Opinion
No. 18-14975
10-20-2020
J-B WELD COMPANY, LLC, Plaintiff – Appellant, v. The GORILLA GLUE COMPANY, Defendant – Appellee.
Neil C. Jones, Nelson Mullins Riley & Scarborough, LLP, Greenville, SC, Ashley B. Summer, Nelson Mullins Riley & Scarborough, LLP, New York, NY, Mark S. VanderBroek, Lucas Allen Westby, Nelson Mullins Riley & Scarborough, LLP, Atlanta, GA, for Plaintiff–Appellant. April Lynn Besl, Govinda Mirabai Davis, Dinsmore & Shohl, LLP, Cincinnati, OH, Joseph Schaeff, Susan D. Solle, Dinsmore & Shohl, LLP, Dayton, OH, Norman A. Crain, Wesley A. Roberts, Thomas Horstemeyer LLP, Atlanta, GA, for Defendant–Appellee.
Neil C. Jones, Nelson Mullins Riley & Scarborough, LLP, Greenville, SC, Ashley B. Summer, Nelson Mullins Riley & Scarborough, LLP, New York, NY, Mark S. VanderBroek, Lucas Allen Westby, Nelson Mullins Riley & Scarborough, LLP, Atlanta, GA, for Plaintiff–Appellant.
April Lynn Besl, Govinda Mirabai Davis, Dinsmore & Shohl, LLP, Cincinnati, OH, Joseph Schaeff, Susan D. Solle, Dinsmore & Shohl, LLP, Dayton, OH, Norman A. Crain, Wesley A. Roberts, Thomas Horstemeyer LLP, Atlanta, GA, for Defendant–Appellee.
Before BRANCH, TJOFLAT, and ED CARNES, Circuit Judges.
TJOFLAT, Circuit Judge:
This case arises from a dispute over product packaging and marketing between two manufacturers of two similar heavy-duty adhesives. The plaintiff, J-B Weld Company, appeals from the District Court's grant of summary judgment on all counts of its amended complaint in favor of the defendant, the Gorilla Glue Company. J-B Weld's amended complaint is premised on the following claims: (1) trade dress infringement under the Lanham Act, Georgia law, and the common law of unfair competition; (2) trade dress dilution under Georgia law; and (3) false advertising under the Lanham Act and Georgia law. For the reasons that follow, we affirm the grant of summary judgment for Gorilla Glue as to J-B Weld's false advertising claims, but reverse and remand with respect to its trade dress infringement and trade dress dilution claims.
The District Court's grant of summary judgment in favor of Gorilla Glue on the claims in J-B Weld's amended complaint is before us as a final judgment pursuant to Fed. R. Civ. P. 54(b). The District Court left one permissive counterclaim asserted by Gorilla Glue unresolved on summary judgment, but found the criteria of Rule 54(b) satisfied and certified the judgment as to all other claims as final and immediately appealable. See part II, infra .
I.
J-B Weld and Gorilla Glue are both manufacturers specializing in adhesives and other similar products. J-B Weld has manufactured and sold a "two-part epoxy adhesive" since 1969, a formula known as "J-B Weld Original" or "J-B Weld Cold Weld." Two-part epoxy adhesives, like J-B Weld Original, differ from traditional adhesives because they require creating a mix of two different pastes, a resin and a hardener, which is then applied as any other adhesive would be. When the two pastes are mixed together, a chemical reaction begins to occur and, once the adhesive has set, results in a particularly strong and temperature-resistant bond. On a molecular level, J-B Weld Original's resin paste contains members of "epoxy groups," which belong to a chemically distinct category of polymer that catalyzes the curing process – that is, the transformation from liquid goop to rock-solid bond. J-B Weld Original's epoxy resin paste also contains iron dust as a filler, purportedly to strengthen and support the adhesive when cured, which J-B Weld recognizes by referring to the product as a "steel reinforced epoxy." J-B Weld indicates that its two-part epoxy is suitable to bond a variety of surfaces to one another effectively: "Metal, Wood, Plastic, Tile, PVC, Ceramic, Fiberglass, Concrete & More."
Central to this litigation is the way in which J-B Weld Original is packaged. J-B Weld describes its product's packaging, or trade dress, as follows: (1) two squeezable tubes in a blister package, with the tubes angled inwardly to create a "V-shape;" (2) a black-bannered tube on the left side of the package and a red-bannered tube on the right side of the package; (3) black and white caps on each respective tube; (4) a clear "blister" style protective package that angles inward in the same manner as the tubes; (5) a background card with a width of five inches; (6) a "technical information box," located in between the two tubes on the background card, including four lines of information separated by white lines; (7) colored banners stretching across the top and bottom portions of the background card; (8) the capitalized/emphasized word "WELD" inside the upper banner on the background card; (9) a list of potential uses for the product in the bottom-right corner of the background card. J-B Weld has used this particular packaging design continuously since 2012, though it notes that some of the listed features have been a part of J-B Weld Original's packaging since the product's inception in 1969. J-B Weld Original's current packaging is reproduced below.In 2017, Gorilla Glue introduced a two-tube adhesive under the brand name "GorillaWeld." Similar to J-B Weld Original, GorillaWeld supplies users with two tubes, a resin and a hardener, which are to be mixed together in order to begin the curing process of the adhesive. But GorillaWeld's adhesive differs from J-B Weld's in that GorillaWeld uses methyl methacrylate chemistry ("MMA"), which, chemically, is not an epoxy-group polymer, despite the similarities in the reactive process. GorillaWeld's resin also does not contain any iron or steel. GorillaWeld is currently being marketed and sold as an "epoxy" adhesive and, on its packaging, is advertised as a "steel bond epoxy."
Gorilla Glue recognized that it was entering the epoxy adhesive market segment as a competitor of J-B Weld. It expressly considered this fact when designing GorillaWeld's trade dress and planning the GorillaWeld marketing campaign. After considering numerous potential designs for the GorillaWeld packaging, Gorilla Glue ultimately settled on the design shown below. As Gorilla Glue's graphic designer stated: "The objective of this project was to go straight up against the top competitor (J-B Weld) and create packaging that mimics the competitor's architecture. I was able to pull subtle elements into our package, but still keep the package looking tough and geared towards the Gorilla brand." GorillaWeld's current packaging is reproduced below:
II.
