercial impression to consumers apart from any surrounding matter or is likely to serve primarily as a designator of origin, factfinders are bound to account for context, prominence, emphasis, and other indicia of use. Courts and the USPTO ought to retain that factor more consistently in assessing nonverbal marks and take a similar approach to assessing text marks. By simply importing those considerations into the test for word mark protectability, they will be assessing whether the matter is inherently distinctive and whether it is used in such a way that consumers are likely to perceive it as a trademark in a single, unified inquiry. The distinctiveness analysis would thus consider: (1) what is the relationship between the matter for which protection is sought and the goods or services with which it is used; and (2) to what extent does the matter, as it is used in commerce (for common law rights) or in the specimen submitted (for registration), create a distinct commercial impression?

It would oversimplify these issues to say that any matter used as a mark—as Lee, et al.’s “wonderful” is used as a mark—will be perceived as a mark and should thus qualify for protection. Distinctiveness doctrine denies protection for merely descriptive matter without secondary meaning not only because consumers are unlikely to perceive it as a trademark, but also because overprotecting descriptive terms and phrases chills speech and raises barriers to entry for competitors without justification. Relatedly, even if descriptive matter is used as a mark, consumers will be unlikely to be confused by the use of the same or similar descriptive matter by two different producers if neither has acquired distinctiveness. Conversely, a flat rule for assessing use as a mark that treats fanciful and arbitrary word marks the same way it treats descriptive or generic terms or nonverbal matter would produce bad outcomes. As courts and the USPTO have held, an inherently distinctive mark like TRULICITY for pharmaceuticals requires fewer obvious trappings of trademark use than does an arbitrary mark like SKITTLES for candy or BUMBLE & BUMBLE for hairspray or a suggestive mark like SUAVE for shampoo, which in turn require fewer indicia of trademark use than a descriptive term like SPECTRUM or a domain name like IRESTMYCASE.COM. The below chart maps on a number of the examples discussed in this Article to show how neither use as a mark nor distinctiveness is a binary determination; rather, each exists on a spectrum. The relationship between them predicts consumer perception and thus ought to dictate protectability, as shown here265:

Image 24 Roberts

Incorporating use considerations, then, will be most crucial for those word and stylized marks that fall on the border between inherently distinctive and merely descriptive.266 All of the trappings of trademark use in the world shouldn’t render a merely descriptive mark protectable without acquired distinctiveness. But if a term occupies liminal territory under Abercrombie, strong trademark use increases the likelihood that consumers will perceive it as a mark, and thus ought to nudge it toward protectable status.267 The TTAB in Lilly acknowledges as much, distinguishing the fanciful mark TRULICITY from those that fail to function because they “consist[]of descriptive or highly suggestive wording, such that they could be perceived merely as information about the goods.”268 And the USPTO and courts are more likely to merge use as a mark and distinctiveness analyses for word marks when the proposed mark is a descriptive one.269

Incorporating use considerations is also particularly important for those marks whose primary role is something other than source indication. A purported mark that doubles as a domain name, a trade name, a hashtag, a decoration, a celebrity nickname, or a model number, to give just a few examples, will only also be perceived as a mark if it is truly used as one. #ShareTheSilence may technically be an arbitrary phrase for alcoholic beverages, but if its use reflects only use as a hashtag, consumers will view it as merely that. An inherently distinctive domain name will not support registration if its only use is to communicate the address of a website. “X” is not descriptive for surf apparel, nor is HONEY BADGER DON’T CARE for hats, but if their use is ornamental, then consumers will perceive them as decorations rather than trademarks. Kim Kardashian West may be famous, but to support federal trademark rights in her name, Ms. West needs to show sufficient indicia of use to suggest consumers will view KIM KARDASHIAN WEST as a source indicator for entertainment services and not just the name of a person.270 And the sequences of numbers that tripped up the USPTO in ChaCha and courts in Tenneco won’t be seen as marks unless they’re truly used as marks.

Factfinders struggle with borderline marks, and the tests and proxies they apply fail to produce consistent decisions.271 The determination of whether matter is descriptive or suggestive is a subjective one, and judges and examining attorneys often use their own reactions as proxies for consumers at large. The more descriptive a borderline term appears to be, or the more it serves some purpose other than source indication, the more its use must bear all the trappings of trademark use if consumers are to understand it as a source indicator from the time they first encounter it.272 And the less the use calls attention to the matter as a mark or creates a separate commercial impression, the greater the inherent distinctiveness would need to be in order to presume that consumers will understand it as serving a trademark function. Context and manner of use provide additional clues to how consumers will respond to matter used in connection with the sale of goods or services; accounting for them increases the accuracy of those judgments. Even more crucially, matter must be used in a trademark way to come within the auspices of the federal trademark statute.

If the USPTO and courts were to pay use as a mark the attention it deserves, a more radical proposal arguably follows: courts could limit the scope of registered trademark or trade dress protection to the uses identified in specimens submitted to and accepted by the USPTO in connection with registration. After all, the scope of registered rights is already limited to those goods and services identified in the application. Mark owners seeking to enforce rights based on use in connection with additional goods or services not covered in their registrations must rely on their common law rights, rather than claiming their registered rights expand more broadly than the registration specifies. Similarly, owners seeking to enforce rights based on use different from or broader than the type of use they demonstrated in their initial applications would be free to establish common law rights, but the presumptions that accompany registration would extend only to the use shown in the approved specimens. Such practice might also help minimize the slippage that occurs in cases where a registered composite or stylized mark asserted in litigation is assigned the broader scope of a word mark.273

VI.     Conclusion

Use as a mark and distinctiveness are coextensive. Together they ask and answer the fundamental question of trademark protectability without which no trademark rights can accrue: do consumers perceive the matter for which protection is sought as an indication of source?274 If they don’t, extending protection and enforcing the mark against others burdens both speech and competition and fails to serve the interests of either consumers or producers.

Use as a mark often receives inadequate weight at the USPTO and in court. Empirical studies reveal that context and type of use can have a significant impact on whether consumers perceive matter to be a mark and paying heed to such cues helps counterbalance the tendency of factfinders in trademark matters to rely solely on inherent distinctiveness in assessing word mark protectability. The fourth Seabrook factor, the indicia of use discussed throughout, and the charts in this Article offer concrete, articulable ways to analyze use and assess its impact on overall protectability. Integrating use as a mark with distinctiveness assessments would reinvigorate both doctrines and help ensure that statutory and common law requirements for protection are met by protecting only matter that consumers actually perceive as source indicating. When it neglects consumer perception, ostensibly its core concern, trademark law itself fails to function.


 Appendix A

The below chart shows the relationship between inherent distinctiveness and use as a mark:

 Image 25 Roberts

The below graph plots examples from the Article to illustrate the relationship between inherent distinctiveness and use as a mark. The examples are reproduced and explained further below:

 Image 26 Roberts

242242: Between descriptive and suggestive; slight use as a mark (set apart from surrounding matter)

 Image 27 RobertsCOLD-EEZE: Just above descriptive; strong use as a mark (large block font outlined in black, centered, different colors, stylized letters with borders, in the trademark spot)


CHERRY FLAVOR: Generic; no use as a mark

 Image 28 RobertsDUNKIN’ DONUTS: Suggestive; maximum use as a mark (colors, font, prominent position, large size, ®)

 Image 29 Roberts

SUAVE: Suggestive; strong use as a mark (font, size, color, stylized rectangle setting mark apart from surrounding matter, ®)

 Image 30 Roberts

BUMBLE AND BUMBLE.: Arbitrary; moderate use as a mark (separate; initial capital; slightly enlarged font)

 Image 31 Roberts

SPECTRUM: Suggestive; moderate use as a mark (color, initial capital)

 Image 32 Roberts

TRULICITY: Fanciful; just above slight use as a mark (TM symbol, initial capital)

 Image 33 Roberts

#SHARETHESILENCE: Arbitrary or descriptive; slight use as a mark (all capitals; separate)

 Image 34 Roberts

#BEUNPRECEDENTED: Descriptive; slight use as a mark (color, capital letters)

 Image 35 RobertsCHIPSET-FREE CHARGING: Descriptive; slight use as a mark (TM symbol, italics)

 Image 36 Roberts

Wonderful I: Descriptive; strong use as a mark (size; font; oval with shadow; prominent position; initial capital)

 Image 37 Roberts

Wonderful II: Descriptive; slight to moderate use as a mark (initial capital; size; prominent position)

 Image 38 Roberts

Wonderful IV: Descriptive; no use as a mark

 Image 39 Roberts

X: Suggestive or arbitrary; no use as a mark

 Image 40 Roberts

SKITTLES: Arbitrary; strong use as a mark (size, prominence, outlined block font with shadow, stylization, color, border, ®) (not shown in Article)

 Image 41 Roberts

Appendix B


 Image 42 Roberts



  1. [1]. 242242, Registration No. 3,504,586 (image above is a partial screenshot of the specimen). Red arrows have been added to specimens throughout this Article to direct readers’ attention to the matter being discussed.

  2. [2]. ChaCha Search Inc. v. Grape Tech. Grp. Inc., 105 U.S.P.Q.2d (BNA) 1298, 2012 WL 6929402, at *2 (T.T.A.B. 2012). Grape argued that 242242 was merely descriptive and possessed neither inherent nor acquired distinctiveness. Id. “Distinctiveness” refers to a trademark’s ability to indicate source based on its innate characteristics (inherent distinctiveness) or consumers’ familiarity with it as a result of exposure over time (acquired distinctiveness). See U.S. Patent & Trademark Office, U.S. Dep’t of Commerce, Trademark Manual of Examining Procedure
    § 1209.01(b) (Oct. 2018) [hereinafter TMEP],
    #/current/d1e2.html (describing the difference between distinctive and merely descriptive marks); see also id. § 1212 (defining acquired distinctiveness or secondary meaning).

  3. [3]. Barton Beebe & C. Scott Hemphill, The Scope of Strong Marks: Should Trademark Law Protect the Strong More Than the Weak?, 92 N.Y.U. L. Rev. 1339, 1346 (2017) (“In general, a mark is inherently distinctive in inverse proportion to the degree to which it describes the product with which it is used.”).

  4. [4]. Abercrombie & Fitch Co. v. Hunting World,Inc., 537 F.2d 4, 9 (2d Cir. 1976). A fanciful mark is a made-up term or phrase, like EXCEDRIN for an analgesic or KODAK for a camera. See TMEP § 1209.01(a). An arbitrary mark is an existing term or phrase that bears no logical relationship to the goods or services with which it’s used, such as PENGUIN for books or STRAWBERRY for retail clothing stores. See id. A suggestive mark is a term or phrase that subtly hints at some quality of the goods or services, like GREYHOUND for train services. See id. These three types of marks are categorized as inherently distinctive and thus protectable from their earliest use in commerce. See id. A descriptive mark is a term or phrase that directly describes some feature or characteristic of the goods, like 24-HOUR ENERGY for a caffeinated drink or CALIFORNIA CLOSETS for a California-based company’s closet-building services. See id.
    § 1209.01(b). Descriptive trademarks can only earn protection upon a showing of acquired distinctiveness (also called secondary meaning), i.e., proof that consumers have come to view the mark as an indicator of source. See id. A generic term or phrase is a name for the genre of goods or services, like THE PILLOW STORE for a store that sells pillows or BLINDED VETERANS ASSOCIATION for a charity that benefits blinded veterans. See id. § 1209.01(c). Generic terms and phrases are incapable of being or becoming trademarks. Id. In practice, the USPTO does not typically classify a mark more specifically than deeming it inherently distinctive, merely descriptive, or generic, although courts may do so in assessing inherent distinctiveness as a component of trademark strength as part of a likelihood of confusion analysis. See id. § 1209.01; Abercrombie, 537 F.2d at 9.

  5. [5]. Lanham Act, Pub. L. No. 79-489, 60 Stat. 427 (codified as amended at 15 U.S.C.
    §§ 1051–1141 (2012)). For example, matter might be barred from registration as functional or deceptive, or because it creates a false association with an individual or a likelihood of confusion with a previously registered mark. Id. § 1052(a), (d).

  6. [6]. To merit protection, a trademark must be (1) used in commerce, (2) used in connection with the goods or services for which trademark rights are asserted, and (3) used in a trademark way. See id. § 1051(a)–(b); see also infra notes 9­-11 (discussing the requirements for trademark protection). This Article focuses on the third requirement, but courts and the USPTO often collapse any two or all three of these categories of “use.” See infra notes 9­-11 and accompanying text.

  7. [7]. ChaCha, 105 U.S.P.Q.2d at *6.

  8. [8]. Grape eventually tried to add failure to function as an additional ground for cancellation, but it was too late; because Grape did not timely file its motion for leave to amend its counterclaim, the TTAB considered only the distinctiveness challenge. Id. at *3.

