Opinion
Case No. 2:22-cv-00085-SK
03-04-2024
SNAP INC. v. Katherine K. VIDAL, et al.
David Hal Bernstein, Debevoise and Plimpton LLP, San Francisco, CA, Jared I. Kagan, Pro Hac Vice, Parker C. Eudy, Pro Hac Vice, Quenique Newbill, Pro Hac Vice, DeBevoise and Plimpton LLP, New York, NY, for Snap Inc. Jasmin Yang, AUSA — U.S. Attorneys Office, Los Angeles, CA, Lucy Grace D. Noyola, DOJ-Civ, Civil Division, Washington, DC, Carrie Rosato, DOJ-Civ, Civil Division — Office of Commercial Litigation, Washington, DC, Christina Hieber, Mai-Trang Dang, Mary Beth Walker, Pro Hac Vice, U.S. Patent and Trademark Office Office of the Solicitor, Alexandria, VA, for Katherine K. Vidal.
David Hal Bernstein, Debevoise and Plimpton LLP, San Francisco, CA, Jared I. Kagan, Pro Hac Vice, Parker C. Eudy, Pro Hac Vice, Quenique Newbill, Pro Hac Vice, DeBevoise and Plimpton LLP, New York, NY, for Snap Inc.
Jasmin Yang, AUSA — U.S. Attorneys Office, Los Angeles, CA, Lucy Grace D. Noyola, DOJ-Civ, Civil Division, Washington, DC, Carrie Rosato, DOJ-Civ, Civil Division — Office of Commercial Litigation, Washington, DC, Christina Hieber, Mai-Trang Dang, Mary Beth Walker, Pro Hac Vice, U.S. Patent and Trademark Office Office of the Solicitor, Alexandria, VA, for Katherine K. Vidal.
Proceedings: (IN CHAMBERS) ORDER DENYING DEFENDANTS' MOTION FOR SUMMARY JUDGMENT [ECF 68]
Steve Kim, United States Magistrate Judge.
I.
Plaintiff Snap, Inc. seeks registration of the mark SPECTACLES for use with its "smart glasses" product. (ECF 1 at 3, 14). To that end, Snap filed federal trademark applications in 2016, which the U.S. Patent and Trademark Office (USPTO) denied on the grounds that the proposed mark is generic or, if not generic, highly descriptive with insufficient acquired distinctiveness. (Id. at 7-10). The Trademark Trial and Appeal Board (TTAB) affirmed the registration denials, agreeing with the USPTO's genericness and descriptiveness findings. (Id. at 10; CAR at 2099-2130). Snap now seeks judicial review of the TTAB's adverse decision under 15 U.S.C. § 1071(b). (ECF 1 at 4, 13-14). Meanwhile, the USPTO and its Director Katherine Vidal move for summary judgment on genericness, arguing that there is no genuine dispute of material fact about whether relevant consumers of "smart glasses" perceive "spectacles" to be a generic name for the goods defined in Snap's applications— computer hardware and peripherals wearable as eyeglasses. (ECF 68). As permitted by § 1071(b), however, both sides have adduced new evidence on top of the administrative record to support their opposing views on the fact-bound consumer perception question. That combined evidence, viewed as a whole and construed in the light most favorable to Snap, reveals a genuine dispute of material fact over whether "spectacles" is a generic name for "smart glasses." Thus, the court must deny defendants' motion for summary judgment.
II.
Snap is a technology company that runs the social media messaging app Snapchat. (ECF 1 at 4-5). In 2016, Snap introduced a product called "SPECTACLES," which has camera technology attached to eyeglasses enabling wearers to capture and post videos and photos in the Snapchat
app. (Id. at 5). Soon after, Snap applied for registration of SPECTACLES as both a word mark and a stylized mark with the USPTO. (Id. at 7-9; CAR at 1-7, 2198-206). In its applications, Snap described the class of products for which it sought to register SPECTACLES as:
Computer hardware; computer peripherals; wearable computer hardware; wearable computer peripherals; computer hardware and peripherals for remotely accessing, capturing, transmitting and displaying pictures, video, audio and data; downloadable computer software, namely, software for setting up, configuring, and controlling wearable computer hardware and peripherals; downloadable computer software and software applications for use in uploading, downloading, capturing, editing, storing, distributing and sharing photographic and video content and other digital data via global and local computer networks and via mobile devices; downloadable multimedia files containing digital audio and video files featuring user generated images, videos, multimedia files, and other digital data, all in the fields of entertainment, photography and online social networking; computer software for accessing and transmitting data and content among consumer electronics devices and displays.