On October 6, 2017, J-B Weld filed a three-count complaint in the Northern District of Georgia, alleging that Gorilla Glue was infringing on J-B Weld's trade dress rights in violation of (1) Section 43(a) of the Lanham Act (Count I), (2) Georgia common law of unfair competition (Count II), and (3) Georgia statutory law, O.C.G.A. § 10-1-372 (Count III). According to J-B Weld, its trade dress is entitled to protection under these three doctrines because it "is not functional" and "is either inherently distinctive and/or has acquired secondary meaning." J-B Weld further claims that the similarities between the J-B Weld Original and GorillaWeld packaging are likely to confuse consumers into believing that J-B Weld produces, or is otherwise connected to, the GorillaWeld product. J-B Weld moved the District Court for a preliminary injunction preventing Gorilla Glue from further use of the GorillaWeld packaging design. The Court denied J-B Weld's motion, finding that J-B Weld had not shown a substantial likelihood of success on the merits – i.e., that Gorilla Glue was infringing J-B Weld's trade dress rights. The Court reasoned that, while J-B Weld Original's packaging was entitled to "at least moderate protection," J-B Weld would be unable to prove that any packaging similarities between J-B Weld Original and GorillaWeld were likely to confuse consumers, in large part because "the overall impression of the [two trade dresses] is not similar."
In this litigation, "trade dress" refers to the packaging elements of J-B Weld Original and GorillaWeld, as reproduced above.
J-B Weld amended its complaint, adding three more counts while essentially replicating the three counts of the original complaint. J-B Weld's six-count amended complaint contained the following allegations: (1) trade dress infringement based on the Lanham Act (Count I, tracking Count I of the original complaint), Georgia common law of unfair competition (Count III, tracking Count II of the original complaint), and O.C.G.A. § 10-1-372 (Count IV, tracking Count III of the original complaint); (2) trade dress dilution based on O.C.G.A. § 10-1-451 (Count V); and (3) false advertising under the Lanham Act (Count II) and O.C.G.A. § 10-1-421 (Count VI). As a basis for Counts II and VI, the amended complaint added the new allegation that Gorilla Glue was falsely advertising GorillaWeld as an epoxy adhesive containing steel by describing the product as a "Steel Bond Epoxy," thereby materially misleading consumers about the nature of the product.
In its answer, Gorilla Glue denied the allegations of wrongdoing, pled sixteen affirmative defenses, and presented a four-count counterclaim alleging that J-B Weld was deceiving consumers by describing J-B Weld Original as the "World's Strongest Bond" and as "steel reinforced." Both parties moved for summary judgment; Gorilla Glue on all of J-B Weld's claims, and J-B Weld on Gorilla Glue's counterclaims. The District Court granted Gorilla Glue's motion in its entirety, and partially granted J-B Weld's motion. The District Court found that (1) J-B Weld failed to show a sufficient likelihood of confusion to survive summary judgment on its trade dress infringement claims presented in Counts I, III, and IV, (2) J-B Weld had not shown trade dress dilution under Georgia law based on the claim presented in Count V; and (3) Gorilla Glue was entitled to summary judgment on J-B Weld's false advertising claims under both the Lanham Act (Count II) and Georgia law (Count VI) because J-B Weld had not put forth any evidence that the statements at issue were material to consumer purchasing decisions. J-B Weld appeals the District Court's summary judgment in favor of Gorilla Glue.
Gorilla Glue's affirmative defenses included allegations that J-B Weld Original's trade dress is not inherently distinctive, lacks secondary meaning, is functional, and is unprotectable under the Lanham Act. Gorilla Glue also asserted that J-B Weld's claims are barred by the doctrine of unclean hands due to J-B Weld's own false advertising and trademark infringement, as alleged in Gorilla Glue's counterclaims. Gorilla Glue asserted that J-B Weld had not proven any likelihood of confusion between the two products and had failed to show any damages entitling it to relief, nor had it shown that GorillaWeld's use of the phrase "steel bond epoxy" was deceptive and misleading to consumers.
The counterclaims alleged in Count One that the GorillaWeld trade dress was non-infringing under the Lanham Act and Georgia common law, and requested a declaratory judgment in favor of Gorilla Glue on that ground; in Count Two that the GorillaWeld trade dress was non-infringing under O.C.G.A. § 10-1-371, and requested a declaratory judgment in favor of Gorilla Glue on that ground; in Count Three that J-B Weld was infringing Gorilla Glue's trademark rights under the Lanham Act and Georgia common law by using the phrase "Impact Tough" on its SuperWeld product; and in Count Four that J-B Weld Original was falsely advertised as the "World's Strongest Bond" and "steel reinforced" in violation of the Lanham Act, 15 U.S.C. § 1125(a). None of these counts is before us in this appeal.
The parties’ summary judgment briefing consolidated the state and federal trade dress infringement claims because the parties acknowledged that the state-law variants of these claims are governed by the federal Lanham Act standard. The District Court's summary judgment order accordingly treated the trade dress infringement claims (Counts I, III, and IV) and the false advertising claims (Counts II and VI) as identical, similarly acknowledging that these claims are evaluated under the same standards whether based on Georgia law or federal law. See Univ. of Ga. Athletic Ass'n v. Laite , 756 F.2d 1535, 1539 n.11 (11th Cir. 1985) ; B&F Sys., Inc. v. LeBlanc , 2011 WL 4103576 at 21 n.13 (M.D. Ga. Sept. 14, 2011).
We note that the District Court's order states that J-B Weld's motion is denied in its entirety – on both the "World's Strongest Bond" and "Steel Reinforced" counterclaims. The Court later amended the judgment to correct the error and reflect that, consistent with its legal analysis, it was granting J-B Weld summary judgment on the "World's Strongest Bond" claim, but denying summary judgment on the "Steel Reinforced" claim. J-B Weld's other counterclaims had been abandoned, as they had not been argued in any of the briefing before the Court.
While the District Court's analysis treated these three claims as one indistinguishable claim for "trade dress infringement," see n.5, supra , we will continue to reference the three claims separately to be faithful to the delineations present in J-B Weld's amended complaint.
On motion by J-B Weld, the District Court certified the resolved claims as a final judgment pursuant to Fed. R. Civ. P. 54(b). Thus, the claims resolved on summary judgment were immediately appealable, despite the existence of one counterclaim—the "Steel Reinforced Claim"—that was not resolved on summary judgment. The unresolved claim was stayed by the District Court's order.
On appeal, J-B Weld argues that the District Court improperly "evaluat[ed]" and "weigh[ed]" the evidence in determining that there was no genuinely disputed issue of material fact on any of J-B Weld's claims, and that the Court thereby applied the "improper legal standards in granting summary judgment." First, regarding the trade dress infringement claims presented in Counts I, III, and IV, J-B Weld contends that the District Court improperly rejected evidence of actual confusion and failed to give any weight to evidence tending to show that Gorilla Glue "intentionally chose to copy" aspects of J-B Weld Original's trade dress. Additionally, J-B Weld contends that the District Court erred by giving improper weight to its conclusion that the two trade dresses were dissimilar in its likelihood of confusion analysis and failing to consider other relevant factors. As a result, J-B Weld says, the District Court's conclusion that there was no likelihood of confusion between the two products’ trade dresses, and therefore no cognizable infringement claim, was error.
