Select Comfort Corporation v. John Baxter

996 F.3d 925

• Court: Court of Appeals for the Eighth Circuit
• Decided: May 11, 2021
• Docket: 19-1077
• Status: Published

Opinion

United States Court of Appeals For the Eighth Circuit ___________________________

No. 19-1077 ___________________________

Select Comfort Corporation; Select Comfort SC Corporation

lllllllllllllllllllllPlaintiffs - Appellants

v.

John Baxter; Dires, LLC, doing business as Personal Touch Beds and Personal Comfort Beds; Digi Craft Agency, LLC; Direct Commerce, LLC, doing business as Personal Touch Beds; Scott Stenzel; Craig Miller

lllllllllllllllllllllDefendants - Appellees ___________________________

No. 19-1113 ___________________________

lllllllllllllllllllllPlaintiffs - Appellees

John Baxter

lllllllllllllllllllllDefendant

Dires, LLC, doing business as Personal Touch Beds and Personal Comfort Beds

lllllllllllllllllllllDefendant - Appellant Digi Craft Agency, LLC; Direct Commerce, LLC, doing business as Personal Touch Beds

lllllllllllllllllllllDefendants

Scott Stenzel; Craig Miller

lllllllllllllllllllllDefendants - Appellants ___________________________

No. 19-1178 ___________________________

lllllllllllllllllllllDefendant - Appellant

Dires, LLC, doing business as Personal Touch Beds and Personal Comfort Beds; Digi Craft Agency, LLC; Direct Commerce, LLC, doing business as Personal Touch Beds; Scott Stenzel; Craig Miller

lllllllllllllllllllllDefendants ____________

Appeals from United States District Court for the District of Minnesota ____________

Submitted: May 13, 2020 Filed: May 11, 2021 ____________

Before SMITH, Chief Judge, MELLOY and SHEPHERD, Circuit Judges. MELLOY, Circuit Judge.

Plaintiffs and Defendants sell competing adjustable air mattresses and related products. Plaintiffs’ registered trademarks include “SLEEP NUMBER”, “WHAT’S YOUR SLEEP NUMBER”, “SELECT COMFORT”, and “COMFORTAIRE”. Plaintiffs allege Defendants used similar and identical marks in several different capacities online to sell competing products. Plaintiffs also allege Defendants compounded internet-related confusion by making fraudulent misrepresentations and failing to dispel confusion when consumers contacted Defendants’ call centers. At summary judgment the district court rejected as a matter of law an infringement theory based on presale or initial-interest confusion. 4 J. McCarthy, Trademarks and Unfair Competition, § 23:6 (4th Ed. 2010) (hereinafter McCarthy) (initial-interest confusion is “confusion that creates initial customer interest, even though no actual sale is finally completed as a result of the confusion”). The case proceeded to a trial on trademark infringement and dilution claims and on unfair competition and false advertising claims. Consistent with the summary judgment ruling, the district court instructed the jury that infringement liability depended on a showing of a likelihood of confusion at the time of purchase. The trial resulted in a mixed verdict.

Both sides appeal. Plaintiffs raise nine issues. Defendants contest all nine issues, raise one additional issue in a cross-appeal, and raise a second issue in a “conditional cross-appeal.” Because we conclude the district court erred by finding as a matter of law that the relevant consumers were sophisticated and that a theory of initial-interest confusion could not apply, we reverse. As a result, certain issues fall away. Several additional issues remain, however, and we address them below.

I. Background

Plaintiffs are the owners of the heavily advertised Select Comfort and Sleep Number brands of adjustable air mattresses sold online, over the phone, and

-3- (primarily) through hundreds of company-owned stores nationwide. Defendant Dires, LLC, and its principals and predecessor or affiliated companies, actually made adjustable air beds at an earlier date. Defendants have evolved into an online retailer (“personalcomfortbed.com”) that utilizes internet advertising and a call-center-based sales model to sell their own brand of lower-priced adjustable air beds. The individual defendants are executives or owners of Dires or related companies, all of whom had input into marketing strategy and advertising design. Defendants are a distant second to Plaintiffs in adjustable-bed sales volume.