  9. [9]. 15 U.S.C. § 1127; 1 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 3:1 (5th ed. 2017), Westlaw (“[R]equirements for qualification of a word or symbol as a trademark can be broken down into three elements:
    (1) ... [a] symbol ... (2) ... [used] as a mark ... (3) ... to identify and distinguish the seller’s goods from goods made or sold by others.”).

  10. [10]. Trademark use is a threshold requirement for federal protection, but the concept of use is broad and extends far beyond use as a mark, which is the focus of this Article. To come within the jurisdiction of the Commerce Clause, and therefore the Lanham Act, a trademark must be in use in connection with goods or services that travel in interstate commerce or the sale of which affects interstate commerce. Trade-Mark Cases, 100 U.S. 82, 97 (1879); Christian Faith Fellowship Church v. Adidas AG, 841 F.3d 986, 989, 992 (Fed. Cir. 2016); Larry Harmon Pictures Corp. v. Williams Rest. Corp., 929 F.2d 662, 667 (Fed. Cir. 1991); In re Gastown, Inc., 326 F.2d 780, 783 (C.C.P.A. 1964). And since 1989, use in commerce requires “bona fide use of a mark in the ordinary course of trade” and not mere token use made to reserve rights in a mark. Trademark Law Revision Act of 1988, Pub. L. No. 100-667, 102 Stat. 3935 (codified as amended at 15 U.S.C. § 1127).

  11. [11]. The “use as a mark” requirement comes from the Lanham Act itself and refers to the way in which matter is used. See 15 U.S.C. §§ 1051–1052, 1127. To be protectable, matter must be used in such a way that consumers will understand it as a trademark. Sheldon H. Klein & N. Christopher Norton, The Role of Trademark Use in US Infringement, in Trade Mark Use 330 (Jeremy Phillips & Ilanah Simon eds., 2005) (“The mark must be used in a trade mark sense, that is, it must be used to identify the source of goods or services and distinguish that source from other sources.”).

  12. [12]. Anything can gain protection as a trademark if it functions as one, including words, images, product packaging, décor, three-dimensional product features, scents, sounds, colors, or motions. This Article focuses initially on the use as a mark requirement as applied to word marks and other two-dimensional visual marks, such as logos, stylized words, and composite word/design marks, and later turns to protectability assessments for nonverbal marks as offering a model that word mark assessments could benefit from emulating. The use as a mark requirement applies equally to marks of all types.

  13. [13]. From 2003 through 2016, the USPTO issued 26,450 failure to function refusals. See infra Appendix B. Barton Beebe and Jeanne Fromer compiled a dataset of all trademark office actions issued during that period by systematically downloading them from the USPTO website; to determine the rate of failure to function refusals, they autocoded the office action language for any mention of “failure to function,” “fails to function,” or “fail to function.” See E-mail from Barton Beebe, Professor, N.Y. Univ. Sch. of Law, to Alexandra J. Roberts, Assoc. Professor, Univ. of N.H. Sch. of Law (Aug. 15, 2018) (on file with author). For more on their dataset and methodology, see Barton Beebe & Jeanne C. Fromer, Are We Running Out of Trademarks? An Empirical Study of Trademark Depletion and Congestion, 131 Harv. L. Rev. 945, 973–74 (2018). I am indebted to Professors Beebe and Fromer for generating these failure to function statistics from their dataset.

  14. [14]. See United Drug Co. v. Theodore Rectanus Co.,248 U.S. 90, 97 (1918)(“There is no such thing as property in a trade-mark except as a right appurtenant to an established business or trade in connection with which the mark is employed.”); Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 415 (1916); Columbia Mill Co. v. Alcorn, 150 U.S. 460, 463–64 (1893); Trade-Mark Cases, 100 U.S. at 92–94 (holding that a trademark is “[t]he right to adopt and use a symbol or a device to distinguish the goods or property made or sold by the person whose mark it is, to the exclusion of use by all other persons,” which “grows out of its use, and not its mere adoption.”); In re Bose Corp., 546 F.2d 893, 896–97 (C.C.P.A. 1976); 7 Louis Altman & Malla Pollack, Callmann on Unfair Competition, Trademarks & Monopolies § 26:11 (4th ed. 2018) (“The first requirement for registration of matter is that the material covered by the application functions as a mark.” (footnotes omitted)).

  15. [15]. See Joseph Scott Miller, Abercrombie 2.0—Can We Get There from Here? Thoughts on “Suggestive Fair Use, 77 Ohio St. L.J. Furthermore 1, 2 (2016) (“The Abercrombie hierarchy of distinctiveness for trademarks [is] approaching its fortieth birthday with [explicit] citations in more than 850 cases and more than 580 law review pieces.”). Those numbers likely do not reflect the thousands of cases that apply the framework set forth in Abercrombie without explicitly citing to it, and certainly do not account for the thousands or millions of USPTO filewrappers that call on the Abercrombie categories in assessing trademark registrability.

  16. [16]. TMEP, supra note 2, § 1201.

  17. [17]. Barton Beebe describes the linkages between signifier, signified, and referent, including the linkage regulated by use requirements, and notes “case law has substantially liberalized [them] ... [t]he trademark, in short, is falling apart.” Barton Beebe, The Semiotic Account of Trademark Doctrine and Trademark Culture, in Trademark Law & Theory: A Handbook of Contemporary Research 49–50 (Graeme B. Dinwoodie & Mark D. Janis eds., 2008).

  18. [18]. Abercrombie & Fitch Co. v. Hunting World,Inc., 537 F.2d 4 (2d Cir. 1976). The well-known Second Circuit case articulated the trademark distinctiveness hierarchy that continues to dominate distinctiveness and strength analyses to this day. In this Article, “Abercrombie” refers to that scale of distinctiveness, often described as a spectrum or inverted pyramid.

  19. [19]. See, e.g., Thomas R. Lee et al.,An Empirical and Consumer Psychology Analysis of
    Trademark Distinctiveness
    , 41 Ariz. St. L.J. 1033, 1060 (2009) (“[T]rademarkcommentators
    have,forthemostpart,unquestioninglyembracedthe Abercrombie taxonomy and its underlying assumptions about consumer perceptions of meaning.”).

  20. [20]. See Margreth Barrett, Finding Trademark Use: The Historical Foundation for Limiting Infringement Liability to Uses “In the Manner of a Mark, 43 Wake Forest L. Rev. 893, 903–04 (2008) (citing Restatement (First) of Torts § 715 (Am. Law Inst. 1939)).

  21. [21]. Id. at 903–05 (citing Restatement (First) of Torts §§ 715–716).

  22. [22]. Id. at 905–11.

  23. [23]. For example, printing the word “crazy” on a tag inside a t-shirt would be considered use as a mark, while displaying the same word in large font on the front of the shirt would be deemed merely ornamental use (and thus not use as a mark).

  24. [24]. Lee et al., supra note 19, at 1060.

  25. [25]. A number of prominent scholars are engaged in ongoing debate about defendants’ trademark use, a debate that informs this Article. The focus of that debate is whether and when infringement plaintiffs ought to be required to show that defendants used as a mark matter confusingly similar to plaintiffs’ own trademark, not in the context of plaintiffs’ or registrants’ use. See Margreth Barrett, Internet Trademark Suits and the Demise of “Trademark Use,39 U.C. Davis L. Rev. 371, 373 (2006); Graeme B. Dinwoodie & Mark D. Janis, Confusion over Use: Contextualism in Trademark Law, 92 Iowa L. Rev. 1597, 1599 (2007); Stacey L. Dogan & Mark A. Lemley, Grounding Trademark Law Through Trademark Use, 92 Iowa L. Rev. 1669, 1670 (2007); Mark P. McKenna, Trademark Use and the Problem of Source, 2009 U. Ill. L. Rev. 773, 774; Uli Widmaier, Use, Liability, and the Structure of Trademark Law, 33 Hofstra L. Rev. 603, 607–08 (2004).

  26. [26]. “Foundational critiques of modern trademark law tend not to address the role of registration.” Rebecca Tushnet, Registering Disagreement: Registration in Modern American Trademark Law, 130 Harv. L. Rev. 867, 870 (2017).

  27. [27]. It is a matter of black letter law that trademark rights are a function of a party’s use of a mark. McKenna, supra note 25, at 779 (“One acquires rights ... through use of the mark in connection with particular goods or services, and priority as between competing users is determined by first use.”); Widmaier, supra note 25, at 606 (“[T]he concept of use ... is both a plain and simple statutory requirement and the foundational structural principle of American trademark law.”); Robert C. Cumbow, United States: Registrable Trademark or Merely Ornamental?, Mondaq,
    +or+Merely+Ornamental (last updated Jan. 28, 2003) (“It’s fundamental that trademark registration is available to only those words, phrases, designs or combinations that functionastrademarks.”).

  28. [28]. Ownership of a valid trademark is a prerequisite for a trademark infringement action. 15 U.S.C. § 1114(1)(a) (2012) (requiring for liability the “use in commerce [of] any reproduction, counterfeit, copy, or colorable imitation of a registered mark “ (emphasis added)); Rock & Roll Hall of Fame & Museum, Inc. v. Gentile Prods.,134 F.3d 749, 753 (6th Cir. 1998) (“A plaintiff must show that it has actually used the designation at issueas a trademark” and thus the designation or phrase must be used to “perform[ ] the trademark function of identifying the source of the merchandise to the customers.” (emphasis added)); Union Nat’l Bank of Texas, Laredo, Tex. v. Union Nat’l Bank of Texas, Austin, Tex., 909 F.2d 839, 844 (5th Cir. 1990) (“The threshold issue in any action for trademark infringement is whether the word or phrase is initially registerable or protectable.” (quoting Zatarains, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786, 790 (5th Cir. 1983))). But valid U.S. trademark rights are not required to serve as the basis for an unfair competition or false association claim under Section 43(a) of the Lanham Act. Belmora LLC v. Bayer Consumer Care AG, 819 F.3d 697, 708 (4th Cir. 2016) (“Nothing in Lexmark can be read to suggest that § 43(a) claims have an unstated requirement that the plaintiff have first used its own mark (word, term, name, symbol, or device) in U.S. commerce before a cause of action will lie against a defendant who is breaching the statute.” (citing Lexmark Int’l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 129 (2014))).

  29. [29]. “[T]rademark use disputes frequently center on determining when use occurred, not whether use occurred.” Graeme B. Dinwoodie & Mark D. Janis, Trademarks and Unfair Competition: Law & Policy 251 (4th ed. 2014). But see Gamers, Inc. v. Game-Xpert, Inc., Opposition No. 91164969, 2008 WL 3873421, at *3 (T.T.A.B. Aug. 6, 2008) (“[E]stablishing priority is different from establishing use to support registration. For example, while trade name usage is not sufficient to support an application for registration, it is sufficient to establish priority in an inter partes dispute.”).

  30. [30]. For more on the scourge of trademark bullying, see generally Stacey Dogan, Bullying and Opportunism in Trademark and Right-of-Publicity Law, 96 B.U. L. Rev. 1293 (2016) (arguing that certain features of trademark law enable and even reward bullying); Leah Chan Grinvald, Shaming Trademark Bullies, 2011 Wis. L. Rev. 625 (proposing that shaming can be an effective way to combat trademark bullying); Jessica M. Kiser, To Bully or Not to Bully: Understanding the Role of Uncertainty in Trademark Enforcement Decisions, 37 Colum. J.L. & Arts 211 (2014) (identifying how uncertain duties and inherent cognitive biases interact to culminate in trademark bullying); Irina D. Manta, Bearing Down on Trademark Bullies, 22 Fordham Intell. Prop. Media & Ent. L.J. 853 (2012) (proposing a model to prevent frivolous claims while still allowing trademark owners to record evidence of policing); Kenneth L. Port, Trademark Extortion: The End of Trademark Law, 65 Wash. & Lee L. Rev. 585 (2008) (concluding that trademark extortion and strike suits deter market entrants, profoundly changing trademark litigation); Jeremy N. Sheff, Fear and Loathing in Trademark Enforcement, 22 Fordham Intell. Prop. Media & Ent. L.J. 873 (2012) (exploring the legal and psychological justifications for aggressive trademark enforcement).

  31. [31]. Small business owners, in particular artists competing in the “print on demand” industry, are increasingly banding together to contest “frivolous” or overreaching trademark applications for matter that fails to function by filing Letters of Protest with the USPTO Commissioner. See E-mail from Morgan Reece, Author/Advocate, to Alexandra J. Roberts, Assoc. Professor, Univ. of N.H. Sch. of Law (Oct. 14, 2018) (on file with author) (discussing membership of “Makers Against Frivolous Trademarks,” a group that includes 4,800 small business owners and other members as of February 27, 2019, and “Trademark Watch Dawgs,” a group that includes 7,300 small business owners and other members as of February 27, 2019).