With respect to J-B Weld's trade dress dilution claim under Georgia law, presented in Count V, J-B Weld argues that the District Court applied the wrong legal standard by requiring proof of likelihood of confusion as a necessary element of the claim.
Finally, concerning J-B Weld's false advertising claims, presented in Counts II and VI, J-B Weld argues that a factual dispute exists as to whether GorillaWeld's "Steel Bond Epoxy" descriptor is material to purchasing decisions, and that the District Court erred in granting summary judgment on the false advertising claim based on its conclusion that such statements were not shown to be material.
III.
We review a grant of summary judgment de novo, "viewing all facts in the light most favorable to the nonmoving party and drawing all reasonable inferences in favor of that party." McCullum v. Orlando Reg'l Healthcare Sys., Inc. , 768 F.3d 1135, 1141 (11th Cir. 2014). Summary judgment is appropriate when "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). Given this posture, we must evaluate all the record evidence in the manner most favorable to J-B Weld's claims.
a.
We first review the District Court's grant of summary judgment on J-B Weld's trade dress infringement claims, as presented in Counts I, III, and IV of its amended complaint. Although the posture of the case required the District Court to view the evidence in the light most favorable to J-B Weld, the District Court failed to do so in analyzing the "likelihood of confusion" between J.B. Weld Original's trade dress and GorillaWeld's trade dress. Therefore, we reverse the Court's disposition with respect to these claims.
Section 43(a) of the Lanham Act forbids a seller of any good or service from using any "word, term, name, symbol, or device" in connection with their product that is "likely to cause confusion" with the goods or services of another seller, including confusion as to the "origin, sponsorship, or approval" of the product. 15 U.S.C. § 1125(a)(1)(A). The statutory terms "word, term, name, symbol, or device" are referred to as designations of origin — marks used to "identify and distinguish the source of goods and services of a person or company." See J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition §§ 4:3 ; 27:18 (5th ed. 2019) (internal quotations omitted). One such designation of origin is a product's trade dress, which "involves the total image of a product and may include features such as size, shape, color or color combinations, texture, graphics, or even particular sales techniques." John H. Harland Co. v. Clarke Checks, Inc. , 711 F.2d 966, 980 (11th Cir. 1983). Typically, "trade dress" refers to the "packaging or labeling of goods." Id. It is well settled that the Lanham Act's protections against infringement apply to trade dress. Id.
A plaintiff must prove each of the following three elements to make out a trade dress infringement claim: (1) "its trade dress is inherently distinctive or has acquired secondary meaning"; (2) "its trade dress is primarily non-functional"; and (3) the defendant's trade dress is so similar to the plaintiff's that it is likely to cause confusion. AmBrit, Inc. v. Kraft, Inc. , 812 F.2d 1531, 1535 (11th Cir. 1986). The District Court concluded that J-B Weld failed to satisfy the third prong—likelihood of confusion between J-B Weld Original's trade dress and GorillaWeld's trade dress. Thus, J-B Weld could not make out a claim of trade dress infringement under the Lanham Act, the common law of unfair competition, or O.C.G.A. § 10-1-372. In our Circuit, we typically look to seven factors to determine whether there is a likelihood of confusion between two trade dresses: (1) the strength of the trade dress, (2) the similarity of design, (3) the similarity of the product, (4) the similarity of retail outlets and purchasers, (5) the similarity of advertising media used, (6) the defendant's intent, and (7) actual confusion. Dippin’ Dots, Inc. v. Frosty Bites Distribution, LLC , 369 F.3d 1197, 1207 (11th Cir. 2004) (citing AmBrit , 812 F.2d at 1538 ). While all seven factors must be considered, they are not necessarily exhaustive if other evidence is probative of a likelihood of confusion. See Tana v. Dantanna's , 611 F.3d 767, 781 (11th Cir. 2010) (holding that "[t]he district court did not err in considering the geographic proximity of use as an eighth factor demonstrating the unlikelihood of confusion.") Because this test "presupposes that various factors will point in opposing directions," it is the job of the Court to determine the relative importance of the evidence probative of each factor in an effort to decide whether, "in light of the evidence as a whole, there is sufficient proof of a likelihood of confusion to warrant a trial of the issue." Id. at 775 n.7. Given that we are reviewing the summary judgment the District Court granted Gorilla Glue, we must evaluate the evidence pertaining to each of these factors, and the relative importance of each factor, in a way most favorable to J-B Weld's position that confusion was likely.
As previously noted, see nn.5, 7, supra , the District Court opted not to delineate the three separate causes of action for trade dress infringement presented by Counts I, III, and IV because our case law signals that all three causes require proof of likelihood of confusion. See Univ. of Ga. Athletic Ass'n v. Laite , 756 F.2d 1535, 1539 n.11 (11th Cir. 1985) (noting that the standards governing deceptive and unfair trade practice claims under Georgia law are "similar, if not identical, to those under the Lanham Act."); Energy Four, Inc. v. Dornier Med. Sys., Inc. , 765 F. Supp. 724, 731 (N.D. Ga. 1991) ("The [Georgia] Uniform Deceptive Trade Practices Act involves the same dispositive questions as the Federal Lanham Act." (citing Jellibeans, Inc. v. Skating Clubs of Ga., Inc. , 716 F.2d 833, 839 (11th Cir. 1983) )).
J-B Weld argues that the District Court failed to properly credit evidence relevant to three of the seven enumerated factors: the similarity of the designs, the defendant's intent, and actual confusion. Additionally, J-B Weld contends that the Court failed to adequately address all seven of the enumerated likelihood of confusion factors in its analysis. We consider these arguments in turn.
i.
First, we discuss the District Court's treatment of the "similarity of the designs" factor. The Court determined that any similarities between the two product packaging designs are insufficient to create an issue of fact because the presence of names, logos, and different color schemes on each establishes, as a matter of law, that the two trade dresses were not similar. In so doing, the Court found — as it had to — that there is no possibility that any reasonable juror could have found that the two designs were similar.
We have stated that the similarity of two designs is determined by "considering the overall impression created by the mark[s] as a whole," and that it is a "subjective eyeball test." AmBrit , 812 F.2d at 1540. The appropriate comparative process evaluates the mark's general impression and is not a "dissection of individual features." Sun-Fun Prods., Inc. v. Suntan Research & Dev. Inc. , 656 F.2d 186, 189 (5th Cir. Unit B 1981).
The District Court acknowledged the similarities between the packages — "the V-shape tube arrangement, the use and emphasis of the term ‘Weld,’ and the location of the product specifications." The District Court found, however, that despite those similarities, the presence of the Gorilla Glue logo, brand name, and color scheme meant that no reasonable person could find that "the overall impression created by the [GorillaWeld] mark as a whole" was similar to the impression created by the J-B Weld Original trade dress. AmBrit , 812 F.2d at 1540.