Plaintiffs and Defendants have somewhat of a shared history in that Defendant- Appellee Craig Miller had at one point (from 2006 to 2011) worked as a consultant for Plaintiffs. Also, Defendant-Appellees Baxter and Stenzel had worked for Plaintiff Comfortaire, another air bed company, and Plaintiff Select Comfort SC Corporation later purchased Comfortaire. In fact, Stenzel and Baxter developed their advertising and sales techniques when working for Comfortaire.

Plaintiffs’ overall theory of the case alleges Defendants employed words or phrases identical or confusingly similar to Plaintiffs trademarks in various online advertising formats including: website urls; search inquiry paid terms; embedded links in third-party sites; and general use of identical or similar phrases in text advertisements or combined graphic-and-text advertisements that could be viewed by users or detected organically by search engines. According to Plaintiffs, Defendants used these means to divert customers to their own website and phone lines where Defendants (1) failed to dispel consumer confusion or made statements that caused further source confusion and (2) made false representations about their own products and Plaintiffs’ products in order to promote their own products. In this way, Plaintiffs assert trademark infringement, trademark dilution, and false-advertising theories that rely upon common facts.

-4- As relevant to claims on appeal, Plaintiffs asserted federal trademark infringement and dilution claims based on their registered trademarks, federal unfair competition and false advertising claims, and a state law deceptive trade practices claim. In a declaratory judgment counterclaim, Defendants argued Plaintiffs had no trademark rights in the unregistered phrase “NUMBER BED” because the phrase was either generic or merely descriptive but lacking secondary meaning. Defendants sought a summary judgment ruling on their “NUMBER BED” counterclaim and also argued the incontestable, registered trademark “SLEEP NUMBER” had become generic and was no longer protectable. Defendants also argued Plaintiffs could not maintain an infringement claim based on presale or initial-interest confusion. Both parties sought summary judgment.

In summary judgment rulings, the district court found outstanding questions of fact as to whether “SLEEP NUMBER” had become generic and as to whether “NUMBER BED” had acquired secondary meaning and gained status as a protectable trademark. Regarding trademark infringement, the district court found generally that outstanding questions of fact precluded summary judgment. Regarding the specific question of trademark infringement in the form of initial-interest confusion, the district court first noted that Plaintiffs expressly disavowed any theory of trademark infringement that relied exclusively on Defendants’ use of Plaintiffs’ trademarks as paid search terms with search engine providers such as Google. Rather, Plaintiffs alleged infringement based on that use coupled with Defendants’ several and varied other uses of similar and identical trademarks in multiple forms of online advertising. The district court then relied on our case, Sensient Techs. Corp. v. SensoryEffects Flavor Co., 613 F.3d 754 (8th Cir. 2010), noting that the Eighth Circuit had neither expressly adopted nor rejected a theory of initial-interest confusion as a general matter, but had refused to apply the theory in a case where consumers were sophisticated.

-5- The District Court next held as a matter of law that retail purchasers of mattresses were sophisticated consumers because mattresses are expensive. As a result, the District Court held as a matter of law that a claim alleging initial-interest confusion could not proceed and Plaintiffs would have to show a likelihood of confusion at the time of purchase. The district court ruled as to several other issues, notably denying Plaintiffs’ motion for summary judgment regarding a false advertising claim in which Plaintiffs argued that Defendants’ Rule 30(b)(6) witness admitted the literal falsity of certain statements. The district court held that the statements, in context, were equivocal, presented a factual question, and did not support summary judgment.

