Congratulations to Mark McKenna, whose amicus brief made the key arguments endorsed by the court here. A jury awarded over $1.2 million on Groeneveld’s trade dress claims, but the majority found that Lubecore’s Rule 50 motion should’ve been granted. Starting from the premise that trademark law “is designed to promote brand recognition, not to insulate product manufacturers from lawful competition,” the court allowed a competitor to copy the shape of a grease pump because (1) the plaintiff didn’t show nonfunctionality, and (2) no reasonable consumer of these expensive and niche products could be confused based on the shape.
The court framed the issue as a battle between an established business and an upstart. The parties make grease pumps for automated lubrication systems (ALS) used in commercial trucks. An ALS delivers controlled amount of lubricant to different parts of a machine while the machine is in operation; its primary component is a grease pump.
Groeneveld has been making the grease pump at issue since the 1980s. Lubecore was founded in 2007 by Jan Eisses, who had previous business/employment relations with Groeneveld. It made a very similar-looking pump:
The jury found that Groeneveld proved by a preponderance of the evidence that its trade dress (“the external shape and appearance of the pump, including logo and color”) was nonfunctional and had secondary meaning, and that there was likely confusion as to the source of the pumps. It also found the infringement willful.
The majority held that the overall shape was nonfunctional. Groeneveld didn’t argue that individual parts were protectable. The pump was a black cast aluminum base containing the pump mechanism topped by a clear reservoir. The pump mechanism was connected by wires and hoses to the rest of the ALS. Both pump and reservoir clearly served essential functions. The trial testimony from Groeneveld witnesses made “clear that not only the basic manufacture of the grease pump’s components, but also their size and shape, are closely linked to the grease- pumping function.” The shape of the base minimized the amount of material needed in construction (functional). The volume of the reservoir was dictated by the amount of grease needed during each servicing interval (functional). The clear reservoir allowed users to easily see how much grease was left (functional). Because the surface area (determined by the need to fit on the base) and volume were functionally determined, so was the height. All the elements were there for a practical reason, with nothing ornamental.
“Because Groeneveld presented no evidence showing that the individual components of its grease pump or their overall configuration are nonfunctional, it failed to carry its burden of creating a triable issue of fact with respect to nonfunctionality.” Combining individual functional components in a nonarbitrary manner is functional.
The court rejected Groeneveld’s argument that this particular design wasn’t necessary for effective competition in the ALS business, as shown by the fact that no other competitors made a similar-looking pump. But that’s exactly the argument that the Supreme Court rejected in TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001). Competitive necessity is an appropriate inquiry in aesthetic functionality cases, but not in utilitarian functionality cases where a design is “essential to the use or purpose of a device” or affects its cost or quality. When the design is functional under those standards, there is no need for further consideration of competitive necessity. “In other words, the question is whether the overall shape of Groeneveld’s grease pump was substantially influenced by functional imperatives or preferences. We accordingly reject Groeneveld’s invitation to drift back into the error of inquiring about possible alternative designs.”
Groeneveld’s VP of design and production van der Hulst, who was involved in the pump design, testifed that Groeneveld did not “have to make its pump look this way on the inside because of the way it works on the outside.” But that was insufficient to create a triable issue of fact under TrafFix Devices because it improperly focused on the possibility of alternative designs. It was also conclusory: van der Hulst didn’t explain “why the chosen design was nonfunctional, and certainly did not speak with any particularity about the functional considerations that, as outlined above, apparently dictated the pump’s design.” The bare fact of putting a manager on the stand to claim nonfunctionality doesn’t create a triable issue of fact.
Groeneveld also relied on van der Hulst’s testimony that the “commercial people” affected design, but the majority thought that was unclear and unhelpful. If they were executives in charge of evaluating commercial viability, their involvement said nothing about functionality. And even if it meant marketing/design employees, “[e]very viable mass-market product is presumably designed with marketing considerations in mind, and this unremarkable fact says nothing about whether the product design is nonfunctional.” Van der Hulst also testified that other grease pumps on the market looked “terrible” and that Groeneveld’s founder had a good design sensibility; Groeneveld hadn’t switched to alternative designs, even though they might be cheaper, because the current pump is “a very nice pump” and “[e]verybody knows this pump.” This too didn’t address nonfunctionality: something might look cleanly designed and be made of functional parts. To the extent it was about alternative designs, it wasn’t relevant. This was insufficient, especially in light of testimony from Groeneveld’s own witnesses to show that “the pump’s volume, shape, and materials are all essentially influenced by the dictates of function.”