(CAR at 4, 2201). With its applications, Snap submitted specimen photographs depicting its product in the form of eyeglasses. (CAR at 12-13, 2205-06). As shorthand, then, the parties often maintained during agency registration proceedings below —and continue to maintain here—that "smart glasses" is an acceptable category descriptor for the class of products over which Snap seeks registration of the mark SPECTACLES. (CAR at 2110-11; see ECF 68 at 25; ECF 69 at 31).
In 2020, the USPTO denied Snap's applications, finding that the term "spectacles" is generic for the category of "smart glasses" and that, if not generic, it remains highly descriptive of Snap's product with insufficient secondary meaning capable of identifying the source of the product. (ECF 1 at 10; CAR at 2040-47, 4236-42). On administrative appeal, the TTAB affirmed the USPTO's refusal to register the proposed mark SPECTACLES. (ECF 1 at 10; CAR at 2130). Relying on dictionary definitions and media publications, the TTAB found that such evidence "clearly establishes that the term 'spectacles' is currently used and understood by the relevant public as a reference to hardware and peripherals wearable as glasses generally, i.e., 'smart spectacles' or 'AR spectacles.'" (CAR at 2118). It thus agreed with the USPTO that the term "spectacles" is generic for "smart glasses." (Id.). Also agreeing with the USPTO, the TTAB found alternatively that even if the term "spectacles" is not generic, it still just describes Snap's product with no acquired distinctiveness that is adequately source-identifying for relevant purchasers. (CAR at 2127-28).
Exercising its rights under 15 U.S.C. § 1071, Snap elected to challenge the TTAB's adverse agency decision through a civil complaint filed in federal district court as permitted by § 1071(b). And under that provision, both sides conducted discovery and adduced new evidence not presented originally to the USPTO. As a result, at trial, "the district court must make de novo factual findings that take account of both the new evidence and the administrative record before the [US]PTO." Kappos v. Hyatt, 566 U.S. 431, 446, 132 S.Ct. 1690, 182 L.Ed.2d 704 (2012); see B&B Hardware, Inc. v. Hargis Indus., Inc., 575 U.S. 138, 144, 135 S.Ct. 1293, 191 L.Ed.2d 222 (2015). In the meantime, though, the USPTO and its director seek summary judgment on the agency's genericness determination, arguing that Snap—even with its new evidence—cannot refute the USPTO's conclusion that relevant
consumers of "smart glasses" primarily understand the term "spectacles" to be generic for the entire product category of "smart glasses" rather than as an identifier for the source of any particular "smart glasses" product, including Snap's. (ECF 68 at 8).
III.
"Because of the intensely factual nature of trademark disputes, summary judgment is generally disfavored in the trademark arena." Interstellar Starship Servs., Ltd. v. Epix Inc., 184 F.3d 1107, 1109 (9th Cir. 1999). Even in non-trademark cases, the court may grant summary judgment only if "there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a). Any dispute about a material fact is considered "genuine" so long as there is an adequate evidentiary basis on which a rational trier of fact could find for the nonmoving party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). And when assessing the adequacy of the evidence, the court must view it in the light most favorable to the nonmoving party and draw all justifiable inferences in its favor. See id. at 255, 106 S.Ct. 2505. That means the court may not make credibility determinations, weigh the conflicting evidence, or draw inferences from the disputed facts in favor of the moving party. See id. at 249, 255, 106 S.Ct. 2505. In short, summary judgment is appropriate only when "the record taken as a whole could not lead a rational trier of fact to find for the nonmoving party." Ricci v. DeStefano, 557 U.S. 557, 586, 129 S.Ct. 2658, 174 L.Ed.2d 490 (2009).