At trial, Plaintiffs presented evidence which showed Defendants had used Plaintiffs’ actual trademarks as paid search terms and as identical phrases in their own web-based advertising in text pages, combined text and graphical pages, as terms embedded in linked internet address urls, and in other fashions. Examples included website links that presented Plaintiffs’ trademarks as identical phrases (e.g. personalcomfortbed.com/vSleepNumber or www.personalcomfortbed.com/ cComfortaire). In addition, Defendants used phrases similar to Plaintiffs’ trademarks, often with words broken up in a grammatically non-sensical fashion. Examples included the use of terms such as “Sleep 55% Off Number Beds” and “Comfort Air Beds on Sale” in online advertisements. Survey evidence demonstrated actual consumer confusion, although the parties disputed the relevancy and value of the survey evidence based on percentages of participants who were confused, whether the survey participants were actual or potential consumers, and how the questions were presented. Evidence also included instances of actual confusion, often from transcripts of call-center interactions, messages from customers, or messages from call-center employees. The transcripts and recordings of call-center interactions appeared to show that Defendants’ call-center employees at times attempted to promote confusion and at other times attempted to dispel confusion. Finally, evidence included statements from Defendants’ principals in which they described

-6- confusion as between Plaintiffs’ and Defendants’ brands as a “good thing” and, in response to reports of confusion, indicated that their advertisements were “working.”

At the end of the day, the district court submitted the case to the jury. Based on the summary judgment ruling, the district court instructed the jury that a likelihood of confusion must exist at the time of purchase to support a trademark infringement claim. The court also submitted fifteen alleged false advertising claims to the jury. Over Defendants’ objection, the district court instructed the jury as to false advertising using an instruction that allowed a presumption as to the element of materiality if the jury concluded a statement was literally false.

The jury rejected the trademark infringement claims as to the registered trademarks based on the jury instruction that limited the possibility of a likelihood of confusion to the time of purchase. In addition, the jury found that Plaintiffs held no trademark rights in the phrase “NUMBER BED” and that Defendants’ use of the phrase “NUMBER BED” was not unfair competition. Regarding dilution, the jury found the mark “SLEEP NUMBER” famous, indicating it was a strong and well- known mark, but the jury also found Defendants had not diluted the mark. Finally, the jury found for Plaintiffs on seven of the false advertising claims and for Defendants on the remaining eight. The jury awarded a total of approximately $160,000 in damages on the seven false advertising claims based on a wrongful benefit received by Defendants. The jury awarded no damages on the false advertising claims based on Plaintiffs’ alleged lost profits. The district court denied several post-trial motions, and the parties appeal.

II. Discussion

A. Initial-Interest Confusion

The primary issue in this appeal is the availability of a theory of initial-interest confusion on the trademark infringement claim and the resulting limitation in the

-7- instruction requiring any likelihood of confusion to exist at the time of purchase. As noted, initial-interest confusion is “confusion that creates initial customer interest, even though no actual sale is finally completed as a result of the confusion.” 4 McCarthy §23:6. Most circuits that have addressed the question “recognize the initial interest confusion theory as a form of likelihood of confusion which can trigger a finding of infringement.” Id. (collecting cases). In general, the theory of initial- interest confusion recognizes that a senior user’s goodwill holds value at all times, not merely at the moment of purchase. The theory protects against the threat of a competitor “receiving a ‘free ride on the goodwill’ of [an] established mark.” Checkpoint Systems, Inc. v. Check Point Software Technologies, Inc., 269 F.3d 270, 295 (3d Cir. 2001) (quoting Mobil Oil Corp. v. Pegasus Petroleum Corp., 818 F.2d 254, 260 (2d Cir. 1987)). This free ride may result in the consumer falsely inferring an affiliation between the junior and senior users, provide the junior user with an opportunity it otherwise would not have achieved, or deprive the senior user of an actual opportunity. Id. at 293–95. At least one circuit has “equated initial interest confusion to a ‘bait and switch scheme.’” Id. at 294 (quoting Dorr-Oliver, Inc. v. Fluid Quip, Inc., 94 F.3d 376, 382 (7th Cir. 1996)).

In the present case, the parties dispute as a general matter whether a theory of initial-interest confusion is a viable theory of infringement in our circuit. They also dispute whether the relevant consumers—consumers investigating mattresses and online shoppers in general—are so sophisticated that the issue of consumer sophistication could properly be removed from the jury. To address these questions, it is necessary first to review more generally the test for confusion and what our Court has said about when confusion must exist.