  32. [32]. Beebe & Fromer,supra note 13, at 978; Alexandra J. Roberts,Howto DoThingswithWordMarks: A Speech-Act Theory of Distinctiveness, 65 Ala. L. Rev. 1035, 1055 (2014) (explaining that when the USPTO and courts grant trademark protection for matter that should not qualify for it “they risk depleting the language available to competitors and chilling speech by granting a property right in a term that serves no trademark function”).

  33. [33]. Tushnet, supra note 26, at 871–72.

  34. [34]. See, e.g., Park ‘N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189 (1985); Rebecca Tushnet, Fixing Incontestability: The Next Frontier?, 23 B.U. J. Sci. & Tech. L. 434, 438–39 (2017).

  35. [35]. Jonathan Masur, Patent Inflation, 121 Yale L.J. 470, 472, 473–75 (2011) (arguing that the nature of the patent prosecution process creates an asymmetric incentive for the USPTO to grant rather than refuse applications to minimize appeals and reversals).

  36. [36]. B & B Hardware, Inc. v. Hargis Indus., Inc., 135 S. Ct. 1293, 1309–10 (2015) (“So long as the other ordinary elements of issue preclusion are met, when the usages adjudicated by the TTAB are materially the same as those before the district court, issue preclusion should apply.”).

  37. [37]. Linda Hamilton Krieger & Susan T. Fiske,Behavioral Realism in Employment Discrimination Law: Implicit Bias and Disparate Treatment, 94 Cal. L. Rev. 997, 999 (2006).

  38. [38]. See Alan Durant, ‘How Can I Tell the Trade Mark on a Piece of Gingerbread from All the Other Marks on It?’ Naming and Meaning in Verbal Trade Mark Signs, in Trade Marks and Brands: An Interdisciplinary Critique 109 (Bently et al. eds., 2008) (“[N]otions such as ‘use in a context’ and correspondingly ‘comprehension in a context’ are as relevant in analysing trade mark problems as an abstract, semiotic scale. Commercial signs require trade mark protection when used in particular discourse contexts ... rather than in the abstract. What they mean [and] what effects they have ... all depend on how they are used as much as on the inherent meaning potential of the signs themselves.”).

  39. [39]. “[C]ourts regularly say that the question [of] whether a symbol meets the standards for protectability as a mark is the same as whether it meets the standards for registrability.” Tushnet, supra note 26, at 881.

  40. [40]. While inherent distinctiveness is treated as binary for the purposes of protectability assessment, courts and commentators often deem suggestive marks “less inherently distinctive” or weaker than fanciful or arbitrary marks. See Fact Sheets: Introduction to Trademarks, Int’l Trademark Ass’n,
    FactSheet.aspx (last updated Aug. 2014).

  41. [41]. Appendix A offers additional discussion of the chart and includes a second version that maps examples from the paper onto it.

  42. [42]. In re Bose Corp., 546 F.2d 893, 896 (C.C.P.A. 1976) (emphasis added) (citing In re Standard Oil Co., 275 F.2d 945, 947 (C.C.P.A. 1960)); see also Clairol Inc. v. Gillette Co., 389 F.2d 264, 269 (2d Cir. 1968) (“It is elementary that the function of a trademark is to indicate the origin of the products to which it is attached ... . Clairol must at least show that the term ‘Innocent Beige’ serves this function in order to obtain protection of it as a valid trademark.”); Tushnet, supra note 26, at 876 (“Trademark registration is available when the PTO determines that a symbol is functioning as a trademark by indicating the source of a product or service (or will immediately do so upon use, in the case of an ITU application).”).

  43. [43]. Dinwoodie & Janis, supra note 29, at 247 (“Either actual or constructive trademark use is a prerequisite for applying for rights under a federal registration.”). While a registration will not be granted until the applicant demonstrates actual use, applications based on an intent to use a trademark in the future will be examined and published, and registrations benefit from the priority date of the ITU application under Section 1(b). See 15 U.S.C. § 1051(b) (2012).

  44. [44]. MicroStrategy Inc. v. Motorola, Inc., 245 F.3d 335, 341 (4th Cir. 2001) (“The Lanham Act ... does require that in order to obtain trademark protection ‘a designation must be proven to perform the job of identification: to identify one source and distinguish it from other sources.’” (quoting McCarthy, supra note 9, § 3:3)); see also Graeme B. Dinwoodie & Mark. D. Janis, Use, Intent to Use & Registration in the USA, in Trade Mark Use, supra note 11, at 316 (“US courts generally apply the same methodology to questions of actual use whether the trade mark claimant is claiming registered or (unregistered) common law rights.”).

  45. [45]. In most jurisdictions, trademark rights stem from registration rather than use. Before the U.S. enacted the ITU provision, it was the only country other than the Philippines “to require use as a prerequisite for the protection of trademarks.” Rudolf Rayle, The Trend Towards Enhancing Trademark Owners’ Rights–A Comparative Study of U.S. and German Trademark Law, 7 J. Intell. Prop. L. 227, 245–46 (2000) (citing Robert J. Eck, Statement of the United States Trademark Association in Support of S. 1883 (DeConcini), The Trademark Law Revision Act, 78 Trademark Rep. 382, 387 (1988)). While most other countries do not require use before registration,they do require that owners of registered marks eventually use them as marks, and marks canbe cancelled for nonuse. See Susan Neuberger Weller, Use It or Lose It: When Can a Trademark Registered Under Section 44(e) or 66(a) Be Deemed “Abandoned” in the US?, Mintz (Feb. 3, 2015),

  46. [46]. Foreign applicants can also obtain U.S. trademark registrations without actually making use in the United States. See 15 U.S.C. § 1051(e).

  47. [47]. The applicant’s statement of bona fide intent to use the mark on the listed goods and services in the future is considered “constructive use” sufficient to put potential competitors on notice and support a claim of priority dating back to the filing of the ITU application. See id.
    § 1057. The applicant has six months to demonstrate actual use or seek an extension, with a maximum of five extensions available after that initial period, and will not obtain a registration until they file a statement of use and appropriate specimens. See Jake Linford, Trademark Owner as Adverse Possessor: Productive Use and Property Acquisition, 63 Case W. L. Rev. 703, 722 (2013).

  48. [48]. McCarthy, supra note 9, § 3:3 (“Some of the common markers of whether a word, phrase or picture is being used as a trademark are: larger-sized print, all capital letters or initial capitals, distinctive or different print style, color, and prominent position on label or advertising copy.”).

  49. [49]. Where multiple specimens are available, factfinders also consider whether use is consistent. See, e.g., MicroStrategy Inc., 245 F.3d at 341 (finding that plaintiff’s submission of 24 separate documents did not demonstrate use as a mark of “Intelligence Everywhere,” where use was not consistent across documents and “follow[ed] no particular design or sequence”).

  50. [50]. Dunkin’ Brands owns federal registrations for the trademarks DUNKIN’ DONUTS, the design mark shown, and AMERICA RUNS ON DUNKIN’, among many others.

  51. [51]. The temperature warning likely fails to function both because it isn’t used in a trademark way and because it constitutes unprotectable informational matter. See TMEP, supra note 2, § 1202.04.

  52. [52]. 15 U.S.C. § 1127 (2012); In re Michelin N. Am., Inc., Serial No. 85077031, 2015 WL 5675633, at *2 (T.T.A.B. Sept. 11, 2015); Nextel Commc’ns, Inc. v. Motorola, Inc., Opposition No. 91161817, 2008 WL 902843, at *1, *6 (T.T.A.B. Feb. 27, 2008) (“The starting point for our analysis is Section 45 of the Lanham Act, as amended, where ‘trademark’ is defined ... .”); TMEP, supra note 2, § 904.07(b) (“The statutory bases for [failure to function] refusal are §§ 1, 2, 4, and 45 of the Trademark Act.”). But see In re Esso Standard Oil Co., 305 F.2d 495, 500 (C.C.P.A. 1962) (Kirkpatrick, J., dissenting) (“Although the Lanham Act does not spell out in its definition of the term ‘trademark’, found in Section 45, a requirement that the symbol or device must be such as will be recognized by the purchasing public as a trademark, it is obviously not the intention of the Lanham Act to eliminate such fundamental requirement for trademark significance.”).

  53. [53]. Dinwoodie & Janis, supra note 29, at 217 (“Long before the advent of the modern federal trademark regime, the common law of trademarks in the United States required trademark use as a precondition to trademark ownership.”); McKenna, supra note 25, at 788 (“To avoid interfering with legitimate uses, courts had to differentiate between uses of a term in its source-designating (trademark) capacity and uses of the term in its ordinary, non-source-designating capacity.” (citation omitted)).

  54. [54]. 15 U.S.C. § 1127 (emphasis added). “Service mark” is defined similarly in the same section. See id.

  55. [55]. To merit protection, trademark use must be in connection with the goods or services for which trademark rights are asserted. See supra note 10 and accompanying text. That technical requirement at times overlaps with either the use in commerce or the use as a mark requirement or both, but a trademark use can fall short on one, two, or all three of those fronts. Id. If a producer applies to register a trademark and the specimen provided does not meet the requirements articulated in the Lanham Act and specified in detail in the TMEP and case law to demonstrate use with particular goods or services, the USPTO will issue a “specimen rejection.” See generally In re Michelin, 2015 WL 5675633 (distinguishing a specimen rejection from a refusal to register a mark based on the applicant’s failure to demonstrate its use as a mark). It is important to note that not all courts and commentators distinguish use in connection with goods or services from use as a mark. See, e.g., In re Brunswick Corp., Serial No. 78875524, 2008 WL 4155515, at *3 (T.T.A.B. June 25, 2008) (intertwining the two analyses in discussion even though the refusal to register is affirmed entirely based on the applicant’s failure to use the desired mark specifically in connection with boats, rather than its failure to show use as a mark).

  56. [56]. TMEP, supra note 2, Foreword (“The [TMEP] is published to provide trademark examining attorneys in the USPTO, trademark applicants, and attorneys and representatives for trademark applicants with a reference work on the practices and procedures relative to prosecution of applications to register marks in the USPTO. The Manual contains guidelines for Examining Attorneys and materials in the nature of information and interpretation and outlines the procedures which Examining Attorneys are required or authorized to follow in the examination of trademark applications.”).

  57. [57]. “[S]ource distinctiveness describes the extent to which a trademark’s signifier is distinctive of its signified,” while “differential distinctiveness describes the extent to which a trademark’s signifier is distinctive from other signifiers in the trademark system.” Barton Beebe, The Semiotic Analysis of Trademark Law, 51 UCLA L. Rev. 621, 625 (2004).

  58. [58]. See TMEP, supra note 2, § 904.07(b) (“The examining attorney [will] evaluate the specimen to determine whether the applied-for mark is used in a way that shows that:(1) the applied-for mark identifies the goods/services of the applicant and distinguishes them from the goods/services of others; and (2) the applied-for mark indicates the source of those goods/services. If use on the specimen fails in either regard, the record lacks the requisite evidence that the applied-for mark functions as a mark.” (citation omitted)).

  59. [59]. Dogan & Lemley, supra note 25, at 1673–74 (“[N]either judges nor scholars have yet articulated a satisfying definition of trademark use.”).

  60. [60]. Scholastic Inc. v. Speirs, 28 F. Supp. 2d 862, 870 (S.D.N.Y. 1998), aff’d mem., 199 F.3d 1323 (2d Cir. 1999) (quoting treatise with approval); McCarthy, supra note 9, § 3:3; see also Dep’t of Parks & Recreation v. Bazaar Del Mundo Inc., 448 F.3d 1118, 1127 (9th Cir. 2006) (finding that because use in brochures was not “designed to attract the attention of the viewer to the marks themselves, they fail to create any association between the marks and the ... services”).

  61. [61]. New W. Corp. v. NYM Co. of Cal., 595 F.2d 1194, 1200 (9th Cir. 1979).

  62. [62]. Fleischer Studios, Inc. v. A.V.E.L.A., Inc., 925 F. Supp. 2d 1067, 1072 (C.D. Cal. 2012) (citing Proctor & Gamble Co. v. Keystone Auto. Warehouse, Inc., 191 U.S.P.Q.2d (BNA) 468, 1976 WL 20910, at *6 (T.T.A.B. 1976)).

  63. [63]. In re Aerospace Optics, Inc., 78 U.S.P.Q.2d (BNA) 1861, 2006 WL 1087849, at *2 (T.T.A.B. 2006); In re Remington Prods., Inc., 3 U.S.P.Q.2d (BNA) 1714, 1987 WL 124304, at *1 (T.T.A.B. 1987); see also Restatement (Third) of Unfair Competition § 9 (Am. Law Inst. 1995) (stating that a trademark must be “used in a manner that identifies [a producer’s] goods or services and distinguishes them from the goods or services of others”).