Here, reasonable minds could disagree as to which of these features contributes most to the overall impression conveyed by the two marks. Where one consumer may think that the color scheme and Gorilla Glue logo are central to the trade dress's impression, another consumer may believe that the particular placement and angling of the black-and-red labeled tubes, the identical location of product specifications such as hold strength and set and cure time, and the presence of "WELD" in large, bolded text comprised the primary impression of the two products’ packaging. With this amount of conflicting evidence as to the similarity of the two designs, it was error for the District Court to conclude that, as a matter of law, J-B Weld had not shown that the two products’ trade dress designs were similar.
ii.
Second, we consider J-B Weld's argument that the District Court erred in evaluating evidence of Gorilla Glue's intent to copy J-B Weld Original's trade dress. We agree with J-B Weld; the Court erred in its treatment of the "defendant's intent" factor by failing to draw all inferences in J-B Weld's favor and by failing to make J-B Weld's "best case." See Stephens v. DeGiovanni , 852 F.3d 1298, 1313–14 (11th Cir. 2017).
A defendant's intention to bolster its own reputation by trading off of the goodwill associated with the plaintiff's trade dress supports a finding of likelihood of confusion. AmBrit , 812 F.2d at 1542. Intent to benefit from a competitor's goodwill can rarely be proven with a smoking gun, so "[t]here is nothing unusual about a finding of intent based on circumstantial evidence." Id. at 1543. Thus, while we have held that there is a distinction "between intentional copying and intentional copying with intent to cause confusion," Yellowfin Yachts, Inc. v. Barker Boatworks , LLC , 898 F.3d 1279, 1293 (11th Cir. 2018) (citation and emphasis omitted), a finder of fact may nevertheless infer intent to derive a benefit from a competitor's goodwill—and, accordingly, an intent to cause confusion—from evidence of intent to copy. AmBrit , 812 F.2d at 1543 (citing John H. Harland Co. v. Clarke Checks, Inc. , 711 F.2d 966, 977 n.16 (11th Cir. 1983) ). Indeed, we have stated that where a defendant attempts to copy a plaintiff's product "as closely as possible" and uses the plaintiff's product design as a model, it may be "inferred that [defendant] purposely chose a mark which was very similar to [plaintiff's] in order to benefit from the reputation [plaintiff]’s mark had already achieved." John H. Harland Co. , 711 F.2d at 977. Such a finding of intent is "especially fitting" where a review of the trade dress of the plaintiff and defendant "reveals substantial similarities." AmBrit , 812 F.2d at 1543.
In this case, J-B Weld presented evidence in the form of communications from Gorilla Glue's packaging design team that repeatedly referenced J-B Weld Original's packaging and expressed a desire to use similar elements for GorillaWeld's packaging. Emails reveal that the GorillaWeld design team considered three package designs: one that was "[a]ligned" to Gorilla brand, one that would "follow" the J-B Weld brand "closely," and one that would "marry[ ]" the two designs. The team further described certain GorillaWeld packaging options as "[c]lose to JB Weld brand" and aspiring to "go[ ] directly after [J-B Weld Original]," and the team stated their target market was consumers that had used the J-B Weld Original product in the last six months. In fact, one Gorilla Glue employee later called the GorillaWeld design a "knock off" of J-B Weld.
J-B Weld contends that this evidence shows—or, at the very least, that a finder of fact could infer from this evidence—that Gorilla Glue intended to copy those elements of the J-B Weld Original trade dress that were reputationally beneficial and recognizable to consumers who were familiar with J-B Weld Original. The District Court rejected this argument, implicitly finding that such an inference was unreasonable. In doing so, the District Court placed significant weight on Heather Tonne's—Gorilla Glue's graphic designer—performance review, where Tonne stated, "I was able to pull subtle elements [of J-B Weld's Dress] into our package, but still keep our package looking tough and geared towards the Gorilla brand." Relying on Tonne's review, the Court found that "Gorilla Glue did not intend to confuse consumers" because Gorilla Glue's "use of its own well-known color scheme and distinct logo clearly indicates that Gorilla Glue did not intend to confuse consumers as to GorillaWeld's origin." Consequently, the District Court held that the "intent" factor did not favor a likelihood of confusion. This was error.
A comparison of this case to Yellowfin Yachts, Inc. v. Barker Boatworks, LLC highlights the missteps in the District Court's analysis. In Yellowfin Yachts , Yellowfin, a manufacturer of high-end fishing boats, brought claims alleging trademark infringement by its competitor, Barker Boatworks, which was founded by a former Yellowfin employee. Yellowfin Yachts, Inc. , 898 F.3d at 1287–88. The District Court in that case found that the record contained "no evidence" that Barker Boatworks copied Yellowfin's design "in an attempt to confuse a potential buyer" and granted summary judgment to Barker. Yellowfin Yachts, Inc. v. Barker Boatworks, LLC , 237 F. Supp. 3d 1230, 1238 (M.D. Fla. 2017). Indeed, the testimony Yellowfin relied on to show Barker's intent in the District Court actually demonstrated that Barker's design was "completely different" from Yellowfin's, not a copy. Id. at 1239.
On Yellowfin's appeal, we affirmed the District Court's grant of summary judgment and noted that Yellowfin "put forth no evidence showing Barker's intent to copy Yellowfin's [design] in order to deceive consumers as to the source of Barker Boatworks’ boats." Yellowfin Yachts, Inc. , 898 F.3d at 1294 (emphasis added). Yellowfin's evidence of Barker's "intent" to copy consisted primarily of a set of design notes that only briefly mentioned Yellowfin's boats and of prior "significant business dealings" between Barker and Yellowfin. Id. at 1293. This evidence, taken together, was not enough to create a genuine issue of material fact on Barker's intent to confuse: "Viewing the evidence in Yellowfin's favor allows us, at most, to infer that Barker Boatworks intended to copy some aspects of Yellowfin's boats in order to construct a worthy competitor in a niche market." Id. at 1293–94.
Here, however, J-B Weld has offered substantial evidence that Gorilla Glue intended to "mirror," "copy," and "knock off" J-B Weld Original's trade dress, not simply "construct a worthy competitor." And this is not a case in which consumer confusion depends on "sophisticated consumers" distinguishing between the "nuanced refinements" of the sheer lines of two boats that, presumably, consumers do not often view side-by-side. Id. at 1296. Rather, J-B Weld proffered evidence that Gorilla Glue intended to copy J-B Weld's trade dress in order "successfully mirror JB Weld at shelf "—in other words, GorillaWeld was designed with the knowledge that it would be sold on shelves near its competitor in retail stores. (emphasis added). This evidence of Gorilla Glue's intent to copy creates an inference that Gorilla Glue intended to capitalize on J-B Weld's goodwill, see AmBrit , 812 F.2d at 1543, and that evidence is probative of the likelihood of confusion issue. See id. ; Brooks Shoe Mfg. Co. v. Suave Shoe Corp. , 716 F.2d 854, 859 n.13 (11th Cir. 1983). The District Court, however, entirely ignored the possibility that a jury could infer Gorilla Glue's intent to benefit from J-B Weld's goodwill based on Gorilla Glue's intent to "copy aspects of J-B Weld's Dress."