The Lanham Act provides several forms of protection for commercial goodwill. Trademarks are protected against infringement, that is, the use of similar marks on similar or related products or services if such use creates a likelihood of confusion. 15 U.S.C. § 1125(a)(1)(A); see also 15 U.S.C. § 1114(1)(a). If a trademark is deemed

-8- “famous” it may be protected against dilution in the form of “tarnishing” or “blurring.” Id. § 1125(c)(2)(C) (the “tarnishing” of a famous mark is a general damaging of the goodwill associated with a mark due to a “similarity between a mark or trade name and a famous mark that harms the reputation of the famous mark”); id. § 1125(c)(2)(B) (the “blurring” of a famous mark is the “impair[ment] [of] the distinctiveness of the famous mark” “arising from the similarity between a mark or trade name and a famous mark”). The prohibition of dilution “protects the holder of a famous trademark from misappropriation of its investment in the mark” regardless of confusion. Everest Capital Ltd. v. Everest Funds Mgmt., 393 F.3d 755, 762 (8th Cir. 2005). Finally, the Lanham Act also protects more generally against false advertising, false representations, and unfair competition. 15 U.S.C. § 1125(a)(1)(B).1

To assess the likelihood of confusion as required for a showing of infringement, our circuit employs a list of nonexclusive factors for addressing a core inquiry: whether the relevant average consumers for a product or service are likely to be confused as to the source of a product or service or as to an affiliation between sources based on a defendant’s use. See Anheuser–Busch, Inc. v. Balducci Publ’ns, 28 F.3d 769, 774 (8th Cir.1994) (noting that protection extends “‘against use of [plaintiff’s] mark on any product or service which would reasonably be thought by the buying public to come from the same source, or thought to be affiliated with, connected with, or sponsored by, the trademark owner’” (quoting McCarthy § 24.03 (3d. 1992) (alteration in original))). The factors we consider come from SquirtCo v. Seven–Up Co., 628 F.2d 1086, 1091 (8th Cir. 1980), and include: (1) the strength of the owner’s mark; (2) the similarity of the owner’s mark and the alleged infringer’s

1 The parties do not meaningfully challenge the judgment below as to the dilution claims and we do not vacate that portion of the judgment in which the jury rejected the dilution claim but held Plaintiffs’ “SLEEP NUMBER” mark to be famous (strong, well-known, and heavily advertised).

-9- mark; (3) the degree to which the products compete with each other; (4) the alleged infringer’s intent to “pass off” its goods as those of the trademark owner; (5) incidents of actual confusion; and (6) the type of product, its costs and conditions of purchase.

We have repeatedly emphasized that no one factor is controlling and different factors will carry more weight in different settings. SquirtCo, 628 F.2d at 1091 (“[R]esolution of this issue does not hinge on a single factor but requires a consideration of numerous factors to determine whether under all the circumstances there is a likelihood of confusion.”); Lovely Skin, Inc. v. Ishtar Skin Care Prods., LLC, 745 F.3d 877, 887 (8th Cir. 2014) (“[T]he relative weight of the factors depends on the facts of the individual case.” (quoting First Nat. Bank in Sioux Falls v. First Nat. Bank, South Dakota, 153 F.3d 885, 888 (8th Cir. 1998))); Frosty Treats, Inc. v. Sony Comput. Ent. Am., Inc., 426 F.3d 1001, 1008 (8th Cir. 2005) (“factors do not operate in a mathematically precise formula” (citation omitted)). We have also noted that the factors are not truly independent—depending on the context, a strong showing as to one factor may serve to make a different factor more or less important. See Kemp v. Bumble Bee Seafoods, Inc., 398 F.3d 1049, 1054 (8th Cir. 2005) (“[T]he factors are not entirely separable. For example, it is inappropriate to conduct a side-by-side comparison of the elements of two products’ trade dress . . . without reference to the senior mark’s strength or the market conditions under which likely consumers would see the marks.”); see also ConAgra, Inc. v. George A. Hormel & Co., 990 F.2d 368, 371 (8th Cir. 1993) (“[W]hen ‘products are closely related, less similarity in trademarks is necessary to support a finding of infringement.’” (quoting SquirtCo., 628 F.2d at 1091)). Ultimately, application of the factors is a highly fact- intensive inquiry both as to the assessment of the evidence concerning each factor and as to the overall synthesis of factors and the evidence.