  64. [64]. Barrett, supra note 25, at 375; see also Barrett, supra note 20, at 894 (“‘Trademark use’ can be generally understood as use of a word or symbol in close association with goods or services being offered for sale, in a manner that is likely to communicate the source of those goods or services to consumers.”).

  65. [65]. McKenna, supra note 25, at 779.

  66. [66]. McCarthy, supra note 9, § 3:4.

  67. [67]. The phrase “trademark use” can also refer to a defendant’s use of a confusing or diluting term or phrase in the context of litigation. See Barrett, supra note 25, at 373; Dinwoodie & Janis, supra note 25, at 1626 n.133; Dogan & Lemley, supra note 25, at 1701; McKenna, supra note 25, at 776. Courts and commentators are divided on whether a defendant is required to make use of a term or phrase as a mark in order to be liable for infringement or dilution. See Stacey L. Dogan & Mark A. Lemley, The Trademark Use Requirement in Dilution Cases, Santa Clara High Tech. L.J. 542–45 (2008) (discussing the varying views on trademark use and liability for infringement or dilution). Those who deem the requirement a crucial check on trademark rights expansion and over-enforcement position it as a parallel or mirror-image of the affirmative use required for protection. See id. at 544. Unlike a registrant’s or plaintiff’s use, a defendant’s use need not necessarily be consumer-facing to be deemed trademark use. See id. at 545. The concept of use as a mark also comes into play when infringement defendants assert the affirmative defense of descriptive fair use, also known as statutory fair use. Id. at 542–43. A party arguing that its use constitutes a descriptive fair use and is thus non-infringing must establish that, among other requirements, it used the matter in question other than as a mark. See, e.g., Cosmetically Sealed Indus., Inc. v. Chesebrough-Pond’s USA Co., 125 F.3d 28, 30 (2d Cir. 1997).

  68. [68]. According to McCarthy’s treatise, “to create trademark or trade dress rights, a designation must be proven to perform the job of identification: to identify one source and distinguish it from other sources.” McCarthy, supra note 9, § 3:4.

  69. [69]. McKenna, supra note 25, at 773, 776.

  70. [70]. Beebe, supra note 57, at 680.

  71. [71]. The 1984 amendments to the Lanham Act codified the anonymoussource rule by revising the definitions of trademark and service mark to include indicating thesourceof the goods or services “even if thatsourceis unknown.” 15 U.S.C. § 1127 (2012); A.J. Canfield Co. v. Honickman, 808 F.2d 291, 300 (3d Cir. 1986)(explaining how the 1984 amendment codifies “the long-recognized anonymoussourcerule”).

  72. [72]. See Graeme W. Austin,Trademarks and the Burdened Imagination, 69 Brook. L. Rev. 827, 844 n.70 (2004)(citingManhattan Shirt Co. v. Sarnoff-Irving Hat Stores, Inc., 164 A. 246, 250 (Del. Ch. 1933)); Beebe, supra note 57, at 663; Deven R. Desai, From Trademarks to Brands, 64 Fla. L. Rev. 981, 1011 (2012); Jeremy N. Sheff, The (Boundedly) Rational Basis of Trademark Liability, 15 Tex. Intell. Prop. L.J. 331, 344 n.80 (2007).

  73. [73]. In re Eagle Crest, Inc., 96 U.S.P.Q.2d (BNA) 1227, 2010 WL 3441109, at *2 (T.T.A.B. 2010); see also In re Aerospace Optics, Inc., 78 U.S.P.Q.2d (BNA) 1861, 2006 WL 1087849, at *2 (T.T.A.B. 2006) (stating “[a] critical element in determining whether matter sought to be registered is a trademark is the impression the matter makes on the relevant public”).

  74. [74]. See Kelly Servs., Inc. v. Creative Harbor, LLC, 124 F. Supp. 3d 768, 775 (E.D. Mich. 2015) (“‘[T]here has to be an “open” use, that is to say, a use has to be made to the relevant class of purchasers or prospective purchasers... . [A]n “internal” use ... cannot give rise’ to priority rights to a mark.” (citing Sterling Drug, Inc. v. Knoll A.G. Chemische Fabriken, 159 U.S.P.Q. (BNA) 628, 631 (T.T.A.B. 1968))); see also Widmaier, supra note 25, at 626 (2004) (“Courts consistently require open and visible use of the mark in close proximity to the goods sold under it so as to permit consumers to perceive goods and mark together and thus to form a source-indicating association between the two.”).

  75. [75]. Blue Bell, Inc. v. Farah Mfg. Co., 508 F.2d 1260, 1265 (5th Cir. 1975) (stating “ownership of a trademark accrues when goods bearing the mark are placed on the market”); Kelly Servs., Inc., 124 F. Supp. 3d at 776 (“But Kelly has not shown that its submission of [the product bearing the trademark] to Apple was sufficiently open or public to identify or distinguish its application in the minds of consumers. To the contrary, the bilateral exchange between Kelly and iTunes Connect provided no notice of the [branded product] to potential consumers—i.e., persons who might eventually download [it] from the Apple App Store.”).

  76. [76]. Aycock Eng’g, Inc. v. Airflite, Inc.,560 F.3d 1350, 1360 (Fed. Cir. 2009); In re Cedar Point, Inc., 220 U.S.P.Q. (BNA) 533, 1983 WL 51881, at *3 (T.T.A.B. 1983); Intermed Commc’ns, Inc. v. Chaney, 197 U.S.P.Q. (BNA) 501, 1977 WL 22648, at *6 (T.T.A.B. 1977).

  77. [77]. Analogous use doctrine provides a narrow exception to this rule, allowing producers to tack early use that technically falls short of requirements onto later use for purposes of priority. FN Herstal S.A. v. Clyde Armory Inc., 838 F.3d 1071, 1081 (11th Cir. 2016). But even analogous use must be outward facing: Qualifying pre-sales activity must be “of such a nature and extent as to create an association in the mind of the consuming public between [t]he mark and the services to be rendered.” Selfway, Inc. v. Travelers Petroleum, Inc., 579 F.2d 75, 79 (C.C.P.A. 1978); see also T.A.B. Sys. v. Pactel Teletrac, 77F.3d1372, 1376 (Fed. Cir. 1996) (explaining that for a use to qualify as analogous use, the court must “determine whether it was sufficiently clear, widespread and repetitive to create the required association in the minds of potential purchasers between the mark as an indicator of a particular source and the service to become available later”). The analogous use doctrine therefore arguably allows use as a mark to rescue matter that has not been sufficiently used in commerce, so long as it is eventually used in commerce, by granting the user priority based on its earlier pre-commerce use as a mark.

  78. [78]. Mountain Top Beverage Grp., Inc. v. Wildlife Brewing N.B., Inc.,338 F. Supp. 2d 827, 835 (S.D. Ohio 2003),aff’d, 432 F.3d 651 (6th Cir. 2005) (citation omitted).

  79. [79]. This Article focuses on failure to function refusals that are context-specific, i.e. dependent upon how matter is used. However, the USPTO often employs the same phrase when it refuses registration for matter that it deems incapable of functioning as a mark regardless of how it is used. For example, TMEP Section 904.07(b) provides that ubiquitous phrases or symbols may be deemed informational slogans or incapable matter and presumes that consumers will never associate them with a single source. TMEP, supra note 2, § 904.07(b); see also, e.g., In re Volvo Cars of N. Am., Inc., 46 U.S.P.Q.2d (BNA) 1455, 1998 WL 239298, at *1 (T.T.A.B. 1998) (finding “DRIVE SAFELY” incapable of functioning as a mark for automobiles and parts);In re Remington Products, Inc., 3 U.S.P.Q.2d (BNA) 1714, 1987 WL 124304, at *1 (T.T.A.B. 1987)(stating that “PROUDLY MADE IN THE USA” was incapable of functioning as a mark for electric shavers).

  80. [80]. See, e.g.,In re Melville Corp., 228 U.S.P.Q. (BNA) 970, 1986 WL 83650, at *3 n.2 (T.T.A.B. 1986)(“If matter proposed for registration does not function as a mark, it is not registrable in accordance with Sections 1 and 2 of the Act because the preambles of those sections limit registration to subject matter within the definition of a trademark.”).

  81. [81]. SeeSengoku Works Ltd. v. RMC Int’l, Ltd.,96 F.3d 1217, 1219 (9th Cir. 1996)(“To acquire ownership of a trademark it is not enough to have invented the mark first or even to have registered it first; the party claiming ownership must have been the first to actually use the mark in the sale of goods or services.”);AB Electrolux v. Bermil Indus. Corp.,481 F. Supp. 2d 325, 330 (S.D.N.Y. 2007); Bell v. Streetwise Records, Ltd., 640 F. Supp. 575, 580 (D. Mass. 1986) (“Priority is established not by conception but by bona fide usage.”).It is important to note, however, that use insufficient to satisfy registration requirements may nonetheless suffice to establish priority. See, e.g., Gamers, Inc. v. Game-Xpert, Inc., Opposition No. 91164969, 2008 WL 3873421, at *3 (T.T.A.B. Aug. 6, 2008); Taboca AS v. Philip Morris USA Inc., Opposition No. 91172571, 2007 WL 4616272, at *3 (T.T.A.B. Dec. 18, 2007); TuTorTape Labs., Inc. v. Halvorson, 155 U.S.P.Q. (BNA) 268, 1967 WL 7265, at *4 (T.T.A.B. 1967). But see Exec. Coach Builders, Inc. v. SPV Coach Co., 123 U.S.P.Q.2d (BNA) 1175, 2017 WL 3034059, at *24 (T.T.A.B. 2017) (holding that domain name usage has no evidentiary weight on the issue of priority).

  82. [82]. USPTO, Protecting Your Trademark: Enhancing Your Rights Through Federal Registration 24–25 (2018),

  83. [83]. See TMEP, supra note 2, §§ 1202, 1202.02(b)(ii).

  84. [84]. From 2003 through 2015, more than half of all applications that received a failure to function refusal—approximately 60% of them—were ultimately published for opposition, a rate that is lower than the publication rate among the general pool of applications, but far from low. See Appendix B (providing data on the number of applications receiving failure to function refusals each year, the publication rate for that subset of applications, and the publication rate for the broader pool, all extracted from Beebe and Fromer’s data set).

  85. [85]. The latter approach is less common and changes the posture of the case to list the USPTO Director as the respondent.

  86. [86]. See TIPS for Responding to a USPTO Office Action, USPTO (last updated Feb. 11, 2016),

  87. [87]. See id.

  88. [88]. See TMEP, supra note 2, §§ 701–720.

  89. [89]. See infra Section II.D.

  90. [90]. The decisions of other examining attorneys are neither binding nor searchable, rendering inconsistency across applications inevitable. On the value of consistency, see Michael Grynberg,The Judicial Role in Trademark Law, 52 B.C. L. Rev. 1283, 1305 (2011) (“Even if judges do no more than apply heuristics of questionable quality to the disposition of trademark claims,channeling the process through a consistent framework aids litigants in identifying and accommodating the factors that guide fact finding.”).

  91. [91]. Masur, supra note 35, at 470, 472–75.

  92. [92]. Examining attorneys also have quotas and must review a high volume of applications quickly. The rule of doubt often leads them to side with the applicant in ex parte decisions because parties with standing have the opportunity to oppose those decisions upon publication. See, e.g., In re Mavety Media Grp. Ltd., 33 F.3d 1367, 1374 (Fed. Cir. 1994), abrogated by In re Tam, 808 F.3d 1321 (Fed. Cir. 2015) (finding doubt as to whether a mark is scandalous is resolved in the applicant’s favor); In re Merrill Lynch, Pierce, Fenner, & Smith, Inc., 828 F.2d 1567, 1571 (Fed. Cir. 1987) (finding doubt as to whether a mark is generic is resolved in the applicant’s favor); In re, Inc., Serial No. 76001834, 2004 WL 2368420, at *3 (T.T.A.B. Sept. 29, 2004) (finding doubt as to whether a mark is suggestive or descriptive is resolved in the applicant’s favor); In re Jose Remacle, 66 U.S.P.Q.2d (BNA) 1222, 2002 WL 31563187, at *2–3 (T.T.A.B. 2002) (concluding that where there are doubts due to inconclusive evidence, such doubts are resolved in the applicant’s favor).

  93. [93]. See Barton Beebe, Is the Trademark Office a Rubber Stamp?, 48 Hous. L. Rev. 751, 763–64 (2011) (tracking publication and registration rates for trademark applications by mark type and application type).

  94. [94]. Id. at 764. Publication is followed by registration unless the application is opposed or abandoned.

  95. [95]. Stuart Graham et al., The USPTO Trademark Case Files Dataset: Descriptions, Lessons, and Insights 46 (2013),

  96. [97]. Beebe,supra note 93, at 769–70 (“Generally speaking, the fact that the mark has been published in the Official Gazette shows that that mark has survived PTO examination.”).