The majority noted the public policy behind functionality: “to channel the legal protection of useful designs from the realm of trademark to that of patent. Such channeling ensures that the high public costs of monopoly are not imposed without an assurance that the design satisfies the rigorous requirements of patentability, including novelty and nonobviousness, and is protected for only a limited period of time.” This isn’t trademark’s job.
The court nonetheless also addressed likely confusion. The factors in the multifactor test “are helpful guides rather than rigid requirements,” with likely confusion the ultimate question. The factors “are therefore not always weighed consistently in this court’s caselaw, and a particular factor might receive a greater or lesser weight depending on the circumstances.” Sometimes cases can be resolved with reference to only one or a few factors.
Here, the presence of house marks was essentially dispositive. The parties’ logos and trademarks appeared on the pumps, as well as on sales and marketing literature, and were “unmistakeably different.” “In light of such a stark visual difference in branding, no reasonable consumer would think that the two grease pumps belong to the same company.” This was reinforced by the fact that, at about $2500 apiece, ALS systems were “expensive industrial products that are not likely to be purchased without substantial care and research.” “[P]otential purchasers of grease pumps are knowledgeable and sophisticated people. Such purchasers are therefore not likely to ignore the stark difference in labeling and mistakenly purchase a Lubecore ALS when they intend to purchase a Groeneveld ALS.”
Groeneveld argued that “(1) labels are not the predominant brand identifiers in the industry; (2) corporate mergers and acquisitions are frequent, so competitor affiliations are constantly changing; (3) many pumps bear multiple company names; (4) a pump’s label does not identify who made it or where it was manufactured; (5) Lubecore is a newcomer and has no independent brand recognition; and (6) witnesses who saw the Lubecores were still confused, notwithstanding the Lubecore label.”
Nope. Mergers and acquisitions (2) and Lubecore’s newcomer status (5) had no bearing on the distinguishability of the pumps via the labels. The claim that labels weren’t the predominant brand identifiers was “irrelevant” and not supported by record evidence. “The most that the cited testimony shows is that certain witnesses were able to distinguish Groeneveld’s pumps from pumps other than Lubecore’s without even looking at the labels, which says nothing about whether a consumer would be able to distinguish a Groeneveld pump from a Lubecore pump with the labels.” And points (3) and (4) didn’t apply to the labels at issue in the present case. Only (6), actual confusion, was relevant—and wrong (see below). Groeneveld at most “identified possible reasons why—hypothetically—differential branding might not be sufficient to distinguish the sources of competing products,” but it didn’t show those reasons applied.
The “starkly different branding” plus the “high degree of care presumably exercised by the pumps’ sophisticated consumers” compelled the conclusion that there could be no likely confusion as a matter of law. The majority nonetheless ran through the rest of the factors.
Of note is its analysis of intent, the key point in Groeneveld’s analysis. Yes, the similar appearance was circumstantial evidence of intent to copy, but “Groeneveld is mistaken about the legal significance of such copying.” Its argument that intentional copying showed a bad intent and created a presumption of confusion “indicates a fundamental misapprehension of the purposes of trademark law.”
Trademark doesn’t bar copying as such—that’s for trademark and patent. If a manufacturer wants protection against copying, it should get one of those. No harm is done to trademark’s incentive structure (which “incentivizes manufacturers to create robust brand recognition by consistently offering good products and good services”) by the copying of a product design that doesn’t confuse consumers as to source. “As long as a consumer can easily identify the source based on the trademark, the consumer will still be able to allocate his or her capital freely and efficiently, and manufacturers will retain the incentive to improve their offerings and solidify their brands.” Thus, copying in the absence of confusion isn’t just legal; it’s often a positive good.
Given these baseline principles, the intent that is relevant is not the intent to copy but the intent to confuse. Otherwise, a legal and beneficial act “would be transformed into evidence—or worse, as Groeneveld suggests, a ‘presumption’—of unlawfulness.” This would “subvert the fundamental purposes of trademark law.” Intent to copy might be probative of secondary meaning, but “such intent standing alone has no bearing on the likelihood-of-confusion issue.”
