FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

P AND P IMPORTS LLC, a California No. 21-55013 limited liability company, Plaintiff-Appellant, D.C. No. 8:19-cv-00523- v. DOC-JDE

JOHNSON ENTERPRISES, LLC, DBA

Tailgating Pros, a Virginia limited liability company, Defendant-Appellee,

and

DOES, 1–10, inclusive,

Defendant.

P AND P IMPORTS LLC, a California No. 21-55323 limited liability company, Plaintiff-Appellee, D.C. No. 8:19-cv-00523- v. DOC-JDE

JOHNSON ENTERPRISES, LLC, DBA

Tailgating Pros, a Virginia limited OPINION liability company, Defendant-Appellant. Appeals from the United States District Court for the Central District of California David O. Carter, District Judge, Presiding

Argued and Submitted April 11, 2022 Pasadena, California

Filed August 24, 2022

Before: A. Wallace Tashima and Kenneth K. Lee, Circuit Judges, and Kathleen Cardone,* District Judge.

Opinion by Judge Lee

* The Honorable Kathleen Cardone, United States District Judge for the Western District of Texas, sitting by designation.

P AND P IMPORTS V. JOHNSON ENTERPRISES 3

SUMMARY **

Lanham Act

The district court reversed the district court’s grant of summary judgment in favor of the defendant in an action alleging trade dress infringement, remanded for further proceedings, and dismissed as moot the defendant’s appeal from the denial of attorneys’ fees.

P&P Imports, LLC, maker of a jumbo red-white-and- blue Connect 4 game, sued Johnson Enterprises, LLC, maker of a similar product, for trade dress infringement. The district court granted summary judgment in favor of Johnson Enterprises on the ground that P&P’s trade dress had not acquired secondary meaning, and therefore was not protectable, because consumers did not associate the trade dress with P&P specifically.

The panel held that trade dress does not have to be linked to a particular company. If consumers link the trade dress to any single (even anonymous) source/company, that is enough to constitute secondary meaning. Because P&P’s evidence of intentional copying and a consumer survey created a genuine issue of material fact about whether its trade dress acquired secondary meaning, the panel reversed and remanded.

The panel dismissed Johnson Enterprises’ attorneys’ fees appeal as moot.

** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader. Mark R. Yohalem (argued), Munger Toller & Olson LLP, Los Angeles, California; Xionan April Hu, Munger Toller & Olson LLP, Washington, D.C.; Casey H. Kempner, P&P Imports LLP, Irvine, California; for Plaintiff- Appellant/Cross-Appellee.

James E. Doroshow (argued), Fox Rothschild LLP, Los Angeles, California, for Defendant-Appellee/Cross- Appellant.

OPINION

LEE, Circuit Judge:

Imitation may be the sincerest form of flattery but that does not shield someone from being sued for it. Two competing companies created their own three-feet-wide versions of Connect 4, the classic game in which players drop colored plastic coins into an upright game board in hopes of lodging four coins in a row. This case, however, does not involve the original maker of Connect 4, but rather two companies that lifted the Connect 4 concept to create their own oversized versions.

The question before us is whether a manufacturer’s red- white-and-blue jumbo rendition of this iconic game qualifies as a protectable trade dress. P&P Imports thinks so. It sued a competitor, Johnson Enterprises, whose version of this game looks uncannily like P&P’s. To resolve this question, we must determine whether P&P’s trade dress has acquired “secondary meaning”—i.e., is P&P’s design distinctive enough to be widely recognized in the market?

P AND P IMPORTS V. JOHNSON ENTERPRISES 5

The district court granted summary judgment to Johnson, ruling that P&P’s trade dress had not acquired secondary meaning because consumers do not associate the trade dress with P&P specifically. We reverse because trade dress does not have to be linked to a particular company; if consumers link the trade dress to any single (even anonymous) source/company, that is enough to constitute secondary meaning. And because a genuine issue of material fact exists about whether P&P’s trade dress acquired secondary meaning, we reverse and remand.

BACKGROUND

I. Ten months after P&P begins selling a three-foot,

red-white-and-blue version of Connect 4, Johnson starts selling a virtually identical game.

P&P Imports sells outdoor games and sporting goods under its GoSports brand. One of its games is the GoSports Giant 4 in a Row Game (“P&P Game”), an enlarged, outdoor variation of Connect 4, the classic tabletop game originally made by Milton Bradley (now Hasbro) for nearly 50 years. The P&P Game measures three feet wide and uses a red, white, and blue color scheme.