  97. [98]. One way to increase the level of scrutiny applied to specimens of use submitted to perfect ITU-based applications would be to wait to publish the mark for opposition until after the statement of use and specimens are filed, or else to reopen the opposition period at that stage so that anyone with standing can oppose a mark based on failure to function.

  98. [99]. TMEP, supra note 2, § 904.07(b).

  99. [100]. 15 U.S.C. §§ 1051, 1052, 1127 (2012); see also id. §§ 1051, 1053, 1127 (regarding service mark failure to function refusals).

  100. [101]. See, e.g., In re Shenandoah Growers, Inc., Serial No. 76620753, 2008 WL 885947, at *2 (T.T.A.B. Feb. 14, 2008) (“[T]he overall impression of the phrase ‘For more great Freshherb ideas, visit’ on the specimen is of an invitation to visit a website for ideas on how to use fresh herbs. As such, ‘’ does not function as a source indicator.”).

  101. [102]. See Alexandra J. Roberts, Tagmarks, 105 Cal. L. Rev. 599, 632–41 (2017) (providing examples in which applications to register hashtags as trademarks were refused based on failure to function because the applied-for matter in each was used merely as a hashtag, not as a trademark).

  102. [103]. In re Michelin N. Am., Inc., Serial No. 85077031, 2015 WL 5675633, at *1 (T.T.A.B. Sept. 11, 2015).

  103. [104]. In re West-Com Nurse Call Sys., Inc., Serial No. 78438594, 2007 WL 2972207, at *11 (T.T.A.B. Sept. 13, 2007) (“The proposed Toilet Symbolfails to function as a trademark and is merely a pictorial representation of an important feature or function of the goods.”).

  104. [105]. In re Osterberg, 83 U.S.P.Q.2d (BNA) 1220, 2007 WL 616032, at *3 (T.T.A.B. 2007) (holding CONDOMTOY CONDOM fails to function as a mark for condoms, and noting “because the mark ... appears as part of the sentence, ‘That’s why Inspiral is also called a CondomToy® condom,’ the commercial impression that is conveyed, particularly because of the use of the indefinite article ‘a’ before CondomToy® condom, is that CondomToy condom is a descriptive term for condoms sold under the Inspiral trademark”); In re Aerospace Optics, Inc., 78 U.S.P.Q.2d (BNA) 1861, 2006 WL 1087849, at *3 (T.T.A.B. 2006) (finding SPECTRUM failed to function as a mark where “prospective purchasers would not view [it] as a source identifier, but rather as part of the text listing the features of the product”).

  105. [106]. J.M. Hollister, LLC v. Am. Eagle Outfitters, No. 203CV703, 2005 WL 1076246, at *2–3 (S.D. Ohio May 5, 2005); In re Frankish Enters. Ltd., 113 U.S.P.Q.2d (BNA) 1964, 2015 WL 1227728, at *7 (T.T.A.B. 2015).

  106. [107]. The list includes use as a trade name; use thatidentifies the name or pseudonym of a performing artist or author; use as a character name; and use as a model number or grade designation. But the TMEP also offers examples of trade names, model numbers, grade designations, and artists’ namesthat do successfully function as marks. See, e.g., TMEP, supra note 2, §§ 1202.09(a)(ii)(B), 1202.16.

  107. [108]. Exceptions include generic matter and matter that the USPTO deems “merely informational matter,” such as a widely-used symbol or slogan or well-known religious phrase. TMEP, supra note 2, § 1202.04; USPTO, Examination Guide 2-17: Merely Informational Matter (2017), available at

  108. [109]. See, e.g., Go Pro, Ltd. v. River Graphics, Inc., No. CIVA01CV600JLK, 2006 WL 898147, at *4 (D. Colo. Apr. 5, 2006) (“A trademark ... must indicate source and may, if appropriate, indicate source and be ornamental at the same time.”).

  109. [110]. Widmaier, supra note 25, at 611.

  110. [111]. Ex parte Nat’l Geographic Soc’y, 83 U.S.P.Q. (BNA) 260, 1949 WL 3854, at *1 (Dec. Comm’r Pat. 1949).

  111. [112]. McCarthy, supra note 9, § 3:3.

  112. [113]. Id. At the same time, courts don’t seem to expect consumers to make the distinction themselves. See, e.g., In re West-Com Nurse Call Sys., Inc., Serial No. 78438594, 2007 WL 2972207, at *10 (T.T.A.B. Sept. 13, 2007) (noting that “[t]he absence of any indication that the [relevant consumers] know or are so well versed in trademark law that they could independently form a reliable opinion on whether a designation like applicant’s toilet caricature design is a mark or as used ... functions as a mark”).

  113. [114]. MicroStrategy Inc. v. Motorola, Inc., 245 F.3d 335, 341 (4th Cir. 2001) (quoting In re Morganroth, 208 U.S.P.Q. (BNA) 284, 1980 WL 30167, at *4 (T.T.A.B. 1980)).

  114. [115]. See Timothy Greene,Trademark Hybridity and Brand Protection, 46 Loy. U. Chi. L.J. 75, 80 (2014) (“It is not just the term used that signifies source, but the term when used in combination with relevant disambiguating context, such as colors, typefaces, product packaging, logos, logo placement on packaging, product categories, and so on... . Rarely (if ever) do consumers encounter trademarks acontextually.”); Laura A. Heymann, Naming, Identity, and Trademark Law, 86 Ind. L.J. 381, 398 (2011) (“[I]t is context that indicates whether an unknown term is functioning as a trademark or name or in some other way.”); id. at n.65 (“Color, graphic design, and other nonlexical elements may contribute to (or, indeed, be the primary driver of) this effect ... .”); Roger W. Shuy, Linguistic Thoughts on Trademark Dilution 6 (unpublished manuscript), available at (last visited Mar. 20, 2019) (“Since humans commonly use context to disambiguate and figure out what is meant, it is reasonable to expect them to keep on doing this with trademarks.”).

  115. [116]. McCarthy, supra note 9, § 3:4.

  116. [117]. In re Osterberg, 83 U.S.P.Q.2d (BNA) 1220, 2007 WL 616032, at *3 (T.T.A.B. 2007) (quoting CondomToy’s website) (emphasis omitted).

  117. [118]. In re Aerospace Optics, Inc., 78 U.S.P.Q.2d (BNA) 1861, 2006 WL 1087849, at *2 (T.T.A.B. 2006) (referring to the VIVISUN LED image in the text of the opinion).

  118. [119]. In re Blackstone Audio Inc., Serial No. 85761414, 2016 WL 5407750, at *2 (T.T.A.B. Sept. 16, 2016). Similarly, the Board affirmed the USPTO’s refusal to register SEMICONDUCTOR LIGHT MATRIX for light curing systems, predicting that based on the specimen submitted, consumers would perceive the phrase as identifying the product’s component technology rather than its source. In re Phoseon Tech., Inc., 103 U.S.P.Q.2d (BNA) 1822, 2012 WL 3854070, at *6–7 (T.T.A.B. 2012).

  119. [120]. But see McCarthy, supra note 9, § 7:27 (“There is no requirement that a mark be used alone and by itself in order to be ‘used.’” (citing Gen. Foods Corp. v. Ito Yokado Co., Ltd., 219 U.S.P.Q. (BNA) 822, 1983 WL 50169, at *2 (T.T.A.B. 1983)).

  120. [121]. Thoroughbred Legends, LLC v. Walt Disney Co., No. 1:07-CV-1275-BBM, 2008 WL 616253, at *6 (N.D. Ga. Feb. 12, 2008).

  121. [122]. See, e.g., E.T. Browne Drug Co. v. Cococare Prods., Inc., 538 F.3d 185, 201 (3d Cir. 2008).

  122. [123]. Sam’s Riverside, Inc. v. Intercon Sols., Inc., 790 F. Supp. 2d 965, 984–85 (S.D. Iowa 2011).

  123. [124]. Id. at 984–85. The court found this failure to provide examples demonstrating use as a mark “particularly troubling because Plaintiff’s counsel presumably cherry-picked these particular advertisements as the best evidence they had in support of Plaintiff’s claim that it used ‘Sam’s Riverside’ as a mark prior to the alleged infringement.” Id. at 984 n.36; see also In re Paris Med. Co., 87 F.2d 484, 485–86 (C.C.P.A. 1937) (“[I]t is not shown that the term ‘Tasteless Chill Tonic,’ standing alone, was ever used in a trade-mark sense, but only as descriptive of the goods to which the term was applied... . [O]n the contrary the specimen shows that the term used was ‘Grove’s Tasteless Chill Tonic,’ and we agree with the commissioner that it was the word ‘Grove’s’ that indicated to purchasers the origin or source of manufacture of the goods upon which the label was applied, and not the term ‘Tasteless Chill Tonic.’”). But see Procter & Gamble Co. v. Keystone Auto. Warehouse, Inc., 191 U.S.P.Q. (BNA) 468, 1976 WL 20910, at *7 (T.T.A.B. 1976) (“It is clear ... the phrase ‘BUMPER TO BUMPER’ is almost always associated with the words ‘CAR CARE’ and design and invariably in conjunction with ‘PROCTER & GAMBLE’ in one form or another. However, considering that ‘PROCTER & GAMBLE’ is opposer’s primary or house mark and would be recognized as such by the average purchaser of opposer’s car waxes and detergents; that the phrase ‘BUMPER TO BUMPER CAR CARE’ and the distinctive car design has been prominently displayed on all of this material; and that purchasers and prospective purchasers of opposer’s products have been exposed to this showing on an increasing scale over a number of years, it is reasonable to assume that the phrase ‘BUMPER TO BUMPER CAR CARE’ and design has created a commercial impression separate and apart from ‘PROCTER & GAMBLE’ and that it serves, in and of itself, as an identification symbol for opposer’s goods ... .”).

  124. [125]. Thomas R. Lee et al., supra note 19, at 1055–57 (“[T]he ‘manner’ of trademark use concerns the size and placement of the mark on the product package or label... . [C]ourts are unwilling to hold that a word or phrase is used as a trademark without a ‘distinctive font, color, typeset or ... other method that makes its nature and function [...] readily apparent and recognizable.’” (quoting Microstrategy Inc. v. Motorola, Inc., 245 F.3d 335, 342 (4th Cir. 1988))).

  125. [126]. Id.; see also Rebecca Tushnet, Running the Gamut from A to B: Federal Trademark and False Advertising Law, 159 U. Pa. L. Rev. 1305, 1318 (2011) (explaining the ways in which trademark use relies upon “consumers knowledge about how source and sponsorship get communicated” to indicate source by implication). “[V]anishingly few products say, ‘This is a bottle of Coca-Cola,’ or something similar. The presence of the name itself, in a distinctive font in the most prominent position on the label, necessarily implies the source claim to a culturally competent modern consumer.” Tushnet, supra, at 1318.

  126. [127]. See, e.g., In re CNS, Inc., Serial Nos. 76250116; 76250194; 76250611; 76250613, 2005 WL 3175107, at *4 (T.T.A.B. Nov. 18, 2005) (combining multiple applications and finding that marks lacked trademark significance where they appeared only on the back and spine of a box and on the bottom front near descriptive text).

  127. [128]. Lee et al., supra note 19, at 1076; see also Robert G. Bone, Enforcement Costs & Trademark Puzzles, 90 Va. L. Rev. 2099, 2130 (2004) (noting consumers may be primed to perceive matter used on product labels as source identifiers).

  128. [129]. In re Michelin N. Am., Inc., Serial No. 85077031, 2015 WL 5675633, at *4 (T.T.A.B. Sept. 11, 2015) (“The vanity number is displayed prominently at the top of the specimens
    ... separated from the advertising copy on the materials. It appears in a font size that is at least as large as the word MICHELIN (clearly itself being used as a mark). It does not blend so well with the other matter on the specimens that it is difficult to discern what the mark is, but rather is set apart from the other marks as well as the text on the pages... . by means of a different color scheme. There is a ™ symbol following the telephone number, which although not dispositive as to whether the telephone number is actually perceived as a mark, lends a degree of visual prominence to the term.” (citation omitted)).

  129. [130]. TMEP, supra note 2, § 1202.03; see also “Ornamental” Refusal and How to Overcome This Refusal, USPTO (Apr. 14, 2016, 1:20 PM), (providing guidance to applicants who receive office actions refusing to register matter as a trademark on the basis that it is ornamental).

  130. [131]. West-Com Nurse Call Sys., Inc., Serial No. 78438594, 2007 WL 2972207, at *6–7 (T.T.A.B. Sept. 13, 2007).

  131. [132]. See Dinwoodie & Janis, supra note 29, at 317.

  132. [133]. See Lionel Bently & Robert Burrell, The Requirement of Trade Mark Use, 13 Austl. Intell. Prop. J. 181, 185 (2002).