Here, copying had salutary effects, as posited by the general rule: “the similarity serves the procompetitive purpose of signaling the existence of a competitive alternative by alerting potential consumers that the pumps might work the same because they look the same.” Lubecore’s use of similarity “targeted consumers who are familiar with the Groeneveld pump and offered them a competitive option.” And “using a functional product’s look to promote a competitive offering is a procompetitive practice.” It would be unfair if there were confusion, but Lubecore “scrupulously avoided such confusion by choosing a starkly different logo that it prominently displays on its pumps and on all its sales and marketing literature.” By targeting Groenveld customers, “Lubecore focuses its competitive activity on those who are most interested in such competition, thereby decreasing the consumers’ search costs and intensifying competition where it matters most.” Maybe Lubecore doesn’t make a better pump, but aggressively courting consumers and offering them choices is good, from a public policy viewpoint.
One of Groeneveld’s witnesses, Dean Osborn, proved the point by testifying that he likes the way the Groeneveld pump works but does not particularly care who manufactures it. He testified that he was satisfied with both and never confused. With the labels off, he wouldn’t be able to tell the difference, but to him, “that means I’m comfortable with the product. If it be whose product it is, I don’t care. I have a good experience with it.… I’m looking at the mechanism that functions the grease to go to the spots where the greasing, that’s all I’m -- that’s all I care about.” The fact that the pumps looked similar gave him confidence to choose either, since he knew how the Groeneveld worked.
Groeneveld argued that Lubecore’s pumps were worse, and that Lubecore offered to extend Groeneveld’s warranty and to replace Groeneveld pumps and parts with Lubecore products. But that was no aid to Groeneveld: “If Lubecore’s pumps are in fact inferior, all the better for Groeneveld: Consumers would soon realize the difference in quality on the clearly labeled pumps and flock to Groeneveld.”
The court then rejected an anti-copying ideology:
Groeneveld’s protestations against slavish copying admittedly have a certain emotional appeal and presumably swayed the jury. After all, people generally dislike copycats. But, as the foregoing discussion demonstrates, the proper application of trademark law requires us to focus our analysis not on the intent to copy the product design, but on the likelihood of consumer confusion. Evidence of Lubecore’s intent to copy Groeneveld’s product design is therefore of no help to Groeneveld. Such evidence, if anything, shows the procompetitive benefits of Lubecore’s practices and cautions against allowing the issue to go to the jury. See Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205, 214 (2000) (discussing the desirability of “summary disposition of an anticompetitive strike suit” in the trade-dress context).
Then the majority ran through the other factors. Groeneveld had evidence that its design was recognizable in the industry, which was sufficient to support a factual finding of a strong trade dress—but that was no help without evidence of likely source confusion. Similarity weighed heavily against Groeneveld, because of the “starkly different” labels and trademarks. And Groeneveld had no evidence of likely confusion; it offered the testimony of a single customer, who unequivocally testified that he wasn’t confused. Testimony that Groeneveld’s own employees and affiliates were shocked and surprised by the copying wasn’t consumer testimony, and anyway they weren’t confused.
Overall, the trade dresses (including logos) were dissimilar, the purchases were sophisticated, there was no evidence of actual confusion, and the “intent and effect” of Lubecore’s design copying was procompetitive. The factors that weighed in Groeneveld’s favor were the strength of the trade dress, the relatedness of the products, and the similarity of the marketing channels. But those three factors standing alone didn’t raise a triable issue of fact about likely confusion. No reasonable jury could find that Groeneveld met its burden.
Adding in initial interest confusion didn’t change anything. IIC might make sense “under the right circumstances,” and here the court offered a billboard analogy: you’re lured off the interstate by a sign for McDonald’s, but discover only a Burger King. “You are not so fond of Burger King but, having already made the detour and loath to waste even more time, you reluctantly buy a Whopper and get on with your trip.” Even without purchase point confusion, this harms “consumer interests, brand-development incentives, and efficient allocation of capital.”