In "registration proceedings, the USPTO 'always bears the burden' of establishing that a proposed mark is generic." Booking.com B.V. v. U.S. Pat. & Trademark Off., 915 F.3d 171, 179 (4th Cir. 2019) (quoting In re Cordua Rests., Inc., 823 F.3d 594, 600 (Fed. Cir. 2016)), vacated on other grounds, 591 U.S. 549, 140 S. Ct. 2298, 207 L.Ed.2d 738 (2020). Here, the parties agree (as they must) that whether the term "spectacles" is generic for "smart glasses" is ultimately a factual question. See Royal Crown Co. v. Coca-Cola Co., 892 F.3d 1358, 1364 (Fed. Cir. 2018) ("Whether an asserted mark is generic or descriptive is a question of fact."). "Whether any given [] term is generic," in turn, "depends on whether consumers in fact perceive that term as the name of a class or, instead, as a term capable of distinguishing among members of the class." Booking.com, 140 S. Ct. at 2307. "Evidence informing that inquiry can include not only consumer surveys, but also dictionaries, usage by consumers and competitors, and any other source of evidence bearing on how consumers perceive a term's meaning." Id. at 2307 n.6. As a result, when the parties have adduced competing evidence about consumer perception of an asserted mark, it is hard to resolve that intensely fact-bound question at the summary judgment stage. Compare, e.g., Solid 21, Inc. v. Richemont N. Am., Inc., 2023 WL 3996530, at *4 (S.D.N.Y. June 14, 2023) (declining to decide genericness on summary judgment due to conflicting evidence); with, e.g., BellSouth Corp. v. DataNational Corp., 60 F.3d 1565, 1571 (Fed. Cir. 1995) (finding summary judgment appropriate where all evidence was "uncontested"). This case is no exception. The parties' competing evidence of relevant consumer perception presented so far is not "so one-sided" in the USPTO's favor that defendants are entitled to summary judgment—as a matter of law—on the genericness of Snap's proposed mark. Retail Servs., Inc. v. Freebies Publ'g, 364 F.3d 535, 546 (4th Cir. 2004).
IV.
To be federally registered under the Lanham Act, a mark must "identify
and distinguish" goods from those manufactured or sold by others, such that it "indicates[s] the source of the goods." 15 U.S.C. § 1127. A generic term cannot be registered as a trademark (nor can it even be protected as such) because it is "the common descriptive name of an article or substance" and thus does not distinguish the product and source from others on the market. Park 'N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 193-94, 105 S.Ct. 658, 83 L.Ed.2d 582 (1985). The question of genericness turns on the "primary significance of the registered mark to the relevant public." 15 U.S.C. § 1064(3). The Ninth Circuit has described the primary significance inquiry, colloquially, as the "who are you/what are you" test: a registrable mark identifies the source of the product (species), while a generic term names the type of product (genus). See Filipino Yellow Pages, Inc. v. Asian J. Pubs., Inc., 198 F.3d 1143, 1147 (9th Cir. 1999). The primary significance test for genericness involves, essentially, a two-step inquiry. First, the court must identify the "genus of goods or services at issue." H. Marvin Ginn Corp. v. Int'l Ass'n of Fire Chiefs, Inc., 782 F.2d 987, 990 (Fed. Cir. 1986). Second, the court must determine whether "the term sought to be registered... [is] understood by the relevant public primarily to refer to that genus of goods or services." Id. "The critical issue in genericness cases" is therefore how "members of the relevant public primarily use or understand the term sought to be protected." Id. at 989-90.