  133. [134]. Anyone can use the trademark symbol ™ to convey their own belief that the matter to which it is appended is being used as a trademark. The ® symbol, on the other hand, must only be used by owners of federally registered marks.

  134. [135]. See, e.g., Bobosky v. Adidas AG, 843 F. Supp. 2d 1134, 1145 (D. Or. 2011); In re Mentor Graphics Corp., Serial No. 78325604, 2008 WL 906611, at *3 (T.T.A.B. Mar. 6, 2008). But see Go Pro Ltd. v. River Graphics, Inc., No. CIVA01CV600JLK, 2006 WL 898147, at *6 (D. Colo. Apr. 5, 2006) (“[T]he use of ... [an] intellectual property designation does not, by itself, create a trademark.”); In re A La Vieille Russie, Inc., 60 U.S.P.Q.2d (BNA) 1895, 2001 WL 862510, at *6­–7 (T.T.A.B. 2001)(holding that the use of the term “Russianart” buried in advertisement text does not establish service mark significance merely because the “SM” symbol accompanies it).

  135. [136]. See, e.g., Dessert Beauty, Inc. v. Fox, 568 F. Supp. 2d 416, 424 (S.D.N.Y. 2008), aff’d, 329 F. App’x 333 (2d Cir. 2009) (“[Plaintiff] DBI did not use ‘love potion’ as a trademark because the source of its fragrance products was not identified by that term... . Moreover, [Plaintiff] placed a TM symbol only next to the word ‘Dessert,’ highlighting the non-trademark use of ‘love potion.’ The TM symbol was not placed next to the words ‘love potion.’”); In re Vertex Grp., LLC, 89 U.S.P.Q.2d (BNA) 1694, 2009 WL 398091, at *8 (T.T.A.B. 2009) (“As is the case with any trademark, mere intent that a word, name, symbol or device function as a trademark or service mark is not enough in and of itself.”).

  136. [137]. See, e.g., In re Standard Oil Co., 275 F.2d 945, 947 (C.C.P.A. 1960) (“Whatever may have been the intention of the applicant in using [the words], their use has not accomplished what the applicant wished to do.”); In re McIlhenny Co., 278 F.2d 953, 956 (C.C.P.A. 1960) (“The mere intent of an applicant will not convert ... [a matter] into a trademark.”); Thoroughbred Legends, LLC v. Walt Disney Co., No. 1:07-CV-1275-BBM, 2008 WL 616253, at *6 (N.D. Ga. Feb. 12, 2008) (“[O]ne cannot acquire rights in a trademark by asserting he owns it.”); Go Pro Ltd., 2006 WL 898147, at *6 (“Not all words, devices, symbols, and the like necessarily function as trademarks notwithstanding that they may have been adopted with the intent of doing so.”); In re Brass-Craft Mfg. Co., 49 U.S.P.Q.2d (BNA) 1849, 1998 WL 1120830, at *3 (T.T.A.B. 1998) (“[A]pplicant’s use of the notice indicating that the asserted marks are trademarks did not transform these designations into trademarks ... .”); see also James E. Hawes & Amanda V. Dwight, Trademark Registration Practice § 8:2 (“In determining whether a feature is ornamentation or a mark, it is not the applicant’s intent or purpose which controls, but rather the public’s perception of the feature.”).

  137. [138]. In re Aerospace Optics, Inc., 78 U.S.P.Q.2d (BNA) 1861, 2006 WL 1087849, at *3 (T.T.A.B. 2006).

  138. [139]. Id. at *2.

  139. [140]. Id.

  140. [141]. In re Singer Mfg. Co., 255 F.2d 939, 941 (C.C.P.A. 1958) (“No authority is cited, and none has been found, to the effect that a trademark use requires a display of a design of any particular size or degree of prominence.”).

  141. [142]. Registration No. 4,786,025, USPTO Office Action, June 18, 2015, available at; see also, e.g., MicroStrategy Inc. v. Motorola, Inc., 245 F.3d 335, 343 (4th Cir. 2001) (finding no use as a mark for infringement plaintiff, where phrase was used inconsistently, in small font, and without prominent placement).

  142. [143]. In re Eli Lilly & Co., Serial No. 85183667, 2015 WL 4241138, at *1 (T.T.A.B. June 18, 2015) (alteration in original) (citation omitted) (quoting the Examining Attorney).

  143. [144]. Id. at *3. The specimen did reflect capitalization and use of the TM symbol. Id. at *2.

  144. [145]. See supra Section II.B.

  145. [146]. See supra note 116 and accompanying text.

  146. [147]. See Roberts, supra note 102, at 601; infra Appendix A.

  147. [148]. See Roberts, supra note 102, at 632–41.

  148. [149]. TMEP, supra note 2, § 1202.18.

  149. [150]. See In re DePorter, Serial No. 87229711, at 4 (Jan. 29, 2019), available at http:// #MagicNumber108 failed to function as a mark for shirts on the grounds that “the public will not perceive the term as a trademark that identifies the source of Applicant’s goods but rather only as conveying an informational message”).

  150. [151]. #SHARETHESILENCE, Registration No. 4,749,708.

  151. [152]. #BEUNPRECEDENTED, Registration No. 4,671,869.

  152. [153]. In response to a question asking whether #BeUnprecedented was a hashtag or a trademark based on a portion of the specimen LexisNexis submitted to the USPTO, “only 5% of respondents chose ‘trademark’ or ‘both.’” Alexandra Roberts, Hashtags Are Not Trademarks
    Eksouzian v. Albanese (Guest Blog Post), Tech. & Marketing L. Blog (Aug. 26, 2015),

  153. [154]. Roberts, supra note 102, at 632–41 (citing specimens for registered trademarks including #McCallWeddings, #PowerOfHer, #LetsBowl, #TrendinGHot, #ReachForPeach, #LiveTheGive, #TutLife, #FixItJesus, and #TeamGifted).

  154. [155]. CHIPSET-FREE CHARGING, Registration No. 5,247,769.

  155. [156]. BIG IS THE NEW SMALL, Registration No. 5,162,928.


  157. [158]. HONEY BADGER DON’T CARE, Registration No. 4,505,781 (specimen in text); see also Registration Nos. 4,419,081; 4,281,472; 4,419,079; 5,059,721 (using “Honey Badger Don’t Care” with commercial goods other than clothing).

  158. [159]. Gordon v. Drape Creative, Inc., 909 F.3d 257, 263 (9th Cir. 2018).

  159. [160]. The 26,450 applications that garnered a failure to function refusal between 2003 and 2015 represent only 0.5% to 1% of the total application set for each year. See infra Appendix B.

  160. [161]. Cf. Masur, supra note 35, at 470, 472–85 (asserting that the USPTO “will grant nearly any plausible patent” to avoid appeals and reversals, which bear both reputational and monetary costs). Granting registration for a mark without proper use will almost never lead to challenge or blowback. Compare this situation to that of marks that create a likelihood of confusion or dilution with previously registered marks—in such cases the prior registrants are likely to challenge the new registration, increasing administrative costs.

  161. [162]. On the other hand, one need spend only a few minutes searching for common words on TESS, the USPTO website, in order to turn up examples of marks deemed registrable based on specimens reflecting few to no indicia of trademark use.

  162. [163]. See In re Mogen David Wine Corp., 372 F.2d 539, 543 (C.C.P.A. 1967) (“The duration of a trademark depends on public recognition that the trademark identifies the user’s goods and distinguishes them from the goods of others.Trademark law is replete with instances where the public has both ‘created’ and ‘destroyed’ the trademark rights ... independently of any action or inaction by the trademark owner.”); Jane C. Ginsburg,Euro-Yearnings? Moving Toward A “Substantive” Registration-Based Trademark Regime, 130 Harv. L. Rev. F. 95, 96 (2017) (“[A] trademark fills its role of advancing the public’s interest in avoiding confusion or deceit only if the public perceives that the trademark symbolizes a particular producer’s goods.Without the consumer-perception link, the symbol does not ‘identify and distinguish’one producer’s goods, and it therefore is not a trademark.” (emphasis added) (footnote omitted)); Laura A. Heymann, TheReasonable Person in Trademark Law, 52 St. Louis U. L.J. 781, 781 (2008) (“The meaning of any trademark is ultimately determined by the consumer.”).

  163. [164]. See Coca-Cola Co. v. Busch, 44 F. Supp. 405, 411 (E.D. Pa. 1942) (enjoining use of KOKE-UP for soft drinks).

  164. [165]. See Coca-Cola Co. v. Gemini Rising, Inc., 346 F. Supp. 1183, 1186 (E.D.N.Y. 1972).

  165. [166]. See generally Hana Fin., Inc. v.Hana Bank,135 S. Ct. 907 (2015) (holding that the issue was a question of fact for the jury).

  166. [167]. See, e.g., Jake Linford, Are Trademarks Ever Fanciful?, 105 Geo L. J. 731, 734–35, 750 (2017).

  167. [168]. One exception is the inclusion in many circuits’ likelihood of confusion factors of the degree of consumer sophistication, specialization, or price-related attention that relevant consumers are expected to exercise in making purchasing decisions for the goods or services in question. See Laura A. Heymann,The Grammar of Trademarks, 14 Lewis & Clark L. Rev. 1313, 1316 (2010); Thomas R. Lee et al.,Trademarks, Consumer Psychology, and the Sophisticated Consumer, 57 Emory L.J. 575, 578–82 (2008).

  168. [169]. See, e.g., Multi Time Mach., Inc. v., Inc., 804 F.3d 930, 938 (9th Cir. 2015), cert. denied, 136 S. Ct. 1231 (2016); Toyota Motor Sales, U.S.A., Inc. v. Tabari, 610 F.3d 1171, 1178 (9th Cir. 2010); Nitro Leisure Prods., L.L.C. v. Acushnet Co., 341 F.3d 1356, 1362–63 (Fed. Cir. 2003). Ann Bartow notes judges’ inconsistent assumptions about the sophistication and purchasing behaviors of wine drinkers in trademark infringement litigation and the ways in which those conflicting assumptions have been crystallized in different strands of case law. Ann Bartow, Likelihood of Confusion, 41 San Diego L. Rev. 721, 773 (2004).

  169. [170]. Krieger & Fiske, supra note 37, at 998.

  170. [171]. Id. at 998–99.

  171. [172]. Rik Pieters & Michael Wedel, Attention Capture and Transfer in Advertising: Brand, Pictorial, and Text-Size Effects, 68 J. Marketing 36, 49 (2004); see also Ulrich R. Orth & Keven Malkewitz, Holistic Package Design and Consumer Brand Impressions, 72 J. Marketing 64, 64 (2008) (“A significant body of research attests to the importance that managers and scholars attach to package design... . [P]ackage design is an extremely influential medium because of its pervasive impact on purchasers, its presence at the crucial moment when the purchase decision is made, and consumers’ high level of involvement when they actively scan packages in their decision making.”).

  172. [173]. 15 U.S.C. § 1127 (2012). The definition also includes matter “which a person has a bona fide intention to use in commerce and applies to register on the principal register established by this chapter.” Id. Actual registration, however, can issue only upon a showing of use in commerce.

  173. [174]. Krieger & Fiske, supra note 37, at 999.

  174. [175]. Put another way, it makes assumptions about what the matter does based on what the matter is, rather than assessing what the matter does more directly.

  175. [176]. See generally Deven Desai, Should Trademark Law Protect Non-Traditional Trademarks?, in The Protection of Non-Traditional Trademarks: Critical Perspectives (Irene Calboli & Martin Senftleben eds., 2018) (arguing that trademark law evolves to protect changing business practices). Recent examples include Hasbro’s attempt to register the scent of Play-Doh as a trademark for Play-Doh, General Mills’ attempt to register the yellow Cheerios box as trade dress for cereal, and Gene Simmon’s application to register a hand symbol as a trademark for entertainment services.

  176. [177]. For examples of the myriad athletes who have registered their names or nicknames as trademarks, see Alexandra J. Roberts, Athlete Trademarks: Names, Nicknames, and Catchphrases, in The Oxford Handbook of American Sports Law 471 (Michael A. McCann ed., 2018).

  177. [178]. See Mark P. McKenna & Lucas S. Osborn, Trademarks and Digital Goods, 92 Notre Dame L. Rev. 1425, 1426 (2017).

  178. [179]. For example, see Eggland’s Best eggs imprinted with “EB” logo. Eggland’s Best Announces Their Eggs Now Have 6X More Vitamin D Than Ordinary Eggs, Eggland’s Best (Apr. 25, 2017), Per the company, “[t]he logo is applied by stamping machines which use an FDA-compliant, USDA-approved, food-safe ink that is acceptable as a food additive.” Frequently Asked Questions, Eggland’s Best, (last visited Mar. 20, 2019).