But that wasn’t the case here. First, Groeneveld didn’t explain how there could be any initial interest confusion at all, given the different labels and logos. Nor did Groeneveld explain why any confusion wouldn’t dissipate without harm. “[A] hypothetical chance that a consumer might think for an instant that two products come from the same source is simply not enough.” Groeneveld wasn’t really upset about initial interest confusion, but rather initial interest. It’s natural enough to dislike competition, but not a trademark problem.
The court also refused to consider Groeneveld’s dilution theory of harm (that Lubecore harmed its brand image by introducing low-quality knockoffs) for the first time on appeal.
The majority also rejected false advertising claims based on Lubecore’s practice of offering to extend Groeneveld’s warranty and to replace Groeneveld pumps and parts with Lubecore products. Targeting the customers of a competitor by extending a warranty program or offering the replacement of parts wasn’t unlawful; that’s competition.
Judge Helene White dissented, arguing that the jury could reasonably have found differently than the majority did. The appropriate focus should have been the overall trade dress rather than each component. (I think the majority approach, carving out the house mark, is much better—it avoids the problem of the plaintiff claiming that the overall trade dress including the mark is nonfunctional, then claiming confusing similarity based only on similarities in functional elements.)
The majority overread TrafFix, which only held that the possibility of alternative designs can’t render an otherwise functional trade dress nonfunctional. (Seems to me that’s exactly what the majority says!) But that doesn’t mean the possibility of alternative designs is irrelevant, or relevant only in cases of aesthetic functionality.
The dissent read the testimony to say that the form wasn’t dictated by function. Anyway, a nonfunctional configuration of functional components could be protectable, as long as the features were configured in an “arbitrary, fanciful, or distinctive way.” (Note that this quote, from an earlier case, can’t fully be squared with Wal-Mart’s holding that there’s no such thing as inherent distinctiveness in product design.) There was evidence supporting a finding that the pump’s overall configuration was designed to look distinctive in the industry rather than due to functional concerns. The dissent would’ve found it sufficient fact that the outer appearance wasn’t dictated by internal functioning (even though that’s exactly the reasoning reversed in TrafFix). Here, “the round and cylindrical shape of the clear reservoir, the grooves on the top and bottom of the reservoir, the particular placement of the product label and other features, and the irregular shape of the base” had no inherently functional purpose, even if individual components had functional qualities. And Groeneveld’s witness van der Hulst testified that the pump could be made in different ways—horizontal or vertical pistons, round or square reservoirs—and work the same. (The quoted testimony also says the reservoir could be made in different sizes … because that affects the time between fillings. I see why the majority wasn’t impressed.)
Van der Hulst also explained that the pump was designed to look nice and “completely different” from other pumps made with lots of different “bolts and screws”—the Groeneveld pump was “[o]ne piece worked and finished” and thus looked much nicer than the alternatives. Where “the pump’s design was based on aesthetic considerations, the shape does not dictate the pump’s function, other designs would result in the same function, and the design does not result in superior performance or cost effectiveness,” the dissent would allow a finding of nonfunctionality.
Likewise, the dissent would’ve upheld the confusion finding. There was testimony that the pump was recognizable by its shape, and strong evidence of intent to copy. “When a newcomer to the market copies a competitor’s trade dress, its intent must be to benefit from the goodwill of the competitor’s customers by getting them to believe that the new product is either the same, or originates from the same source as the product whose trade dress was copied.” (Yes, and it’s that either/or that is the issue: as the majority says, it’s not the case that we condemn intent to confuse and are simply neutral about intent to communicate comparability. To the contrary, public policy supports copying, and thus communication of copying, product features, at least where confusion can be avoided. Because one of the two possibilities is a positive good, it would be foolish to find that intent to copy was intent to confuse.) The dissent would’ve used the copying as evidence of secondary meaning (though of course for the same reasons copying might just be evidence of a good product design that, if unprotected by patent, is free for all to try.)
The dissent, rather implausibly for these particular products, evoked the rule that likely confusion has to be judged not side by side, but singly, “to account for the possibility that sufficiently similar [trade dresses] may confuse consumers who do not have both [trade dresses] before them but who may have a general, vague, or even hazy, impression or recollection of the other party’s [trade dress].” And the pumps look similar despite the different labels.