Further, the District Court improperly inferred innocuous motives from the testimony of Gorilla Glue's employees, who stated that the "V-shape" design of GorillaWeld's packaging served purposes other than mirroring J-B Weld's trade dress. From the employees’ testimony, the Court inferred that Gorilla Glue sought only to compete with J-B Weld, not to confuse consumers. However, as the District Court correctly noted, in instances where " ‘there may have been many other motivations for Defendant's actions,’ intentional copying does not necessarily indicate a desire to capitalize on another's goodwill" (emphasis added). The issue, of course, is that the District Court concluded that testimony suggesting innocuous motives must show that Gorilla Glue did not intend to capitalize on J-B Weld's business reputation. But the fact that the intentional copying does not necessarily demonstrate Gorilla Glue's intent to capitalize on J-B Weld's goodwill does not mean that it cannot . The District Court was required to draw this inference in the light most favorable to J-B Weld, not Gorilla Glue. See McCullum , 768 F.3d at 1141.
An example is useful to understand the weight we place on self-serving testimony of "intent." In Jellibeans, Inc. v. Skating Clubs of Ga., Inc. , we affirmed a District Court's conclusion that a roller-skating rink, Lollipops, "intended to capitalize on" the reputation of its competitor, Jellibeans, in spite of the fact that Lollipops’ owners testified that "they had selected the name Lollipops with no intent of benefitting from its similarity to Jellibeans." 716 F.2d 833, 843 (11th Cir. 1983). We reasoned that sufficient circumstantial evidence of intent to confuse can overcome testimony professing innocent intentions, and we noted that even where intent is "innocent," the use of a mark may nevertheless create "a likelihood of unreasonable confusion in the minds of the public." Id. at 843 n.23. This makes intuitive sense, as "self-serving testimony" of intent is often no more than argument masquerading as testimony, and it adds little to the intent inquiry. Cf. Tartell v. S. Fla. Sinus & Allergy Ctr., Inc. , 790 F.3d 1253, 1258 (11th Cir. 2015) (stating that "self-serving testimony" "add[s] no material evidence ... that is relevant to the secondary meaning inquiry" (alterations in original and quotations omitted)). Accordingly, in the case at hand, it was entirely possible that a reasonable jury would have inferred Gorilla Glue's intent to capitalize on J-B Weld's goodwill from the Gorilla Glue employees’ testimony and from circumstantial evidence of intent.
On summary judgment, the District Court was required to view all of the evidence of Gorilla Glue's "intent"—including the testimony of Gorilla Glue's employees—in the light most favorable to J-B Weld, not Gorilla Glue. See McCullum , 768 F.3d at 1141. It simply failed to do so here, and by deciding disputed evidentiary issues on summary judgment, the District Court short-circuited the jury's role as fact finder. Accordingly, the District Court erred in evaluating evidence of Gorilla Glue's intent.
iii.
Third, we address the District Court's conclusion that J-B Weld's evidence was insufficient to establish any proof of actual confusion. This conclusion overlooks reasonable inferences of fact that operate in J-B Weld's favor and was therefore error.
Documented confusion of "someone very familiar with the enterprise," like a "professional buyer," is "relevant evidence" of actual confusion. Frehling Enters., Inc. v. Int'l Select Grp., Inc. , 192 F.3d 1330, 1341 (11th Cir. 1999). J-B Weld's evidence indicated that other industry professionals, including a buyer at a retailer that carries J-B Weld Original, asked J-B Weld representatives whether J-B Weld "had anything to do with" GorillaWeld, or if J-B Weld was making or supplying Gorilla Glue with "private Label Epoxy Twin Tubes" pursuant to some sort of agreement. The District Court stated that this evidence was not probative of actual confusion because the comments show that the representatives "knew the difference" between the two products. Because these professionals knew that the two were different products, the Court reasoned, their statements evince no confusion.
The District Court's conclusion is plausible, but it is not the only reasonable inference based on the evidence presented. It is also reasonable to infer that, even if the industry professionals knew the two products were different, they were confused as to whether GorillaWeld was the product of a collaboration or other liaison between the two companies. For example, a question as to whether J-B Weld was supplying Gorilla Glue with unbranded, "private label" packaging leads to a reasonable inference that the person was confused because he or she believed that the similarities in packaging signified a business relationship or other agreement between the companies. Asking if one company "had anything to do with" another company's product would — or, at least, could — generate a similar inference. Given these two possible interpretations, the District Court erred by favoring one conclusion over the other on summary judgment because that conclusion was adverse to J-B Weld, the non-moving party.
This is important because actionable confusion under the Lanham Act is not limited to confusion as to the literal source of a product or service — it also extends to confusion about perceived "affiliation" or "sponsorship" of a product or service by another company. See Ky. Fried Chicken Corp. v. Diversified Packaging Corp. , 549 F.2d 368, 388 (5th Cir. 1977).
Additionally, the District Court failed to consider any of the circumstantial factors that we have held are integral to determining how much weight should be assigned to any individual instance of actual confusion. We have held that "[i]n reviewing the evidence, there are no set rules as to how much evidence of confusion is needed; rather, a district court ‘must take into consideration the circumstances surrounding each particular case.’ " Lone Star Steakhouse & Saloon, Inc. v. Longhorn Steaks, Inc. , 122 F.3d 1379, 1382 (11th Cir. 1997) (quoting Dieter v. B & H Indus. of Sw. Fla. , 880 F.2d 322, 326 n.3 (11th Cir. 1989) ). Relevant circumstances include the extent of the parties’ advertising, the length of time for which the allegedly infringing product has been advertised, or any other factor that might influence the likelihood that actual confusion would be reported. Jellibeans, Inc. v. Skating Clubs of Ga., Inc. , 716 F.2d 833, 844 (11th Cir. 1983). The District Court did not consider these circumstances, which would have provided additional context for J-B Weld's evidence. For example, the record indicates that the GorillaWeld product was introduced in 2017, and J-B Weld's first Complaint was filed October 6, 2017. Given that the elapsed time since the introduction of the allegedly infringing product had been less than a year when the complaint was filed, in all likelihood the number of reported instances of actual confusion would be on the lower side, making each instance of reported confusion more probative. The District Court did not consider these circumstances relevant, and instead, supplying its own interpretation of the meaning of J-B Weld's evidence, it found that as a matter of law nobody was truly confused. This was error because it did not weigh the circumstances in the light most favorable to J-B Weld.
iv.