A. The "Smart Glasses" Genus
On the first step, as noted, the parties profess to agree about the relevant genus of goods—"smart glasses." (ECF 68 at 25; ECF 69 at 31). Using that "working definition of the genus," the USPTO maintains that "'spectacles' must be found generic if it is generic for any portion of the genus." (ECF 68 at 25) (emphasis added). For sure, a term can be generic "if the relevant public understands the term to refer to part of the claimed genus of goods or services, even if the public does not understand the term to refer to the broad genus as a whole." In re Cordua Rests., 823 F.3d at 605. For instance, "the term 'pizzeria' would be generic for restaurant services, even though the public understands the term to refer to a particular sub-group or type of restaurant rather than to all restaurants." Id. Similarly, "if the public understands ZERO when used in combination with a designated beverage name to refer to a sub-group or type of beverage that carries specific characteristics" —like the "subcategory" of "drinks with few or no calories"—then "that would be enough to render the term generic" for the broader category of pertinent beverages. Royal Crown, 892 F.3d at 1368. So, in the beverages example, "ZERO need not be equated by the general public with the entire broad genus" of "soft drinks, sports drinks, and energy drinks" for the term to still be considered "generic" for that genus. Id. at 1367-68. Relying on that same category-subcategory scheme, the USPTO insists that "spectacles" must be generic for the claimed genus "smart glasses" because glasses is a "portion" or "part" of smart glasses. (ECF 68 at 25-26).
But the USPTO's "working definition of the genus," leaning as heavily as it does on the shorthand "smart glasses," hides (or at least obscures) an unstated—and unproven —factual premise. While not formally expressed as such, the USPTO's working conception of the relevant genus rests on the following syllogism:
Premise 1: "spectacles" is generic for "eyeglasses" Premise 2: "eyeglasses" is a "portion" or "part" of "smart glasses" Conclusion: "spectacles" is therefore generic for "smart glasses"
The second premise, however, is argument by ipse dixit: it presupposes that "glasses" is a necessary subset of—and thus completely subsumed by—the genus of "smart glasses." That is semantic sleight of hand, waving the shared word "glasses" to convey a fixed subgenus-genus relationship between that word and "smart glasses" as if it were axiomatic. It is one thing, of course, to claim that the genus of soft drink beverages (all chemically containing some caloric content) necessarily encompasses "zero-calorie" drinks, or that the genus of restaurants (serving as they must different food types) necessarily includes various restaurant types no matter if they serve pizza, churrasco, or sushi. That claim is logically sound and unassailable as a trademark-law principle. It is quite another thing, however, to claim that the relevant genus here—computer peripherals wearable as glasses—categorically encompasses all eyeglasses. That latter claim, instrumental to the agency's defense of its genericness finding, is beset by logical fallacies of division and composition when applied to the relevant genus here. Put another way, the analogies to beverages and restaurants could be potentially dispositive only if Snap's claimed genus were a broad class of eyewear—for then, even glasses with embedded computer technology could still be "clearly encompasse[d]" by the overarching eyewear genus both as a matter of category logic and trademark law. Royal Crown, 892 F.3d at 1368. But that is not necessarily this case. Nor is there is an inescapable reason why smart glasses couldn't be placed in its own category —overlapping with but distinct from eyeglasses. Indeed, wearable computer hardware and peripherals exist in various non-optical products (FitBit activity trackers or Apple Watches come to mind). So computer peripherals wearable as glasses could comprise a discrete category of goods created by the intersection of traditional optical glasses on the one side and non-optical computer wearables on the other.