This flexible, context-specific, and relative-rather-than-mechanical approach makes sense because the general function of the likelihood-of-confusion factors is to guide the finder of fact towards considerations generally thought to be material to the

-10- consuming public’s understanding of product source or affiliation. Common sense is inherent in the factors, and the factors, properly applied, should try to capture a holistic view of the normal experiences for any given industry, product, or service. The consumer experience differs by products (buying a toothbrush vs. buying a car vs. professional buyers obtaining input goods for a factory), and the relative importance of any given factor is influenced greatly by how the other factors might apply. As a result, we review the likelihood of confusion determination as a finding of fact. See Everest Capital, 393 F.3d at 760.2

Although our test for a likelihood of confusion is well-developed, some uncertainty remains as to when confusion must exist in order to support a trademark infringement claim. Sensient, 613 F.3d at 766. Although not addressing initial- interest confusion specifically, our Court has clearly established that claims of

2 On appeal, the parties frame the primary issues in dramatically different ways. Plaintiffs argue the district court erred in a summary judgment ruling and carried that error forward to the jury instructions, effectively limiting the theories presented to the jury. Plaintiffs characterize jury findings relating to these issues as tainted by underlying legal error and not meriting deference. Defendants argue Plaintiffs’ appeal is better characterized as an instructional issue reviewed merely for abuse of discretion. Defendants also rely heavily on those portions of the jury verdict in their favor to argue that our review should be limited to the sufficiency of the evidence and that certain trial results are effectively shielded by the jury’s interpretation of the evidence. Essentially, the parties dispute the manner in which we must view the evidence when conducting our review. Because we ultimately conclude a summary judgment ruling and subsequent instructions erroneously limited the theories presented to the jury, our review of the evidence speaks largely to the question of harmless error or prejudice and not to the question of sufficiency. As such, we discuss the evidence generally in terms of what the parties presented to the jury rather than limiting our discussion to what the jury found. As a practical matter, the ability to determine the inferences the jury drew from the evidence is substantially clouded by (1) the interrelated nature of the infringement, dilution, and misrepresentation claims in this case, (2) the mixed verdict, and (3) our conclusion that summary judgment and instructional error occurred.

-11- infringement are not limited solely to a likelihood of confusion at the time of purchase. See Insty*Bit, Inc. v. Poly-Tech Indus., 95 F.3d 663, 671–72 (8th Cir. 1996). In Insty*Bit, our Court recognized that a 1962 amendment to the Lanham Act eliminated reference to “purchasers” when describing actionable confusion. Id. (quoting Pub. L. No. 87–772, 76 Stat. 769, 773 (1962)). We interpreted this statutory amendment as permitting claims for post-sale confusion among non- purchasers—generally “consumers”— who witnessed a confusingly marked product. Id.3

Fourteen years later, however, our Court indicated that it was unclear as a general matter whether initial-interest or presale confusion was actionable. See Sensient Tech., 613 F.3d at 766. There, over a dissent, our circuit identified the theory, but neither rejected nor adopted it for general application. Instead, we held that the theory did not apply on the facts of the case because the consumers at issue

3 Our Court stated:

“Post-sale confusion” refers to the association consumers might make between the allegedly infringing item and the familiar product, thereby influencing their purchasing decisions. The Lanham Act’s protection of post-sale confusion stems from the 1962 amendment to § 32 of the Act, 15 U.S.C. § 1114(1), which provides remedies for the infringement of [a] registered trademark. Pub. L. No. 87–772, 76 Stat. 769, 773 (1962). The 1962 amendment included confusion of nonpurchasers as well as direct purchasers by eliminating language in § 32 which had restricted the scope of trademark infringement to confusion of “purchasers as to the source of origin of such goods or services.” 76 Stat. at 773. Thus, an action for trademark infringement may be based on confusion of consumers other than direct purchasers, including observers of an allegedly infringing product in use by a direct purchaser.

Insty*Bit, 95 F.3d at 671–72.

-12- were sophisticated commercial purchasers of inputs for industrial food production who purchased goods with a high degree of care “after a collaborative process.” Id. at 769.