Finally, we turn to the District Court's failure to consider all of the appropriate factors in the likelihood of confusion analysis. Although a "district court's failure to consider all the factors relevant to the issue of whether two marks are confusingly similar does not necessarily constitute reversible error," Dippin’ Dots , 369 F.3d at 1207–08 (quoting Wesco Mfg., Inc. v. Tropical Attractions of Palm Beach, Inc. , 833 F.2d 1484, 1489 (11th Cir. 1987) ), here we agree with J-B Weld that the District Court failed to appropriately discuss four of the seven applicable factors — the similarity of the products, the similarity of retail outlets and purchasers, the similarity of advertising media used, and the strength of the J-B Weld mark — and that this failure was error.
We have said that all seven of the factors are potentially relevant in a likelihood of confusion analysis, and a district court must "fully consider the seven factors." Longhorn Steaks , 122 F.3d at 1382. The court must "evaluate the weight to be accorded the individual factors," which "varies with the circumstances of the case," and then "make its ultimate decision" only after doing so. AmBrit , 812 F.2d at 1538. Here, the District Court did not perform the analysis properly as just described — instead, it discussed the facts supporting, and weight due, only the three factors mentioned supra .
As to the strength of the trade dress factor, the Court noted in a footnote that it had "already determined [at the preliminary injunction stage] that J-B Weld's Dress deserves moderate protection" due to the nature and strength of its trade dress. It had previously found that J-B Weld Original's trade dress is "overall suggestive [on the distinctiveness scale,] and no third party uses the same combination of elements." First, the District Court's reliance on its factual findings at the preliminary injunction stage was improper at the summary judgment posture because "findings of fact and conclusions of law made by a court granting a preliminary injunction are not binding at trial on the merits." Univ. of Tex. v. Camenisch , 451 U.S. 390, 395, 101 S. Ct. 1830, 1834, 68 L.Ed.2d 175 (1981). Second, particularly because the strength of the trade dress is quintessential in "determin[ing] the scope of protection [the trade dress] will receive," the District Court's failure to consider the weight that a jury might assign this factor was error. AmBrit , 812 F.2d at 1539. The District Court appears to have unilaterally decided that this factor deserves little weight by restricting discussion to a footnote, despite J-B Weld's presentation of evidence that J-B Weld Original's dress is recognizable and has retained consistent features for decades. The other three factors — the similarity of the products, the similarity of retail outlets and purchasers, the similarity of advertising media used — received no discussion at all. The District Court did not discuss the facts underpinning any of these factors, or how much weight a jury could properly afford each factor in assessing a likelihood of confusion. Instead, it simply found that no matter how probative the remaining factors were, they would "still fail to demonstrate a likelihood of confusion." In doing so, the Court simply rejected evidence that weighed in J-B Weld's favor, because the record indicates that the other three factors (similarity of products, similarity of retail outlets and purchasers, and similarity of advertising media used) cut in favor of infringement. The two products are both heavy-duty adhesives, both sold in similar major hardware and automotive retailers, mass merchants, and online sites; and both are advertised via print media, customer cooperative ads, and trade shows.
The District Court's failure to discuss potentially relevant evidence by choosing not to discuss four factors, when combined with its impermissible fact-finding with respect to the other three factors, indicates that the Court did not "fully consider" the applicable factors. Longhorn Steaks , 122 F.3d at 1382. Accordingly, we must reverse the Court's holding because J-B Weld has shown enough evidence to make likelihood of confusion a triable issue of fact.
* * *
In sum, the analysis performed here was improper for a summary judgment posture, and for all of the above reasons, we reverse the District Court's order granting summary judgment to Gorilla Glue on J-B Weld's trade dress infringement claims, as presented in Counts I, III and IV of the amended complaint.
b.
We next turn to J-B Weld's argument that the District Court applied an improper standard in evaluating J-B Weld's trade dress dilution claim under Georgia law, as presented in Count V of its amended complaint. Georgia law provides that a party may enjoin the use of "the same or any similar trademark" if there is a likelihood of "injury to [the plaintiff's] business reputation" or "dilution of the distinctive quality of the [plaintiff's] trademark," even if there is no competition between the parties or no "confusion as to the source of goods or services." O.C.G.A. § 10-1-451(b).
J-B Weld argues that the District Court erred by requiring proof of a likelihood of confusion as an element of this claim, despite the statutory provision's explicit indication that confusion need not be proven. J-B Weld correctly notes that while the statute requires that the trade dress be "the same" or "similar" and requires a "likelihood of injury to business reputation or of dilution of the distinctive quality" of the plaintiff's mark, it does not require proof of likelihood of confusion. O.C.G.A. § 10-1-451(b) ; see also Original Appalachian Artworks, Inc. v. Topps Chewing Gum, Inc. , 642 F. Supp. 1031, 1039 (N.D. Ga. 1986) ("In order to prevail under [a Georgia state law] dilution claim, the plaintiff needs to show that the marks in question are similar ... and that the contested use is likely to injure [plaintiff's] commercial reputation or dilute the distinctive quality of its marks.") (internal quotations omitted); Pillsbury Co. v. Milky Way Prods., Inc. , 215 U.S.P.Q. 124, at *14 (N.D. Ga. 1981) (stating that proof of likelihood of confusion is not required under the Georgia anti-dilution statute).
The District Court's abbreviated treatment of this claim leaves us with serious doubt that it applied the correct standard in concluding that J-B Weld was unable to show trade dress dilution. The District Court's remarks about the indistinguishability of the applicable standards indicates that it applied the elements of the trade dress infringement claims to the trade dress dilution claim, thus conflating the two different sets of requirements. Specifically, the District Court indicated that the "federal Lanham Act analysis" governs all the parties’ trade dress claims, including "trade dress dilution in violation of O.C.G.A. § 10-1-451(b)." On this basis, it reasoned that "Gorilla Glue's likelihood of confusion arguments" properly "address[ed] J-B Weld's state trade dress dilution claim." The Court did not discuss this claim any further, instead ostensibly subsuming it into its analysis of the trade dress infringement claims presented by Counts I, III, and IV – an analysis which turned on J-B Weld's lack of evidence of likelihood of confusion. See subpart III.a, supra .
In light of the manner in which the District Court treated this claim, we cannot be sure that it did not erroneously require proof of likelihood of confusion to satisfy the requirements of O.C.G.A. § 10-1-451(b). Therefore, we must reverse the District Court's grant of summary judgment to Gorilla Glue on J-B Weld's Count V claim, and remand with the instruction that the District Court fully evaluate this claim in accordance with the standard set out in O.C.G.A. § 10-1-451(b).
c.