In December 2016, P&P began selling its game through various e-commerce channels such as Amazon and eBay. In under a year, the P&P Game climbed the best seller ranks in Amazon’s Toys and Games category, racking up significant sales within its category. P&P’s success did not go unnoticed. Sometime in 2017, Johnson Enterprises was looking to expand its product offerings in the Yard Games category and decided it too would produce a giant Connect 4-style game. After conducting market research, Johnson discovered that P&P— the most successful Amazon seller in this product category—was selling 700 units per month. So Johnson bought a copy of the P&P Game and sent samples to its manufacturer in China. In October 2017, ten months after the P&P Game hit the market, Johnson began selling an almost identical game, the Tailgating Pros White Connect 4 game (“Johnson Game”). The P&P and Johnson Games featured their respective logos at the top of the white game boards but otherwise looked nearly identical in color, style, and size.

(Blue Br. at 1)

II. P&P sues Johnson for federal trade dress

infringement under the Lanham Act, and unfair competition under California law.

In March 2019, P&P sued Johnson for damages and injunctive relief, bringing claims of (1) trade dress

P AND P IMPORTS V. JOHNSON ENTERPRISES 7

infringement under section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), (2) unfair competition under section 17200 et seq. of California Business and Professions Code, and (3) unfair competition under California common law.

P&P alleged that Johnson appropriated its trade dress for the “almost identical” Johnson Game. As stated in the complaint, P&P’s trade dress consisted of:

the overall appearance of [the P&P Game] which may be described as a combination of individual features, including, but not limited to the unique color combination of flat-white colored square board with evenly spaced round-hole cut-outs, bordered by a thin bas- relief bezel on all four sides, with two mirrored sculpted legs extending half way up the sides of the bezel and joined to it by tee joints which enfold part of the bezel to create a relief on the bezel and extend depth-wise slightly both frontwards and backwards, and which vertically extend slightly below the bezel where they are joined with the feet to create a relief between them on the outside edge, the feet extend depth-wise from the legs with their flat-top extending into rounded shoulders and squared ends with an arch type shape cut into the bottom-center, which are all contrasted with the smooth, circular flat- red and flat-blue featureless chips game pieces. In other words, P&P’s trade dress is defined by its flat-white game board with circular cut-outs and flat, circular red and blue tokens. 1

According to P&P, its “distinctive” trade dress had become “widely known and recognized” such that it had acquired “secondary meaning.” By copying its trade dress, P&P says Johnson intended to “deceive the public as to the source or origin” of its game to benefit from “P&P’s goodwill and reputation in the four in a row market.”

III. P&P’s expert, Robert Wallace, submits consumer

survey evidence purporting to show secondary meaning.

Robert Wallace submitted an expert report in support of P&P. Wallace designed and conducted a “secondary meaning” survey to measure how much consumers associated P&P’s trade dress with a single source or company. The survey exposed 200 respondents to an image of the P&P Game as it appeared on Amazon, with the GoSports logo and other descriptions of source removed. The respondents were then asked a series of questions.

First, the survey asked, “Do you recognize this product?” Of the 200 respondents, 188 answered “Yes.” Next, the survey asked, “Do you believe that this specific product is made by one company or more than one company?” Ninety- two respondents answered, “One company.” The eighty respondents that answered, “More than one company” were

1 According to the district court, when “[i]gnoring the functional aspects of the P&P Game’s design, the alleged trade dress consists of a ‘flat-white colored square board with evenly spaced round-hole cut-outs . . . contrasted with the smooth, circular flat-red and flat-blue featureless chips game pieces.’”

P AND P IMPORTS V. JOHNSON ENTERPRISES 9

asked, “If you believe the products are sold by more than one company, do you believe that they come from the same source or producer?” 2 Thirty-four said, “Yes.” Wallace then added the “92 respondents who believe that the product comes from one company” to the “34 respondents who believe the product comes from one source or producer” to conclude that 126 of 200—63%—of respondents believe that the P&P Game is “from a single source or company.” In Wallace’s opinion, this “clear majority . . . established secondary meaning.”

Wallace also noted evidence of intentional copying. Deposition testimony revealed that Johnson bought the P&P Game because it was the best-selling product in its category, sent samples to its Chinese manufacturer, and then shortly began selling a nearly identical game. According to Wallace, the “most logical explanation . . . is that [Johnson] set out, and did, copy [P&P’s] trade dress.”