The general question of whether presale, initial-interest confusion is actionable, therefore, seemingly pits two opposing views of trademark law against one another. On the one hand, through our application and review of the likelihood of confusion factors, we recognize the varied landscape of commercial transactions and leave the jury to sort through the details. Our factors provide guidance but do not draw bright lines that might constrain the general test for confusion. Similarly, the Court in Insty*Bit refused to place firm constraints on the question of when confusion must exist. On the other hand, in Sensient, our Court acknowledged the possibility of cabining the likelihood-of-confusion test to a particular moment in time, at least under certain circumstances.

We now address the issue left open in Sensient and hold that a theory of initial- interest confusion may apply in our circuit.4 We are, of course, bound by Sensient.5 But, when the particular conditions of Sensient are not present, i.e., when a jury question exists as to the issue of consumer sophistication, a plaintiff should not be barred from proving presale, initial-interest confusion. In reaching this conclusion we find the Lanham Act itself and amendments to its language as cited in Insty*Bit particularly compelling. Other courts addressing the question of initial-interest

4 In so holding, we make no comment as to the impact that such a showing might have on the availability of various remedies or any measurement of damages—entirely separate questions dependent on the proof in a given case. 5 Because Sensient did not address post-sale confusion, it is not inconsistent with Insty*Bit in a way that would preclude it from having precedential value. See Mader v. United States, 654 F.3d 794, 800 (8th Cir. 2011) (en banc) (“when faced with conflicting panel opinions, the earliest opinion must be followed as it should have controlled the subsequent panels that created the conflict.” (cleaned up)).

-13- confusion have relied on this language. Checkpoint, 269 F.3d at 295 (noting that as originally enacted, “the Lanham Act only applied where the use of similar marks was ‘likely to cause confusion or mistake or to deceive purchasers as to the source of origin of such goods or services’” (quoting 1946 Lanham Act) (accord Esercizio v. Roberts, 944 F.2d 1235, 1244 (6th Cir. 1991))); see generally, 4 McCarthy §23:7 (collecting cases) (noting that several courts have interpreted this amendment as expanding trademark protection beyond point-of-sale confusion to reach presale confusion (including initial-interest confusion) and post-sale confusion).

And, in general, adoption of the theory is consistent with the overall practice of recognizing the varied nature of commercial interactions and the importance of not cabining the jury’s analysis of the likelihood of confusion factors. If we do not generally impose strict constraints on the jury’s nuanced assessment of how or whether the consuming public might be confused, it would be odd to presume that all commercial interactions are alike or that, in all settings, trademarks are worthy of protection only in the few moments before the consummation of a transaction.

Of course, as per Sensient, the theory of initial-interest confusion cannot apply in our Circuit where the relevant average consumers are sophisticated at the level of the careful professional purchasers who were at issue in Sensient. In this regard, however, we find several comments by the dissent in Sensient compelling, and we note that a finding of customer sophistication typically will rest with the jury.

In reaching its conclusion, the Court in Sensient relied upon Checkpoint Systems for the proposition that “courts look to factors such as product relatedness and the level of care exercised by customers to determine whether initial interest confusion exists.” Sensient, 613 F.3d at 766. Sensient was an appeal from a grant of summary judgment, and on the summary judgment record, our Court indicated that the parties agreed the relevant consumers were sophisticated. The dissent in Sensient accurately noted, however, that in Checkpoint Systems, the Third Circuit had been

-14- reviewing the issue after trial, not making a determination as to consumer sophistication as a matter of law (or making any likelihood of confusion determinations) at the summary judgment stage. Id. at 773 (Colloton, J, dissenting). The dissent described the theory of initial-interest confusion and emphasized that, even if customers are sophisticated, that fact alone should not automatically defeat the theory. In advocating for this no-blanket-rule point, the dissent cited a Second Circuit case involving professional buyers in a lawsuit between Mobil Oil and an entity that was marketing products under the name “Pegasus Petroleum.” The dissent noted that “[w]hether or not a sophisticated customer eventually would sort out the difference, the doctrine of initial interest confusion prevents an infringer from using another’s mark to gain ‘crucial credibility during the initial phases of a deal.’” Id. at 773 (quoting Mobil Oil Corp. v. Pegasus Petroleum Corp., 818 F.2d 254, 258 (2d Cir. 1987)). And, the dissent also emphasized that the Third Circuit in Checkpoint specifically disclaimed any categorical rule, stating instead that the “significance [of customer sophistication] will vary, and must be determined on a case-by-case basis.” Sensient, 613 F.3d at 773 (quoting Checkpoint, 269 F.3d at 297).