Finally, we address J-B Weld's challenge to the District Court's grant of summary judgment to Gorilla Glue on the claim that GorillaWeld is falsely advertised as a "steel bond epoxy." J-B Weld asserts that this advertising practice violates both the Lanham Act (Count II) and O.C.G.A. § 10-1-421 (Count VI). We affirm the District Court's holding that J-B Weld has not shown that the inclusion of "steel bond epoxy" on GorillaWeld's packaging is material to consumers, and therefore that GorillaWeld's advertising violates neither the Lanham Act nor Georgia state law.
To make out a claim of false advertising under the Lanham Act or O.C.G.A. § 10-1-421, a plaintiff must show five things: (1) a false or misleading advertisement (2) that deceived, or had the capacity to deceive, consumers; (3) the deception had a material effect on purchasing decisions; (4) the misrepresented product or service affects interstate commerce; and (5) the plaintiff has been, or is likely to be, injured by the false advertising. Johnson & Johnson Vision Care, Inc. v. 1-800 Contacts, Inc. , 299 F.3d 1242, 1247 (11th Cir. 2002) ; B&F Sys., Inc. v. LeBlanc , 2011 WL 4103576, at *21 n.13 (M.D. Ga. 2011) (indicating that the same elements apply to claims under O.C.G.A § 10-1-421 ). Here, the District Court based its disposition solely on whether any alleged misrepresentation was material to consumer purchasing decisions.
J-B Weld contends that "steel bond epoxy" is material to purchasers because it misrepresents "inherent quality[ies] or characteristic[s]" of GorillaWeld that are "each material to purchasers in selecting a product." We disagree. J-B Weld has not presented evidence indicating that either "steel bond" or "epoxy" is material to purchasing decisions. A representation in advertising is material when it is a "deception" that is "likely to influence the purchasing decision." Johnson & Johnson , 299 F.3d at 1250 (quoting Cashmere & Camel Hair Mfrs. Inst. v. Saks Fifth Ave. , 284 F.3d 302, 311 (1st Cir. 2002) ). On the other hand, even demonstrably false statements that are irrelevant to consumer purchasing decisions are immaterial and cannot be the basis for a false advertising claim. Id. ; see also id. at 1252 (finding no false advertising claim where plaintiff did not show that defendant's use of the term "eye doctor" instead of "eye care practitioner" had "any effect on consumer behavior," regardless of whether the language was false or misleading).
Because "steel bond epoxy" is a compound term containing two distinct descriptors of the GorillaWeld product, we address the materiality of "epoxy" and "steel bond" in turn, consistent with the flow of the District Court's analysis and the parties’ briefing.
i.
J-B Weld first argues that the term "epoxy" is material because it refers to the chemical composition of an adhesive, which constitutes an "inherent quality or characteristic" of the product. But the "inherent quality or characteristic" formulation adopted by this Circuit does not replace the consumer-oriented nature of the materiality inquiry with a scientific one. See Johnson & Johnson , 299 F.3d at 1250 ("[T]he plaintiff must establish that the defendant's deception is likely to influence the purchasing decision." (emphasis added) (internal quotation omitted)); see also Nat'l Basketball Ass'n v. Motorola, Inc. , 105 F.3d 841, 855 (2d Cir. 1997) (explaining that the requirement that a defendant misrepresent an inherent quality or characteristic of the product "is essentially one of materiality," and holding that because the defendant's inaccurate statements "would not influence consumers at the present time," the statements are "not material and do not misrepresent an inherent quality or characteristic of the product").
J-B Weld appears to be arguing that, since "components or ingredients" of a product are often found to be "inherent qualities or characteristics" that are important to consumer purchasing decisions, it follows that the mere fact that an advertisement concerns a component or ingredient of a product weighs in favor of it being material to buyers. J-B Weld's logic proves too much – not every component or ingredient of a product need be material to consumers, and J-B Weld cannot introduce the possibility of a correlation to circumvent its burden to show materiality.
J-B Weld argues that advertising GorillaWeld as an "epoxy" is material to consumers because "consumer[s] know[ ] that ‘epoxy’ is a specific and desirable category of adhesives." J-B Weld's argument commands the inference that a consumer would consider J-B Weld Original to be an "epoxy" adhesive, but would not consider GorillaWeld's MMA-based adhesive to constitute an "epoxy" adhesive, due to the chemical differences between the two formulas. J-B Weld has not presented any evidence that consumers are so scrupulous about the chemicals in their adhesives. In fact, the evidence presented indicates that consumers likely categorize "epoxies" as all two-part resin-and-hardener adhesives, regardless of the chemical constitution of the resin. John Barrett, J-B Weld's survey expert, admitted that J-B Weld's survey did not ask consumers whether or not they understood epoxy adhesive to have "a specific type of chemistry to it," and Barrett opined that consumers likely only care about whether the product sticks two surfaces together effectively. In addition, Gorilla Glue points to evidence showing that MMA-chemistry based adhesives, such as GorillaWeld, are frequently marketed, and categorized by retailers, as epoxies.
It seems to us that consumers categorize epoxy adhesives not by their chemical constitution, but by their resin-and-hardener method of application and by the strength of their bond on various surfaces. J-B Weld has not presented evidence giving any indication otherwise.
J-B Weld also claims that the District Court erred by "limit[ing] its analysis to whether the false advertising claim was material to .... end-user consumers." J-B Weld is correct that retail purchasers (i.e. "middlemen") can be considered in assessing materiality. See William H. Morris Co. v. Group W, Inc. , 66 F.3d 255, 257 (9th Cir. 1995) (holding that false statements made by a distributor of a dietary supplement were not material because they were unlikely to influence the purchasing decisions of retail pharmacists who were not end-user consumers). But J-B Weld's speculative assertion that retailer demand for GorillaWeld increased in 2017 merely because GorillaWeld began including "epoxy" on its labeling – thereby deceiving retailers – cannot be credited without proof that it was the inclusion of "epoxy," and not some other factor, that increased demand for the product. J-B Weld's estimate that in prior years the non-epoxy-labeled product "should have been [just] as successful" does not constitute proof. In addition, J-B Weld would have had to prove that the chemical constitution of GorillaWeld, rather than the resin-hardener application method, mattered to retail purchasers, which it was unable to do with respect to end-user consumers.
J-B Weld nonetheless contends that chemical epoxies, like J-B Weld Original, and MMA chemistries, like GorillaWeld, have "different physical properties," including "safety and odor differences." Maybe so. But J-B Weld has not made any showing that these differences would matter to a consumer. Without evidence on the issue, it would be pure speculation to hold that, due to some difference in the function or efficacy of the product, consumers are materially deceived when they purchase an adhesive advertised as an "epoxy" and receive an MMA-chemistry product. This speculation cannot sustain J-B Weld's claim of materiality.
ii.