Wallace also mentioned P&P’s advertising efforts. P&P mainly advertised through Amazon’s web-based tools, such as “Deals,” Amazon marketing services (“AMS”), and Amazon marketing allowances. “Deals” is a promotional program in which Amazon features certain products at discounted rates. And AMS allows vendors to advertise their products on the Amazon webpage on a “cost per click” basis. P&P provided Amazon an eight-percent marketing allowance to fund internal and external marketing, such as promotion of P&P’s products on third-party sites. P&P also conducted “grass roots marketing” by displaying the P&P

2 Twenty-eight respondents answered, “Don’t know, not sure” when asked whether they believed the product “is made by one company or more than one company.” Game at Orange County, California beaches on the weekends.

IV. The district court grants summary judgment for

Johnson, ruling P&P failed to present sufficient evidence of secondary meaning.

To prove its trade dress infringement claim, P&P had to show that “(1) the trade dress is nonfunctional, (2) the trade dress has acquired secondary meaning, and (3) there is a substantial likelihood of confusion between [P&P’s] and [Johnson’s] products.” See Art Attacks Ink, LLC v. MGA Ent. Inc., 581 F.3d 1138, 1145 (9th Cir. 2009). In its motion for summary judgment, Johnson argued that P&P had failed to present sufficient evidence on all three elements.

The district court granted Johnson’s motion, ruling that P&P failed to submit sufficient evidence of secondary meaning. According to the district court, our decision in Fleischer Studios, Inc. v. A.V.E.L.A., Inc. required P&P to prove that consumers associate its trade dress with P&P itself, rather than any single (even anonymous) company—a standard which we refer to as “specific association” in this opinion. See 654 F.3d 958, 966–67 (9th Cir. 2011). Therefore, the district court wholesale dismissed the Wallace Report’s survey evidence as irrelevant because the results— that 63% of respondents “believe that Plaintiff’s product is from a single source or company”—did not “show that the trade dress has become associated with [P&P] itself.”

The district court also noted that the “short period of exclusive use (approximately ten months), relatively low volume of sales, and [P&P’s] inconsistent use of the trade dress further cut against any finding of secondary meaning.” And “[w]hatever questions may exist about advertising and intentional copying, they do not rise above the level of a

P AND P IMPORTS V. JOHNSON ENTERPRISES 11

scintilla of evidence.” P&P thus “failed to raise a genuine issue of material fact as to whether its trade dress . . . has acquired secondary meaning.” And because P&P’s state- law claims hinged on its trade dress claim, the district court granted summary judgment for Johnson on all claims.

V. The district court denies P&P’s motion for

reconsideration and Johnson’s motion for attorneys’ fees, and this appeal follows.

P&P moved for reconsideration of the district court’s grant of summary judgment. According to P&P, the district court (1) committed “legal error because the relevant purchasing public need not know the identity . . . of a single source” denoted by the trade dress, and (2) relied on a legal theory not raised by the parties by requiring specific association, thereby denying P&P notice and an opportunity to respond in violation of Federal Rule of Civil Procedure 56(f). The district court denied P&P’s motion, and P&P timely appealed.

Johnson, in turn, moved for attorneys’ fees and costs as the “prevailing party” under the Lanham Act. See 15 U.S.C. § 1117(a). The district court denied the motion because P&P’s case was not “substantially weak.” Although insufficient to preclude summary judgment, P&P presented some evidence of secondary meaning, including “evidence of willful copying,” referring to Johnson’s admission “to buying [P&P’s game] and sending it to [their] Chinese manufacturer.” Johnson timely appealed the district court’s denial of fees. We consolidated both appeals.

STANDARD OF REVIEW

We review de novo a district court’s grant of summary judgment, and we must “determine, viewing the evidence in the light most favorable to the nonmoving party, whether there are any genuine issues of material fact and whether the district court correctly applied substantive law.” Ballen v. City of Redmond, 466 F.3d 736, 741 (9th Cir. 2006) (quoting United States v. City of Tacoma, 332 F.3d 574, 578 (9th Cir. 2003)). We review a district court’s denial of attorneys’ fees and costs under the Lanham Act for an abuse of discretion. Sunearth, Inc. v. Sun Earth Solar Power Co., 839 F.3d 1179, 1181 (9th Cir. 2016) (en banc) (per curiam).