What's more, the USPTO's categorization of "glasses" as a logical subset of "smart glasses" is as factually undeveloped as it is logically unsound. For a mark to be found generic because it names a "subcategory of the claimed genus," it must "refer[] to a key aspect of the genus." Royal Crown, 892 F.3d at 1368. A "key aspect" of a claimed genus is something that is "an integral, if not the paramount, aspect" of that genus. In re Cordua Rests., 823 F.3d at 603-04 (quoting In re Reed Elsevier Props. Inc., 482 F.3d 1376, 1379 (Fed. Cir. 2007)). While seemingly recognizing that legal requirement from the same cases it relies on, the USPTO just assumes that the "key aspect" of the "smart glasses" genus is invariably its "glasses" component but cannot be its computer technology—or so-called "smart"—component. (ECF 68 at 8, 29). Put another way, the USPTO presupposes that just as pizzerias are to restaurants and zero-calorie drinks are to beverages, so too are "glasses" to "smart glasses." But unlike restaurants that must all serve some food type on their menus to operate or beverages that must all contain some calories (on a scale from little to too much), not all eyeglasses contain computer hardware, and not all computer wearables are produced in the form of eyeglasses. What, then, is the key, integral, or paramount aspect of "smart glasses" as a class: is it the computer firmware; is it the wearable glasses; or is it some third altogether emergent
feature? The USPTO's answer is that it must be the glasses "part" or "portion," but nothing in trademark law—that it has cited or could be found anyway— preordains that assumption. So to say that "pizzeria" is generic for the class of restaurant services (given that food service is the paramount feature of restaurants) or that "zero" branded drinks is generic for a class of beverages (since caloric content is chemically integral to such liquids) reveals nothing materially dispositive about whether "glasses" should be considered a logical subcategory of the "smart glasses" genus as a matter of settled trademark law.
B. Consumer Perception of "Spectacles"
Exposing the underlying (if unstated) premises of the USPTO's conception of the relevant genus at the first step of the primary significance test is indispensable to the court's analysis at the critical second step. "By expanding or contracting the definition of a 'genus' of products, a court can substantially affect the final determination of whether a term is 'generic.'" In re Cordua Rests., 823 F.3d at 605 (quoting McCarthy on Trademarks § 12:23). It also helps explain why the parties appear to advance such divergent views of the evidence adduced for summary judgment. They are, in a real sense, talking past each other because they are gauging each other's evidence against different working conceptions of the relevant class of goods—even as they profess to have nominal agreement on how to define that class. For the USPTO, so long as the evidence shows that consumers primarily understand the term "spectacles" to be generic for conventional "eyeglasses"— which is essentially uncontested, if not incontestable, by Snap—it must be also generic for modern "smart glasses." For Snap, on the other hand, so long as the evidence shows that few consumers use the term "spectacles" to name the class of "smart glasses"—considering its unique computer technology elements—and that some tech-sawy (or self-interested) consumers even use the term as source-identifying of Snap's product, there must remain a genuinely disputed fact question about consumer perception of its asserted SPECTACLES mark.
The framing of the relevant public—general or specialized—can also be influential. That is why some courts include this as an intermediate third step in the genericness inquiry, See, e.g., Glover v. Ampak, Inc., 74 F.3d 57, 59 (4th Cir. 1996).
For summary judgment purposes only, Snap is right. The court must view the parties' evidence in the light most favorable to Snap and with all permissible inferences in Snap's favor. And that evidence here consists of (1) equivocal or neutral dictionary definitions; (2) Snap's own use of "glasses" merely to describe an intrinsic feature or function of its own product; (3) different assorted uses of the term "spectacles" by the public in traditional and social media; (4) selectively limited generic competitor use of the term from internet shopping searches; and (5) opposing consumer surveys unavailable at the agency-determination level. The court may not weigh this evidence or draw inferences in the USPTO's favor—especially the new competing expert reports and consumer surveys—to decide what is the most convincing at this stage. Applying these well-worn summary judgment standards, then, the court cannot find that whether relevant consumers perceive the term "spectacles" to be generic for "smart glasses" is so beyond genuine factual dispute that judgment must be entered in defendants' favor with no trial—and associated de novo factfinding—under § 1071(b).
First, the dictionary definitions are not decisive either way for summary judgment.