We turn now to J-B Weld's claim that use of the phrase "steel bond" on GorillaWeld was deceptive and material to consumers. We are skeptical of Gorilla Glue's claim that the phrase "steel bond" is intended to describe "a strong bond that works well on metal," rather than an adhesive that physically contains iron or steel as a reinforcing agent. But it doesn't matter what we think, in the abstract, about the slogan's true meaning, because J-B Weld needed to show that the presence or absence of steel in GorillaWeld resin would be material to a consumer's purchasing decision. J-B Weld has not made that showing. J-B Weld's survey showed respondents the trade dress of both J-B Weld Original and GorillaWeld and asked them to identify which of the products they believed contained steel. However, the survey did not ask respondents whether their conclusions about the presence of steel would have affected their decision to purchase one product or the other. Without asking that question or something similar, the survey fails to address the critical issue of effect on purchasing decisions, and therefore cannot be probative of materiality. See Johnson & Johnson , 299 F.3d at 1250.
J-B Weld attempts to supplement its claims of materiality by invoking internal Gorilla Glue documents, all of which tend to show that Gorilla Glue wanted the GorillaWeld trade dress to make some reference to "steel." For example, Gorilla Glue's marketing team indicated its intention to "bring in some visual reference to steel reinforcement" and "play up on steel." These statements cannot overcome the fact that J-B Weld has not shown that a consumer would be more likely to purchase an adhesive that he or she believed contained steel. Without that, J-B Weld has not shown materiality, no matter how purposeful Gorilla Glue's inclusion of "steel" on the packaging was. * * *
In sum, we agree with the District Court's conclusion that "J-B Weld has not shown that consumers care whether GorillaWeld uses MMA chemistry or epoxy chemistry, or whether GorillaWeld actually contains steel." Therefore, the Court's grant of summary judgment for Gorilla Glue on J-B Weld's claims of false advertising under the Lanham Act and O.C.G.A. § 10-1-421, as presented in Counts II and VI of J-B Weld's amended complaint, was proper.
IV.
For the reasons stated above, the District Court's grant of summary judgment to Gorilla Glue is affirmed with respect to J-B Weld's claims of false advertising presented in Counts II and VI of its amended complaint. We reverse with respect to J-B Weld's claims of trade dress infringement presented in Counts I, III and IV and remand those claims for trial. And we reverse with respect to J-B Weld's trade dress dilution claim presented in Count V and remand that claim for further summary judgment proceedings consistent with this opinion.
AFFIRMED IN PART, REVERSED IN PART AND REMANDED.
ED CARNES, Circuit Judge, concurring:
I concur in the majority opinion in this case but write separately to emphasize the distinction between "intentional copying" and "intentional copying with intent to cause confusion ." Yellowfin Yachts, Inc. v. Barker Boatworks, LLC , 898 F.3d 1279, 1293 (11th Cir. 2018). That distinction matters, and I do not read the majority opinion to suggest that it doesn't.
In the trade dress context, a defendant's intent to copy aspects of a plaintiff's product will not always support a finding of likelihood of confusion; a defendant's intent to copy for the purpose of confusing consumers will. See, e.g. , id. ("[P]roof of intentional copying alone is not conclusive on the likelihood of confusion issue."); Brooks Shoe Mfg. Co. v. Suave Shoe Corp. , 716 F.2d 854, 859 n.13 (11th Cir. 1983) ("[T]here is a difference between intentional copying and adopting a mark or design with the intent of deriving benefit from another person's mark." (quotation marks omitted)); J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 23:110 (5th ed. 2017 & Sept. 2020 update) ("[T]he only kind of intent that is relevant to the issue of likelihood of confusion is the intent to confuse."); id . § 8:19 (evidence that "a junior user copies a competitor's product design because it sells better and consumers seem to like it ... is not evidence of an intent to confuse").
That distinction — copying to compete versus copying to confuse — comports with the intellectual property principle that not all copying is culpable. See B.H. Bunn Co. v. AAA Replacement Parts Co. , 451 F.2d 1254, 1259 (5th Cir. 1971) ("Outright copying is often a civilizing rather than a cannibalizing folkway. The world would be a duller place without the originators, but it would not work without the copyists."); McCarthy on Trademarks § 1:24 ("The popular folklore is that a ‘copycat’ is a pirate and that all commercial copying and imitation must be an illegal form of competition. That is false. Legitimate copying is what makes a free market economy work.").
The majority opinion discusses our Yellowfin Yachts decision and notes the distinction between "intentional copying and intentional copying with intent to cause confusion." Op. at 790, 791–92. It concludes that evidence of Gorilla Glue's intent to copy J-B Weld's packaging supports a finding of likelihood of confusion because a jury could reasonably infer from that evidence that Gorilla Glue intended to confuse consumers about GorillaWeld's origins and benefit from the brand identity of J-B Weld's trade dress. Op. at 792 ("This evidence of Gorilla Glue's intent to copy creates an inference that Gorilla Glue intended to capitalize on J-B Weld's goodwill, and that evidence is probative of the likelihood of confusion issue." (citation omitted)).
I agree that our precedent permits inferring intent to confuse from evidence of intent to copy if the facts of the case allow that inference. See, e.g. , AmBrit, Inc. v. Kraft, Inc. , 812 F.2d 1531, 1543 (11th Cir. 1986) ("Although Kraft was free to copy the Klondike product and the functional packaging features of that product, the finder of fact may infer from evidence of such actions an intent to derive benefit from Isaly's goodwill."). And I agree that a jury could reasonably draw that inference based on the record in this case: as the majority notes, Gorilla Glue sought to "mirror JB Weld at shelf ," Op. 792; its design team described GorillaWeld's trade dress as "[c]lose to the JB Weld brand," Op. 790–91; and a Gorilla Glue employee even called GorillaWeld a "knock off" of J-B Weld, Op. 790–91, 791–92. Based on that evidence a jury could reasonably infer that Gorilla Glue intended to copy J-B Weld's trade dress to confuse consumers about GorillaWeld's origins and benefit from J-B Weld's goodwill. The district court, which had to draw all inferences in J-B Weld's favor, erred in concluding otherwise. See McCullum v. Orlando Reg'l Healthcare Sys., Inc. , 768 F.3d 1135, 1141 (11th Cir. 2014).
For those reasons, I do not read the majority opinion to collapse the distinction between copying to compete and copying to confuse, nor to alter the well-established rule that intentional copying does not — without more — permit an inference of copying with intent to confuse. Instead, it holds that there is evidence of "more" in this case, and the district court erred by failing to consider that evidence in the light most favorable to J-B Weld.