DISCUSSION

I. We reverse the district court’s grant of summary

judgment for Johnson.

Like “distinctive names, logos, packages, or labels,” a product’s design may have a “source-identifying appearance[].” See Blumenthal Distrib., Inc. v. Herman Miller, Inc., 963 F.3d 859, 864 (9th Cir. 2020). Such designs, if distinctive, may receive trade dress protection, and the manufacturer may bring an action for infringement under section 43(a) of the Lanham Act. See Wal-Mart Stores v. Samara Bros., 529 U.S. 205, 209–10 (2000).

Because a product’s design can never be inherently distinctive, a plaintiff must prove that the design has acquired secondary meaning. See id. at 216. The district court granted summary judgment for Johnson on secondary meaning. But the district court applied an incorrect legal standard for determining secondary meaning, and P&P has presented sufficient evidence to survive summary judgment. We thus reverse the district court’s grant of summary judgment and remand for further proceedings.

P AND P IMPORTS V. JOHNSON ENTERPRISES 13

a. The district court applied the wrong legal standard for secondary meaning by requiring evidence of specific association.

Secondary meaning exists when “in the minds of the public, the primary significance of [the trade dress] is to identify the source of the product rather than the product itself.” Inwood Lab’ys v. Ives Lab’ys, 456 U.S. 844, 851 n.11 (1982). The district court required P&P to show that consumers specifically associate the P&P Game’s trade dress with P&P itself. In other words, consumers must both recognize P&P’s trade dress and be able to name P&P as the source. The district court purportedly derived this standard from a single sentence in Fleischer in which we said that the plaintiff “must show that the mark has become identified with the manufacturer . . . .” Fleischer, 654 F.3d at 967 (emphasis added) (internal quotation omitted).

The district court’s interpretation of Fleischer conflicts with our long-established precedents requiring association with only a single—even anonymous—source. See Maljack Prods. v. Goodtimes Home Video Corp., 81 F.3d 881, 887 (9th Cir. 1996) (“[A] showing of secondary meaning only requires proof that the public associates the [mark] with a single source, even if that source is anonymous.”); Bentley v. Sunset House Distrib. Corp., 359 F.2d 140, 147 (9th Cir. 1966) (“To show that some secondary meaning existed, it was necessary for Bentley to establish that the public . . . regard[s] its product as emanating . . . from a single, though anonymous maker.”) (internal quotation omitted)).

Fleischer, decided by a three-judge panel, could not have overturned these earlier binding decisions. Koerner v. Grigas, 328 F.3d 1039, 1050 (9th Cir. 2003). Nor did Fleischer purport to. In the very next sentence, the Fleischer panel clarifies that the “basic element of secondary meaning” is an association “with the same source.” Fleischer, 654 F.3d at 967 (emphasis added) (quoting Levi Strauss & Co. v. Blue Bell, Inc., 632 F.2d 817, 820 (9th Cir. 1980)). When judicial opinions refer to a “single” or “same” source, they are not suggesting that consumers must know “the corporate name of the producer or seller”; rather, they connote that “a single, albeit anonymous, source” suffices. 2 McCarthy on Trademarks and Unfair Competition § 15:8 (5th ed. 2021); see also 8 Trademark Manual of Examining Procedure (TMEP) 1212 (2021) (Secondary meaning only requires association with “an anonymous producer, since consumers often buy goods without knowing the . . . actual name of the manufacturer.”) (quoting Ralston Purina Co. v. Thomas J. Lipton, Inc., 341 F. Supp. 129, 133 (S.D.N.Y. 1972))).

The district court’s reading of Fleischer also clashes with the text of the Lanham Act. Our “anonymous source” test flows directly from the text of that statute, which defines “trademark” as “any word, name, symbol, or device” that “indicate[s] the source of the goods, even if that source is unknown.” 15 U.S.C. § 1127 (emphasis added). Because trade dress is a subcategory of trademarks, see Wal-Mart, 529 U.S. at 209, the same definition applies.

We thus hold that the district court erred by requiring evidence of specific association for secondary meaning.

b. P&P’s intentional copying and consumer survey evidence creates a triable issue of fact about secondary meaning.