(CAR at 2049-51, 2230-32, 4244-47; ECF 68-4 at 22, 25). If anything, they only highlight how the parties are shooting at different targets. On the one hand, these definitions show—and Snap cannot credibly deny—that "spectacles" is generic for "eyeglasses" or "something resembling eyeglasses in shape or suggesting them in function." (ECF 68 at 10, 26; ECF 69 at 31). So if, as the USPTO argues, "eyeglasses" is just a subset—or a key aspect —of "smart glasses," then these definitions support the USPTO's view that "spectacles" is generic for "smart glasses." But whether the form and function of eyeglasses —or some computer hardware feature —is the integral or paramount aspect of "smart glasses" is a genuinely disputed fact question. On the other hand, contrary to Snap's suggestions (ECF 69 at 31), the absence of a dictionary defining "spectacles" as "smart glasses" is hardly decisive either. "A number of generic terms are not found in dictionaries." Door Sys., Inc. v. Pro-Line Door Sys., Inc., 83 F.3d 169, 171 (7th Cir. 1996). To be sure, the USPTO found a recent Cambridge Dictionary definition of "smart glasses." (ECF 68-4 at 25). But viewed in the light most favorable to Snap, that definition spotlights their "computer technology" and emphasizes how "they can be used in a similar way to a smartphone" as much as it recognizes that they are produced and sold as a "pair of glasses." (Id.). And the illustration of that word in a sentence—"A pair of smart glasses will mean that you can take photos with your spectacles" (id.)—"is not conclusive" as a matter of law "if for no other reason than that this would endow editors of such works with the power to destroy trademarks, merely by defining them genetically." In re Minnetonka, Inc., 212 U.S.P.Q. 772, at *5 (T.T.A.B. 1981).
Second, Snap's conceded use of "sunglasses" and "glasses"—by a company employee, in its public advertising, and in certain public filings (ECF 1 at 5; ECF 68-4 at 27, 38, 77, 203)—to describe or refer to its own "smart glasses" product moves the needle little, if at all, on summary judgment. Again, that evidence might support the USPTO's position but only to the extent that any form of eyewear—with or without embedded computer features—is categorically a subtype of the "smart glasses" genus. (ECF 68 at 26-27). And naturally, such usage might be decisive if all types of "eyeglasses" were the relevant genus. But as detailed above, the latter scenario is not necessarily implicated here, and the former depends on a disputed factual basis as much as it rests on a contestable logical premise.
Third, when viewed with all permissible inferences in Snap's favor, the existing and new evidence from traditional and social media sources combined establishes a genuine dispute of material fact about consumer perception of the term "spectacles" as applied to "smart glasses." (ECF 68 at 15-16, 27; ECF 69 at 14-16, 32-34). To be sure, many articles in the administrative record show that "spectacles" is used in certain marketing channels or media outlets to describe or refer to so-called "smart glasses," including uses like "AR spectacles" or "smart spectacles." (CAR 2381-418, 2868-70, 2873-76). But to counter that evidence, Snap has supplied new social media evidence from marketing and linguistic experts. Dr. Peter Golder, for instance, concludes that 86-100% of the digital media he analyzed—including social media posts, product reviews, and Google Trends search data—reveals that "spectacles" is used to identify Snap's product.
Snap also points to two new declarations from a content creator, Kyle Neely, and an AR designer, Lauren Cason, who both state that they understand the term "spectacles" to be source-identifying rather than generic. (ECF 69 at 19-21, 26-27; ECF 70-2; ECF 70-3).
(ECF 69-3 at 12-13, 17). He also finds that 92-99% of his sampled social media posts use spectacles as source-identifying of Snap's product. (Id. at 12-13). The other expert, Dr. Robert Leonard, examined the etymology of the term "spectacles" using dictionaries, historical corpora, and Google Ngram Viewer book data. (ECF 69-6 at 8). Based on that examination, he concludes that the relevant public does not commonly use the term "spectacles" to name the relevant product class of "smart glasses." (Id. at 7-8).