  179. [180]. Kerry Martin, A Cross-Promotion That Means [Monkey] Business, Curley & Pynn (Jan. 11, 2011),

  180. [181]. See, e.g., Football MIDFIELD LOGO Stencil, USSC: US Specialty Coatings, (last visited Mar. 20, 2019) (displaying images of football logo imprints on football fields).

  181. [182]. See Peter J. Karol, Affixing the Service Mark: Reconsidering the Rise of An Oxymoron, 31 Cardozo Arts & Ent. L.J. 357, 391–94 (2013).

  182. [183]. As services in particular evolve and the concept of indicating source becomes more muddled in that context, it becomes increasingly difficult to identify what constitutes a service mark. See McKenna & Osborn, supra note 178, at 1463–69.

  183. [184]. See James Gibson,Risk Aversion and Rights Accretion in Intellectual Property Law, 116 Yale L.J. 882, 908 (2007) (“[W]hat consumers view as the norm becomes the norm because consumer perception is trademark law’s touchstone.” (emphasis omitted)).

  184. [185]. Lee et al., supra note 19, at 1039–54.

  185. [186]. Of course, consumer perception is not the only reason to decline granting protection to descriptive terms without proof of acquired distinctiveness. Considerations like promoting competition and protecting freedom of expression also inform this policy. See Lisa P. Ramsey, Descriptive Trademarks and the First Amendment, 70 Tenn. L. Rev. 1095, 1113–18 (2003) (discussing how trademark protection as applied to descriptive terms has evolved).

  186. [187]. Lee et al., supra note 19, at 1088 (“Participants were presented with a product package that displayed the mark, and they were asked, ‘In the product package shown above ... would you say [the mark]’ ... ‘is a brand name,’ ‘is not a brand name,’ or ‘I don’t know or have no opinion.’”).

  187. [188]. In addition to the “trademark spot” rationale discussed above, the fact that the faux cookie boxes contained no other obvious trademarks probably increased the likelihood that consumers would deem “wonderful” a trademark. Some scholars have advocated restricting trademark law to allow any given producer protection for only one source; such an approach would likely reduce the incidence of producers asserting protection for marks that fail to function, as those marks are often secondary or tertiary to a main mark. David W. Barnes, One Trademark Per Source, 18 Tex. Intell. Prop. L.J. 1, 3­­–7 (2009). Conversely, a heightened emphasis on use as a mark at the USPTO and by courts might lead to less claiming—if one or several marks are large, prominently placed, in distinct font, and otherwise cultivated to attract consumer attention, it logically follows that the number of different protectable marks a product can showcase is finite, and that consumers are unlikely to view more than a few different things on a single sign or package as separate source indicators.

  188. [189].

  189. [190]. Xiaoyan Deng & Barbara E. Kahn, Is Your Product on the Right Side? The “Location Effect” on Perceived Product Heaviness and Package Evaluation, 46 J. Marketing Res. 725, 725 (2009).

  190. [191]. J.S. Burt et al., Brandname Confusion: Subjective and Objective Measures of Orthographic Similarity, 146 J. Experimental Psychol. 1, 34 (2017).

  191. [192]. Pieters & Wedel, supra note 172, at 37.

  192. [193]. Id. at 44, 48.

  193. [194]. Id. at 48; see also Luca Cian et al., This Logo Moves Me: Dynamic Imagery from Static Images, 51 J. Marketing Res. 184, 186 (2014) (“Visual cues specifically affect attention and consumer engagement.”).

  194. [195]. Aparna Sundar & Theodore J. Noseworthy, Place the Logo High or Low? Using Conceptual Metaphors of Power in Packaging Design, 78 J. Marketing 138, 140 (2014) (citing Nathan Novemsky et al., Preference Fluency in Choice, 44 J. Marketing Res. 347–56 (2007)).

  195. [196]. Sundar & Noseworthy, supra note 195, at 138 (citation omitted).

  196. [197]. Orth & Malkewitz, supra note 172, at 64.

  197. [198]. McCarthy, supra note 9, § 3:3.

  198. [199]. See also Barton Beebe,Search and Persuasion in Trademark Law, 103 Mich. L. Rev. 2020, 2029 (2005)(“As perceived by a given consumer, a trademark either is or is not source distinctive; it either qualifies for protection or it does not.”).

  199. [200]. Consumer surveys conducted in the course of litigation also offer insight based on larger sample groups and are specifically tailored to the matter in question. While such surveys are more likely to reflect bias precisely because they are prepared for litigation, they can still prove helpful, especially when experts from both sides are given an opportunity to highlight the strengths and weaknesses of survey design. Surveys are regularly used in infringement and dilution litigation but are far less commonly used to establish protectability.

  200. [201]. See, e.g., Nextel Commc’ns, Inc. v. Motorola, Inc., Opposition No. 91161817, 2008 WL 902843, at *9 (T.T.A.B. Feb. 27, 2008) (“Having determined above that applicant’s 911 Hz chirp fails to function as a trademark, we need not and do not reach the issue of whether the proposed mark is inherently distinctive or has acquired distinctiveness.”).

  201. [202]. This combined approach to distinctiveness may also have been common before the rise of Abercrombie and the creation of the distinctiveness taxonomy. See, e.g., Roux Labs., Inc. v. Clairol Inc., 427 F.2d 823, 826–28 (C.C.P.A. 1970).

  202. [203]. U.S. Trademark Application Serial No. 78/438,594 (filed June 21, 2004).

  203. [204]. In re West-Com Nurse Call Sys., Inc., Serial No. 78438594, 2007 WL 2972207, at *5 (T.T.A.B. Sept. 13, 2007).

  204. [205]. Nor did it address functionality, another potential hurdle for the applicant. That may be because the TTAB is limited on appeal to the issues raised by the examining attorney; the examining attorney in West-Com had initially based his refusal on both failure to function and mere descriptiveness, then subsequently withdrew his finding of mere descriptiveness. See Examining Attorney’s Appeal Brief at *17–20, In re West-Com Nurse Call Sys., Inc., Serial No. 78438594, 2007 WL 2972207 (T.T.A.B. Sept. 13, 2007) (No. 78/438594), 2006 WL 6863586.

  205. [206]. See, e.g., In re Michelin N. Am., Inc., Serial No. 85077031, 2015 WL 5675633, at *2 (T.T.A.B. Sept. 11, 2015).

  206. [207]. One critic complains of trademark bullies who routinely “present specimens with textbook-perfect hang tags ... in applications to register ... popular expressions” they intend to use ornamentally. Once the mark is registered, the registrant files infringement claims “broadly ... against all competitors.” E-mail from Morgan Reece, Author/Advocate, to Alexandra J. Roberts, Assoc. Professor, Univ. of N.H. Sch. of Law (Aug. 10, 2018) (on file with author).

  207. [208]. See, e.g., Hawes & Dwight, supra note 137, § 8:2 (“By appropriately selecting the specimens showing the applicant’s use as a mark ... problems can be cured or avoided.”).

  208. [209]. Self-Realization Fellowship Church v. Ananda Church of Self-Realization, 59 F.3d 902, 907 (9th Cir. 1995).

  209. [210]. In re Mentor Graphics Corp., Serial No. 78325604, 2008 WL 906611, at *3 (T.T.A.B. Mar. 6, 2008).

  210. [211]. Id. at *1.

  211. [212]. Id. at *2 (quoting the arguments of the Examining Attorney).

  212. [213]. Id. at *3 (citing to the Examining Attorney’s appeal brief).

  213. [214]. Id.

  214. [215]. Id.

  215. [216]. Id.

  216. [217]. See, e.g., In re Roberts, 87 U.S.P.Q.2d (BNA) 1474, 2008 WL 1944634, at *4–5 (T.T.A.B. 2008) (refusing applications to register domain name marks based on failure to function); In re Indus. Risk Insurers, Serial No. 76097495, 2008 WL 885952, at *2 (T.T.A.B. Feb. 14, 2008) (same);

  217. [218]. In re Eilberg, 49 U.S.P.Q.2d at *2.

  218. [219]. Id. at *1.

  219. [220]. Id.

  220. [221]. The Federal Circuit has repeatedly held that the addition of a TLD like .com or .net will not render a descriptive mark inherently distinctive or a generic mark protectable. See, e.g., In re Oppedahl & Larson LLP, 373 F.3d 1171, 1172 (Fed. Cir. 2004). However, a court in the Fourth Circuit recently held the opposite in assessing the distinctiveness of BOOKING.COM for hotel reservation and travel agent services. See B.V. v. Matal, 278 F. Supp. 3d 891,
    923–24 (E.D. Va. 2017).

  221. [222]. In re Indus. Risk Insurers, 2008 WL 885952, at *1–2.

  222. [223]. In re Oppedahl & Larson LLP, 373 F.3d at 1172.

  223. [224]. Id.

  224. [225]. Id. at 1173.

  225. [226]. See id. at 1176.

  226. [227]. The Board and Court had their hands tied, as they were limited to the issues raised by the examining attorney.

  227. [228]. See ESPN, Inc. v. Quiksilver, Inc., 586 F. Supp. 2d. 219, 226 (S.D.N.Y. 2008).

  228. [229]. Answer and Counterclaims at 13–17, ESPN, Inc. v. Quiksilver, Inc., 586 F. Supp. 2d 219 (S.D.N.Y. 2008) (No. 08 Civ. 4222).

  229. [230]. Id. Examples include:

  230. [231]. Id. at 21–22.

  231. [232]. Plaintiff-Counterdefendant ESPN Inc.’s Motion to Dismiss Defendant-Counterclaimant Quiksilver, Inc.’s Second, Third, Fourth, and Fifth Counterclaims at 11, ESPN, Inc. v. Quiksilver, Inc., 586 F. Supp. 2d 219, (S.D.N.Y. 2008) (No. 08 Civ. 4222).

  232. [233]. ESPN, Inc., 586 F. Supp. 2d at 226.

  233. [234]. See Rock & Roll Hall of Fame & Museum, Inc. v. Gentile Prods., 134 F.3d 749, 753, 756 (6th Cir. 1998).

  234. [235]. As the scholarship on defendants’ use has discussed, courts seem to reason backward from the likelihood of confusion: if confusion is plausible, then the type and extent of both plaintiff’s and defendant’s use must be sufficient to support the cause of action.

  235. [236]. Viacom Int’l Inc. v. IJR Capital Invs., LLC, 242 F. Supp. 3d 563, 568–69 (S.D. Tex. 2017); DC Comics v. Kryptonite Corp., 336 F. Supp. 2d 324, 332 (S.D.N.Y. 2004);DC Comics, Inc. v. Powers, 465 F. Supp. 843, 845, 847 (S.D.N.Y. 1978). See generally Alexandra Roberts, Viacom Possesses Trademark Rights in ‘Krusty Krab’ Based on Its Central Role in the SpongeBob Universe–Viacom v. IJR (Guest Blog Post), Tech. & Marketing L. Blog (June 9, 2018), https://blog.ericgoldman.

  236. [237]. Defendant IJR Capital Investments, LLC’s Opposition to Plaintiff’s Motion for Summary Judgment and Memorandum Supporting Defendant’s Motion for Summary Judgment Out of Time at *16 n.8, Viacom Int’l Inc. v. IJR Capital Invs., LLC, 242 F. Supp. 3d 563 (S.D. Tex. 2017) (No. 4:16-cv-00257), 2016 WL 8647422.

  237. [238]. Tenneco Auto. Operating Co. v. Kingdom Auto Parts, 410 F. App’x 841, 843–44 (6th Cir. 2010).

  238. [239]. Tenneco Auto. Operating Co. v. Kingdom Auto Parts, No. 08-CV-10467, 2009 WL 1438834, at *8 (E.D. Mich. May 18, 2009), aff’d, 410 F. App’x 841 (6th Cir. 2010).

  239. [240]. 171994, Registration No. 4,014,055.

  240. [241]. See Tenneco Auto. Operating Co., 410 F. App’x at 844 (referencing office actions taken on June 16, 2008, July 31, 2008, April 6, 2009, and April 13, 2011). The Sixth Circuit notes that despite those refusals, at one point the USPTO withdrew its failure to function refusal and granted publication of the mark, only to subsequently withdraw its approval and reinstate its original failure to function refusals. Id.

  241. [242]. TMEP, supra note 2, § 1202.16(a).

  242. [243]. In re Chevron Intellectual Prop. Grp. LLC, Serial No. 78490836, 2010 WL 5010883, at *1 (T.T.A.B. Nov. 23, 2010).

  243. [244]. Seabrook Foods, Inc. v. Bar-Well Foods Ltd., 568 F.2d 1342, 1344 (C.C.P.A. 1977).

  244. [245]. Nola Spice Designs, L.L.C. v. Haydel Enters., Inc., 783 F.3d 527, 540–41 (5th Cir. 2015) (quoting Amazing Spaces, Inc. v. Metro Mini Storage, 608 F.3d 225, 243 (5th Cir. 2010)).