High degree of care wasn’t enough as a matter of law. Adding a house mark isn’t dispositive, and doesn’t always need to be given significant weight. In the Maker’s Mark case, the court credited testimony that many consumers didn’t know about brand affiliations, and that some companies produce multiple types of distilled spirits. When two products are related enough, one might associate with the other and still use a house mark. “Confusingly similar marks may lead a purchaser who is extremely careful and knowledgeable . . . to assume nonetheless that the seller is affiliated with or identical to the other party.” Since the most important factors are similarity of marks and strength of the plaintiff’s mark, consumer care couldn’t overcome them.
The same reasoning was available to the jury here; the district court found that the different markings might not be enough “given the testimony regarding corporate mergers and acquisitions in the industry; the fact that many pumps bear multiple company names; and, the fact that labels may not be the brand identifiers relied on in this industry.” The dissent argued that “facts that undermine presumed consumer care in differentiating between products are relevant to the overall analysis, and dissimilarities in the trade dresses based on labeling have less weight in the context of such industry-specific evidence.” There was testimony that people in the industry didn’t rely on labels to distinguish pumps (though they didn’t need to until Lubecore entered the market—as the majority concludes didn’t and couldn’t are potentially very different things). But the evidence that consumers usually used the design to identify the brand “supports the inference that labels are not necessarily indicative of the product’s manufacturer as the consumer’s focus in this industry is on the design.” (As a matter of policy, given functionality doctrine, perhaps we should discourage consumers from this focus, since for functional designs it will lead them astray.)
Also, Lubecore was a new company with minimal brand recognition, which undercut the effect of the labeling: “even a sophisticated consumer cannot be expected to recognize brands that have little name recognition or have been in existence only a short time.” That short time on the market also undermined the weight the majority gave to the lack of evidence of actual confusion.
As to intent, the dissent agreed that intentional copying isn’t itself actionable. But then the dissent said that a presumption of intent to confuse arises from evidence of copying, unless confusion is unlikely based on other factors. So intent to copy wasn’t irrelevant. Lubecore’s founder testified that he liked the Groeneveld pumps because they had a good reputation and Lubecore didn’t want its pumps to look like those of a company with a bad reputation. This was “compelling circumstantial evidence” that Lubecore intentionally copied to benefit from Groeneveld’s established reputation. Also, Lubecore’s extended warranty specifically targeted Groeneveld customers and products, and not other competitors. This also supported an inference of intent to confuse, “as an unaffiliated company normally does not (and cannot) extend another company’s warranty.”
Anyway, intent never favors the defendant, as the majority seemed to hold. “Intent … is an issue whose resolution may benefit only the cause of a senior user, not of an alleged infringer.” The same was true of lack of evidence of actual confusion, which is rarely present.
Strength of the trade dress and relatedness of the goods, while discounted by the majority, are factors that bear on whether confusion is likely. (This is a core problem with the multifactor test treated as if it were uniformly applicable—fortunately here just by the dissent. Strength of a trade dress and relatedness of goods, along with similarity of marketing channels, may be “significant” in a likely confusion inquiry, but only if other factors are present. Coca-Cola’s trade dress in its bottle epitomizes trade dress strength. But that doesn’t mean that Pepsi’s bottle is likely to infringe!)
Because the dissent would’ve upheld the permanent injunction, it addressed Groeneveld’s appeal arguing that the injunction should’ve been broadened to include Canada. The dissent disagreed. Extraterritorial application of the Lanham Act requires a court to consider “(1) whether the defendant’s conduct has a substantial effect on commerce in the United States; (2) whether the defendant is a citizen of the United States; and (3) whether there exists a conflict between defendant’s trademark rights established under foreign law, and plaintiff’s trademark rights established under domestic law.”
Lubecore is a Canadian corporation, weighing heavily against extraterritorial application of the Lanham Act, though Lubecore had no claim to the trade dress in Canada so barring its use wouldn’t interfere with Canadian sovereignty. Groeneveld’s argument that Lubecore’s Canadian activities had a substantial effect on US commerce was speculative—it just claimed a likelihood that Canadian trucks with Lubecore pumps would cross in and out of the US, affecting sales and reputation in the US. “Even if (as Groeneveld asserts) U.S. consumers can access Lubecore’s web sites and order products online, that does not prove that Lubecore’s Canadian activities have a substantial effect on domestic commerce.” The possibility of a US impact wasn’t enough.