We assess many factors to determine whether secondary meaning exists, including: “direct consumer testimony; survey evidence; exclusivity, manner, and length of use of a mark; amount and manner of advertising; amount of sales

P AND P IMPORTS V. JOHNSON ENTERPRISES 15

and number of customers; established place in the market; and proof of intentional copying by the defendant.” Art Attacks, 581 F.3d at 1145. “Because of the intensely factual nature” of the secondary meaning inquiry, “summary judgment is generally disfavored . . . .” See Soc. Techs. LLC v. Apple Inc., 4 F.4th 811, 816 (9th Cir. 2021) (quoting Rearden LLC v. Rearden Com., Inc., 683 F.3d 1190, 1202 (9th Cir. 2012)).

Though we have not prescribed the precise combination of factors necessary to survive summary judgment, we have found the presence of two factors—intentional copying and survey evidence—sufficient. See Clicks Billiards, Inc. v. Sixshooters Inc., 251 F.3d 1252, 1263–64 (9th Cir. 2001) (reversing summary judgment for defendants because plaintiff provided “sufficient evidence of secondary meaning in the form of both a consumer survey and testimony [about intentional copying]”). By submitting evidence of intentional copying and an admissible consumer survey, P&P created a triable issue of fact about secondary meaning.

i. Johnson’s intentional copying strongly suggests that P&P’s trade dress acquired secondary meaning.

“[P]roof of copying strongly supports an inference of secondary meaning.” Vision Sports, Inc. v. Melville Corp., 888 F.2d 609, 615 (9th Cir. 1989). That is because competitors generally copy “to realize upon a secondary meaning that is in existence.” Audio Fid., Inc. v. High Fid. Recordings, Inc., 283 F.2d 551, 558 (9th Cir. 1960). Johnson conducted market research, ordered a copy of the best-selling three-foot Connect 4-style game made by P&P, sent samples of the P&P Game to its Chinese manufacturer, and began selling a nearly identical game mere months later. This chronology strongly suggests that Johnson intentionally copied the P&P Game.

Johnson does not deny copying but discounts its relevance. While competitors usually copy to appropriate secondary meaning, we have recognized that they may also “copy product features” that are “wholly functional . . . because of those features’ intrinsic economic benefits.” Fuddruckers, Inc. v. Doc’s B.R. Others, Inc., 826 F.2d 837, 844–45 (9th Cir. 1987) (holding that a jury may but is not required to infer secondary meaning from copying). Johnson thus argues that “intentional copying supports a finding of secondary meaning only where the defendant intended to confuse consumers and pass off its product as the plaintiff’s. (quoting Cont’l Lab’y Prods. v. Medax Int’l, Inc., 114 F. Supp. 2d 992, 1010 (S.D. Cal. 2000)).” According to Johnson, there is no evidence of such intent.

Though some circuits have adopted Johnson’s argument about an intent to confuse requirement, see, e.g., Craft Smith, LLC v. EC Design, LLC, 969 F.3d 1092, 1110 (10th Cir. 2020), we have not done so. We have only held that an intent to confuse is required for establishing likelihood of consumer confusion, a separate element of a trade dress claim. See AMF, Inc. v. Sleekcraft Boats, 599 F.2d 341, 349, 354 (9th Cir. 1979). Admittedly though, we have also recognized that “[s]econdary meaning can also be established by evidence of likelihood of confusion” because they are “related determinations . . . rising from the same evidentiary findings.” Transgo, Inc. v. Ajac Transmission Parts Corp., 768 F.2d 1001, 1015–16 (9th Cir. 1985) (emphasis added).

But even if an intent to confuse is required for the intentional copying factor under secondary meaning, Johnson’s argument still falls short. An intent to confuse

P AND P IMPORTS V. JOHNSON ENTERPRISES 17

may be inferred when the defendant copies a product’s design and marketing. See Leatherman Tool Grp., Inc. v. Cooper Indus., 199 F.3d 1009, 1013 (9th Cir. 1999) (citing Versa Prods. Co. v. Bifold Co. (Mfg.), 50 F.3d 189, 207–08 (3d Cir. 1995)). While the universe of functionally optimal product designs may be limited, there are many ways to market a product to consumers. Thus, precise copying of the plaintiff’s marketing suggests that the defendant intended to “pass off its product as the plaintiff’s.” Cont’l, 114 F. Supp. 2d at 1010.

Johnson copied much of P&P’s product description. Like P&P, Johnson says the game board is made from “premium wood”; the tokens are made from “durable plastic” that “will never break”; the carrying case was “durable” or “robust”; the game will provide “giant” or “jumbo” fun for “kids and adults of all ages”; and the game measures “3 feet,” even though the Johnson Game is slightly smaller. While we would expect Johnson to describe the game’s materials, durability, and dimensions, its decision to crib identical language from P&P’s advertisement suggests that Johnson intentionally cast its game as P&P’s.