  245. [246]. In re Chevron Intellectual Prop. Grp. LLC, 96 U.S.P.Q.2d at *3.

  246. [247]. Id. at *1.

  247. [248]. Id. at *5; All. for Good Gov’t v. Coal. for Better Gov’t, 901 F.3d 498, 508 (5th Cir. 2018); Tex. Outhouse Inc. v. Fresh Can, LLC, 266 F. Supp. 3d 928, 935 (S.D. Tex. 2017); In re Arlington Specialties, Inc., Serial Nos. 85851794; 85851833, 2017 WL 914069, at *9 (T.T.A.B. Feb. 21, 2017); In re Icelandic Provisions, Inc., Serial No. 86768287, 2017 WL 914083, at *6 (T.T.A.B. Feb. 16, 2017); In re Fantasia Distrib., Inc., Serial No. 86185623, 2016 WL 5866950, at *2 (T.T.A.B. Sept. 21, 2016); Bd. of Trs. of the Univ. of Ala. v. Pitts, 107 U.S.P.Q.2d (BNA) 2001, 2013 WL 4397047, at *12 (T.T.A.B. 2013); In re Banana Republic (Apparel), LLC, Serial No. 78485048, 2007 WL 1580023, at *4 (T.T.A.B. May 23, 2007).

  248. [249]. See Seabrook Foods, Inc., 568 F.2d at 1344. The fourth Seabrook factor has been read literally and thus construed narrowly by some courts to ask whether matter creates a commercial impression distinct from the surrounding text. See, e.g., In re Ruby Mine, Inc., Serial No. 85287700, 2013 WL 3129909, at *3 n.2 (T.T.A.B. Apr. 16, 2013) (“We do not find [the fourth Seabrook factor] relevant, since the drawing submitted by applicant contains no words.”). That interpretation echoes Seabrook itself, which considered whether the design portion of the mark functioned independently of the word portion to identify and distinguish Seabrook’s goods. Seabrook Foods, Inc., 568 F.2d at 1344.

  249. [250]. See, e.g., Rock & Roll Hall of Fame & Museum, Inc. v. Gentile Prods., 134 F.3d 749, 753 (6th Cir. 1998) (assessing the protectability of the plaintiff’s word mark and building design together in stating “it is not the case that all inherently distinctive symbols or words on a product function as trademarks. Rather, in order to be protected as a valid trademark, a designation must create a separate and distinct commercial impression, which ... performs the trademark function of identifying the source of the merchandise to the customers”); see also, e.g., Procter & Gamble Co. v. Keystone Auto. Warehouse, Inc., 191 U.S.P.Q. (BNA) 468, 1976 WL 20910, at *7 (T.T.A.B. 1976) (“[I]t is reasonable to assume that the phrase ‘BUMPER TO BUMPER CAR CARE’ and design has created a commercial impression separate and apart from ‘PROCTER & GAMBLE’ and that it serves, in and of itself, as an identification symbol for opposer’s goods ... .”).

  250. [251]. McCarthy lists three factors, acknowledging the fourth only in a footnote. McCarthy, supra note 9, § 8:13. Some courts identify four factors but set the fourth aside as addressing something other than distinctiveness. See, e.g., Forney Indus., Inc. v. Daco of Mo., Inc., 835 F.3d 1238, 1246 n.3 (10th Cir. 2016) (“We do not consider this factor because it does not relate to whether the trade dress is inherently distinctive, but instead ‘to a different question: is the design really a separately registerable mark apart from any nearby words.’” (quoting McCarthy, supra note 9, § 8:13 n.9)); Amazing Spaces, Inc. v. Metro Mi­ni Storage, 608 F.3d 225, 243 n.14 (5th Cir. 2010); In re Chippendales USA, Inc., 622 F.3d 1346, 1351–52 (Fed. Cir. 2010) (“The fourth factor, whether the trade dress was capable of creating a commercial impression distinct from the accompanying words, is not applicable here.”). The Second Circuit has seemingly gone so far as to hold that if matter is inherently distinctive under the first three Seabrook factors, the fourth factor becomes moot. See Star Indus., Inc. v. Bacardi & Co., 412 F.3d 373, 383 (2d Cir. 2005) (reversing the district court’s holding that the matter was not inherently distinctive and did not create a separate commercial impression by finding the mark inherently distinctive and noting “the Star ‘O’ design is protectable separately from the other design elements on the Georgi orange-flavored vodka label precisely because the ‘O’ design is itself inherently distinctive.” (emphasis added)).

  251. [252]. Duraco Prods., Inc. v. Joy Plastic Enters, Ltd., 40 F.3d 1431, 1434 (3d Cir. 1994).

  252. [253]. Knitwaves, Inc. v. Lollytogs Ltd., 71 F.3d 996, 1008 (2d Cir. 1995).

  253. [254]. Duraco, 40 F.3d at 1434.

  254. [255]. See, e.g., In re Chem. Dynamics, Inc., 839 F.2d 1569, 1571 (Fed. Cir. 1988) (affirming the TTAB’s refusal to register the dropper and droplet (featured below) as a trademark separate from the surrounding image); In re Michelin N. Am., Inc., Serial No. 85077031, 2015 WL 5675633, at *6 (T.T.A.B. Sept. 11, 2015); In re Wendy’s Int’l, Inc., 227 U.S.P.Q. (BNA) 884, 1985 WL 71946, at *2 (T.T.A.B. 1985) (holding that the shape of a sign did not make a commercial impression separate from the text and images that appear on it).

  255. [256]. The office actions in question use combination headings that referenced both doctrines, such as “Failure to Function as a Service Mark - The Applicant’s Mark Is Not Inherently Distinctive.” Registration No. 78/666,598, USPTO Office Action, Sept. 5, 2007, available at

  256. [257]. Motorola Mobility, Inc. v. Nextel Commc’ns, Inc., Opposition No. 91200355, 2012 WL 2588577, at *5 (T.T.A.B. June 8, 2012); Nextel Commc’ns, Inc. v. Motorola, Inc., Serial No. 78235365, 2009 WL 1741923, at *8 (T.T.A.B. June 12, 2009).

  257. [258]. In re N.V. Organon, 79 U.S.P.Q.2d (BNA) 1639, 2006 WL 1723556, at *15 (T.T.A.B. 2006) (“Even if we were to treat applicant’s application as being for its ‘unique’ orange flavor, we would find that applicant’s flavor fails to function as a mark. Because flavor is generally seen as a characteristic of the goods, rather than as a trademark, a flavor, just as in the cases of color and scent, can never be inherently distinctive.”); see also In re Pohl-Boskamp GmbH & Co. KG, 106 U.S.P.Q.2d (BNA) 1042, 2013 WL 1234849, at *1 (T.T.A.B. 2013) (applying to trademark peppermint flavor and scent).

  258. [259]. In re Fantasia Distrib., Inc., Serial No. 86185623, 2016 WL 5866950, at *1 (T.T.A.B. Sept. 21, 2016).

  259. [260]. Of course, both are examples of matter formerly protected only under unfair competition doctrine, not trademark doctrine. That evolution may help explain how the failure to function conundrum arose—as a casualty of the broad expansion of the concept of “trademark.”

  260. [261]. In re Chevron Intellectual Prop. Grp. LLC, Serial No. 78490836, 2010 WL 5010883, at *1 (T.T.A.B. Nov. 23, 2010).

  261. [262]. In re Fantasia Distrib., Inc., 2016 WL 5866950, at *1.

  262. [263]. Likewise, the Supreme Court in trade dress cases Wal-Mart v. Samara, Qualitex v. Jacobson Products, and Two Pesos v. Taco Cabana uses the language of inherent distinctiveness, but it might just as well use the language of failure to function. See generally Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205 (2000); Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159 (1995); Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992). A holding that product design or color cannot be inherently distinctive, but product packaging or analogous trade dress like restaurant décor can, is also a holding that product design and color will fail to function as trademarks without acquired distinctiveness, while product packaging-type matter is capable of being used in a trademark way and thus being perceived as a mark by consumers from its first use.

  263. [264]. Since overall protectability under the Lanham Act includes numerous additional considerations beyond inherent distinctiveness and use as a mark, it is helpful to distinguish “Abercrombie distinctiveness,” i.e., mark categorization as inherently distinctive or non-distinctive or as fanciful, arbitrary, suggestive, descriptive, or generic, from a more fulsome kind of distinctiveness that predicts whether consumers would perceive matter as a mark based both on what the mark is and how it is used.

  264. [265]. See infra Appendix A (providing images of all of the examples referenced in this chart along with explanations).

  265. [266]. Of course, heightened attention to use as a mark increases the importance of specimens, which carries implications for fraud, renewal, and incontestability. It also requires additional resources. Some scholars argue that given the trade-off between accuracy and efficiency, “additional correct results at the margin may not be worth the expense.” Grynberg,supra note 90, at 1289; see also Bone,supra note 128, at 2101 (arguing trademark law makes heavy use of presumptions and rules of thumb to simplify analyses and minimize enforcement costs, which may result in both over- and under-enforcing trademark rights).

  266. [267]. See, e.g., In re Vicki Roberts, Serial No. 76649075, 2008 WL 1944634, at *5 (T.T.A.B. May 2, 2008) (affirming the refusal to register IRESTMYCASE because it fails to function as a mark based on specimens directing consumers to visit a website at While In re Vicki Roberts focuses on the use displayed in the specimens, the Board acknowledges the relevance of the mark’s lack of distinctiveness to its use as a mark assessment: “[W]hile the inherent or acquired distinctiveness of the proposed mark is not before us, we observe nonetheless that [the mark describes or suggests the services with which it is used and] thus is even more dependent upon specimens displaying proper use of that designation as a mark in order for it to be perceived as such.” Id. at *4 n.2 (emphasis added). The TTAB uses the term “suggestive” but the context seems to imply it finds the mark unlikely to be inherently distinctive; the bracketed text indicates my reading of the opinion.

  267. [268]. In re Eli Lilly & Co., Serial No. 85183667, 2015 WL 4241138, at *3 (T.T.A.B. June 18, 2015).

  268. [269]. For example, the USPTO took into account the phrase’s font, prominence, and separateness from surrounding text in determining FOR TOTS would not serve as a source indicator for applicant’s services, but the TTAB summarized that analysis as follows: “The Examining Attorney’s argument that the term “FOR TOTS” fails to function as a mark is, in essence, a contention that such term is merely descriptive of applicant’s services, rather than one based on the manner in which the term is used on the additional specimen.” In re Language Workshop for Children/Cercle Franco Americain, Inc., Serial No. 75/516,045, 2003 WL 21996100, at *3 (T.T.A.B. Aug. 14, 2003).

  269. [270]. See Complaint at 12, Kimsaprincess, Inc. v. Missguided USA (Finance) Inc., No. 2:19-cv-01258 (C.D. Cal. Feb. 20, 2019) (alleging, inter alia, infringement of registered trademark KIM KARDASHIAN WEST).

  270. [271]. Jake Linford,The False Dichotomy Between Suggestive and Descriptive Trademarks, 76Ohio St. L.J.1367, 1370 (2015); Roberts, supra note 32, at 1056.

  271. [272]. McCarthy, supra note 9, § 3:4 (“For a designation which is not inherently distinctive and requires proof of secondary meaning, use as a trademark is essential.”).

  272. [273]. See, e.g., Park ‘N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 205 (1985).

  273. [274]. Restatement (Third) of Unfair Competition § 16 cmt. b (Am. Law Inst. 1993) (“The imitation or even complete duplication of another’s product or packaging will not create a risk of confusion unless some aspect of the duplicated appearance is identified with a particular source.”).


Associate Professor, University of New Hampshire School of Law; Fellow, Yale Law School Information Society Project.

For helpful discussions and suggestions on earlier drafts, the author thanks Suneal Bedi, Barton Beebe, Graeme Dinwoodie, Roger Ford, Susy Frankel, Jeanne Fromer, Peter Karol, Ashlyn Lembree, Yvette Liebesman, Jake Linford, Mark McKenna, Nicholson Price, Lisa Ramsey, Jessica Silbey, Ashlie Stewart Smith, Paul Tremblay, and Rebecca Tushnet, as well as participants in the 2018 Chicago IP Colloquium and Junior Scholars in IP at Michigan State and the 2017 IP Scholars’ Roundtable at UNH Law, Junior IP Scholars Association meeting, Works in Progress in IP conference (“WIPIP”), BU Journal of Science, Law, & Technology Symposium, Vanderbilt IP Scholars Roundtable, and IP Scholars Conference (“IPSC”). Thanks to Amanda Abelmann, Ryan Masters, and Ashlie Stewart Smith for able research assistance and the team at Iowa Law Review for stellar editing.