Contrary to the USPTO's assertions (ECF 68 at 29-30; ECF 72 at 8-10), Snap's newly adduced media evidence is probative of consumer perception. See Premier Nutrition, Inc. v. Organic Food Bar, Inc., 2008 WL 1913163, at *6 (C.D. Cal. Mar. 27, 2008) ("How the public uses the term is also instructive in whether a term is generic."); Booking.com, 915 F.3d at 182 ("absence of evidence" that consumers frequently used the term to describe the genus was "highly relevant"); Marvin Ginn, 782 F.2d at 989-90 (describing "critical issue in genericness cases" as whether relevant consumers "primarily use or understand the term sought to be protected to refer to the genus of goods or services in question"); In re Merrill Lynch, Pierce, Fenner, & Smith, Inc., 828 F.2d 1567, 1571 (Fed. Cir. 1987) (finding significant that "substantial number" of publications used term as source-identifying); McCarthy on Trademarks § 12:13 ("[W]here some media uses of a term are as a trademark and some are as a generic name, this source of evidence may be so indeterminate as to be insufficient proof of generic usage."). Of course, "not all bare mentions of the mark are equal." Snyder's Lance, Inc. v. Frito-Lay N. Am., Inc., 542 F. Supp. 3d 371, 385-86 (W.D.N.C. 2021) (discounting term usage in financial reports that "simply reflect[s] the fact that Plaintiffs are marketing and selling the product"). But the relative weight and persuasiveness of the public media evidence are matters to be thoroughly tested at trial, not prematurely decided at summary judgment.
Fourth, the USPTO's competitor use evidence —while probative no doubt—is still not overwhelming enough to dictate summary judgment in its favor on genericness. (ECF 68 at 17-18, 27; ECF 68-4 at 79; CAR at 2248-53). Even setting aside Snap's objections to that evidence (ECF 69 at 27-29; ECF 70-1 at 8-9), two cherrypicked instances of competitor generic use from website shopping searches—in the face of other countervailing evidence in the record—hardly builds the kind of "one-sided" record necessary for a finding of genericness on summary judgment. Retail Servs., 364 F.3d at 546; see, e.g., Pom Wonderful LLC v. Hubbard, 2016 WL 3621281, at *7-8 (C.D. Cal. June 29, 2016) (pages of search engine results were an "anemic showing" of competitor usage insufficient to establish generic use). If the tables were turned, the USPTO would surely ask the court to discount such selectively limited evidence at the summary judgment stage.
Fifth and last, the parties' newly adduced consumer surveys—conflicting as they are on the ultimate issue of consumer perception—must preclude summary judgment in defendants' favor. See Clicks Billiards, Inc. v. Sixshooters, Inc., 251 F.3d 1252, 1263-64 (9th Cir. 2001). For its part, the USPTO offers a Teflon consumer survey. (ECF 68 at 18-20). According to that survey, 79.9% of respondents said that they believe "spectacles" is a generic term in the context of "smart glasses." (ECF 68-6 at 135). On the other hand, Snap presents a Thermos consumer survey, finding that only 1.5% of respondents used the term "spectacles" generically to identify "smart glasses." (ECF 69 at 17-19; ECF 69-5 at 34). Contrary to the USPTO's assertion
that Thermos surveys are not competent evidence (ECF 68 at 28-31; ECF 72 at 13), courts have recognized such surveys to be probative of consumer perception. See, e.g., E.T. Browne Drug Co. v. Cococare Prods., Inc., 538 F.3d 185, 197 (3d Cir. 2008) (Thermos survey created a question of fact on genericness); Nightlight Sys., Inc. v. Nitelites Franchise Sys., Inc., 2007 WL 4563873, at *6-7 (N.D. Ga. July 17, 2007) (finding Thermos survey to be admissible evidence on generieness). After all, it only stands to reason that how consumers use—or don't use—a term can reveal their understanding of the term as applied to a product. The USPTO strenuously resists that idea, asserting that only a Teflon survey should be accepted as the industry standard measure of consumer perception. (ECF 68 at 18, 21-22, 28-29). But that assertion is not only contested by Snap, it also places too much near-dispositive weight on consumer surveys at large. See Booking.com, 140 S. Ct. at 2309 (Sotomayor, J., concurring) ("Flaws in a specific survey design, or weaknesses inherent in consumer surveys generally, may limit the probative value of surveys in determining whether a particular mark is descriptive or generic[.]"). Might the USPTO rue that position if it wants or decides to use a Thermos survey one day? In any event, the relative evidentiary weight to be assigned among the parties' opposing consumer surveys is a matter best addressed through trial.
V.
For all these reasons, defendants' motion for summary judgment (ECF 68) is DENIED.