There is also reason to believe that Johnson’s attempt to confuse consumers will succeed. The two games are almost identical, apart from their logos displayed on the product. See Adidas Am., Inc. v. Skechers USA, Inc., 890 F.3d 747, 755 (9th Cir. 2018) (“[T]he greater the similarity . . . the greater the likelihood of confusion.” (quotation omitted)). These nearly identical games are sold through the same e- commerce channels—Amazon and eBay. See Interstellar Starship Servs. v. Epix, Inc., 184 F.3d 1107, 1110 (9th Cir. 1999) (overlapping internet marketing channels are likely to cause confusion). Thus, consumers are likely to encounter these substantially similar games “at the same time, on the same screen,” compounding the risk of confusion. See GoTo.com, Inc. v. Walt Disney Co., 202 F.3d 1199, 1207 (9th Cir. 2000). 3

Ultimately, a jury may find that Johnson’s copying of P&P’s Game and its marketing does not establish secondary meaning. Indeed, it may appear a bit rich that P&P accuses Johnson of copying its game when both are essentially oversized knock-offs of Connect 4. But at the summary judgment stage, we must draw all reasonable inferences in favor of P&P. See Ballen, 466 F.3d at 741. Johnson’s intentional copying of P&P’s trade dress and marketing strongly suggests that secondary meaning exists. See Vision Sports, 888 F.2d at 615. The presence of multiple confusion factors further supports this inference. See Transgo, 768 F.2d at 1015–16.

3 Additionally, P&P provides some evidence of actual confusion in the form of two consumer surveys. See Thane Int’l v. Trek Bicycle Corp., 305 F.3d 894, 902 (9th Cir. 2002) (“Survey evidence may establish actual confusion.”). Both surveys had a “test” and “control” group. The test groups were shown side-by-side pictures of the P&P and Johnson games, with logos and branding removed. The control groups were shown side-by-side pictures of the P&P Game and a different competitor’s game with a “markedly different” trade dress. The respondents were asked whether they believed the products “come from a single source or company” to measure whether the consumer was confused. The disparity between the amount of confusion in the test and control groups—the “net confusion level”—was calculated. The net confusion level was 18% and 16.5% for the first and second surveys, respectively. We have previously found a net confusion level of 11% sufficient to preclude summary judgment on the issue of confusion. See Fortune Dynamic, Inc. v. Victoria’s Secret Stores Brand Mgmt., 618 F.3d 1025, 1036–38 (9th Cir. 2010).

P AND P IMPORTS V. JOHNSON ENTERPRISES 19

ii. The consumer survey is admissible and supports a finding of secondary meaning.

P&P submitted a consumer survey showing that 63% of respondents said that the P&P Game comes from a single source/company. According to P&P’s expert, this strongly suggests that P&P’s trade dress acquired secondary meaning. See Vision Sports, 888 F.2d at 615 (“An expert survey of purchasers can provide the most persuasive evidence of secondary meaning.”).

Consumer surveys are evaluated in two-steps. See Clicks Billiards, 251 F.3d at 1263. First, a court must determine whether the survey is admissible by ensuring that the survey has “a proper foundation . . . and is . . . relevant and conducted according to accepted principles.” Id. Second, once the survey is admitted, issues about “methodology, survey design, reliability, the experience and reputation of the expert, critique of conclusions, and the like go to the weight of the survey rather than its admissibility” and are “for a jury” to decide. Id.

The survey was relevant and admissible at step one. The district court erroneously dismissed P&P’s consumer survey as irrelevant because it failed to prove specific association. But as we explained, secondary meaning requires association with only a single, anonymous source. This is precisely what the survey attempts to measure.

Still, Johnson argues that the survey is irrelevant because it was conducted two-and-a-half years after the Johnson Game was first sold, so the survey does not “measure public perception at the time of Johnson’s first alleged infringing use.” But P&P was not required to preemptively conduct consumer surveys in anticipation of litigation, see Gen. Motors Corp. v. Lanard Toys, Inc., 468 F.3d 405, 419 (6th Cir. 2006), and we have admitted surveys conducted years after the first alleged infringing use, see, e.g., Faberge, Inc. v. Saxony Prods., Inc., 605 F.2d 426, 428 (9th Cir. 1979) (admitting survey from 1974 to gauge secondary meaning in 1970). Surveys conducted within five years of the first infringing use are generally relevant, and the time (zero to five years) between the first infringing use and the survey goes to the weight of the survey evidence. See Converse, Inc. v. ITC, 909 F.3d 1110, 1123 (Fed. Cir. 2018). Here, Wallace conducted his survey two-and-a-half years after Johnson began selling its allegedly infringing product. Wallace’s survey is well within the outer limit of temporally relevant secondary meaning surveys.

To be fair, Wallace’s survey methodology raises some questions. Connect 4 has been sold by Milton Bradley (and now by Hasbro) for nearly 50 years. Large swaths of the public were already familiar with Connect 4 when the P&P Game hit the market. When respondents were shown a picture of the P&P Game—which, apart from its color, substantially resembles Connect 4—some may have recognized the classic Connect 4 design, not P&P’s red- white-and-blue trade dress. This possibility is supported by the fact that, when the survey asked which company made the P&P Game, only 3 out of 200 respondents said GoSports. So when these respondents said the P&P Game comes from a single source, they may have been referring to Hasbro, not P&P. This would mean that the overall look and feel of Connect 4—and not P&P’s trade dress—has a source- identifying appearance.

But Johnson failed to challenge the survey’s design under Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). See Southland Sod Farms v. Stover Seed Co., 108 F.3d 1134, 1143 n.8 (9th Cir. 1997) (Defendants may

P AND P IMPORTS V. JOHNSON ENTERPRISES 21

object under Daubert “to the technical reliability of a survey.”); In re Ford Tailgate Litig., No. 11-cv-02953-RS, 2015 U.S. Dist. LEXIS 159534, at *19 n.3 (N.D. Cal. Nov. 25, 2015) (“Daubert dictates the analysis necessary for a motion to exclude expert testimony in the context of a motion for summary judgment.”). And in any event, these criticisms arguably bear more on the survey’s persuasiveness, not its admissibility. See Clicks Billiards, 251 F.3d at 1263.

Because P&P has presented compelling evidence of intentional copying and an admissible consumer survey, a triable issue of fact exists on secondary meaning. We thus reverse the district court’s grant of summary judgment on all claims. 4

II. We dismiss as moot Johnson’s appeal of the district

court’s denial of attorneys’ fees.

In “exceptional cases,” a district court has discretion to award attorneys’ fees to the “prevailing party” under the Lanham Act. See 15 U.S.C. § 1117(a). Because we reverse summary judgment, Johnson is no longer the prevailing party, so we dismiss as moot Johnson’s appeal of the district court’s denial of fees. See Grouse River Outfitters, Ltd. v. Oracle Corp., 848 F. App’x 238, 245 (9th Cir. 2021).

4 The parties dispute whether the remaining Art Attacks factors, such as the amount of sales, the amount of advertising, and the exclusivity, manner, and length of use of the trade dress weigh in favor of secondary meaning. See Art Attacks, 581 F.3d at 1145. Because the intentional copying and consumer survey evidence here are sufficient to preclude summary judgment, we need not address these additional factors. The district court erred by requiring evidence of specific association. We hold that P&P’s intentional copying and survey evidence were sufficient to create a triable issue of fact about secondary meaning. We REVERSE the district court’s grant of summary judgment, and REMAND for further proceedings consistent with this opinion. 5 Johnson’s appeal is DISMISSED as moot. P&P shall recover its costs on appeal.

5 As alternative grounds for affirmance, Johnson argues that P&P has failed to prove nonfunctionality and the likelihood of consumer confusion. See L.A. News Serv. v. CBS Broad., Inc., 305 F.3d 924, 934 (9th Cir. 2002) (Summary judgment may be affirmed by any ground supported in the record.). Because the remaining elements of P&P’s trade dress claim are also “intensely factual issue[s],” see Zobmondo Ent., LLC v. Falls Media, LLC, 602 F.3d 1108, 1113 (9th Cir. 2010) (quoting KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 408 F.3d 596, 605 (9th Cir. 2005)), and “the district court is in a better position to develop the facts and assess their legal significance in the first instance,” Alexander v. Newland, 20 F. App’x 662, 663 (9th Cir. 2001) (quoting Whalem/Hunt v. Early, 233 F.3d 1146, 1148 (9th Cir. 2000)), we remand for the district court to address these elements as necessary.