Regardless of the relative merits of the positions reflected in Sensient, our general adoption of the theory of initial-interest confusion forecloses summary judgment where a question of fact exists as to the level of consumer sophistication. Here, the parties dispute the issue of consumer sophistication both in reference to shopping for mattresses and shopping online. They also dispute whether consumer sophistication should be measured at the “point of click” for an online shopper, at the point of sale upon final purchase, or at points in between. For the reasons previously discussed, we do not believe it is appropriate to cabin the analysis to any one point in time. And, in any event, authority is mixed as to whether mattress shoppers and online shoppers should be deemed careful, sophisticated consumers.

On the one hand, mattresses are relatively expensive among most consumers’ purchases. See Sleepmaster Prods. Co. v. Am. Auto-Felt Corp., 241 F.2d 738, 741

-15- (C.C.P.A. 1957) (“[T]he average purchaser will exercise such care in the selection of a mattress as to minimize the possibility of confusion as to the origin of the goods.”). On the other hand, most people buy mattresses infrequently, so they enter the marketplace uneducated and susceptible to fast-talking sales people and brand confusion. See Friedman v. Sealy, Inc., 274 F.2d 255, 261–62 (10th Cir. 1959) (“[S]ince a mattress or box spring requires an investment . . ., the degree of care which a customer might be expected to exercise is somewhat greater than if he were buying 5-cent candies. [But] the construction of sleep equipment is not a matter of common knowledge and the consumer buys infrequently. He is thus forced to rely on his memory, more than his inspection, for the recall of names, guarantees, and endorsements. Under such circumstances, confusion can easily arise.”).

Authority is also mixed as to the level of sophistication web-based shoppers bring to the table and how this potentially separate question should influence the general assessment of sophistication. Compare Coca-Cola Co. v. Purdy, No. 02-1782 ADM/JGL, 2005 WL 212797, at *4 (D. Minn. Jan. 28, 2005) (“[T]he quick and effortless nature of ‘surfing’ the Internet makes it unlikely that consumers can avoid confusion through the exercise of due care.”) and GoTo.com, Inc. v. Walt Disney Co., 202 F.3d 1199, 1209 (9th Cir. 2000) (“Navigating amongst web sites involves practically no effort whatsoever, and arguments that Web users exercise a great deal of care before clicking on hyperlinks are unconvincing.”) with Toyota Motor Sales, U.S.A., Inc. v. Tabari, 610 F.3d 1171, 1179 (9th Cir. 2010) (“[I]n the age of [the internet], reasonable, prudent and experienced internet consumers are accustomed to such exploration by trial and error. They skip from site to site, ready to hit the back button whenever they’re not satisfied with a site’s contents. They fully expect to find some sites that aren’t what they imagine based on a glance at the domain name or search engine summary.”). See also Network Automation, Inc. v. Advanced Systems Concepts, Inc., 638 F.3d 1137, 1152 (9th Cir. 2011) (noting that although “‘there is generally a low degree of care exercised by Internet consumers’ . . . the degree of care analysis cannot begin and end at the marketing channel. We still must consider the

-16- nature and cost of the goods, and whether ‘the products being sold are marketed primarily to expert buyers.’” (quoting Brookfield Comm’ns, Inc. v. West Coast Ent. Corp., 174 F.3d 1036, 1060 (9th Cir. 1999))).

At the end of the day, this mix of authority regarding consumer confusion in the context of internet shopping and mattress purchases demonstrates well why a jury rather than a judge should assess the level of consumer sophistication. This point is particularly strong in a case which, like the present case, enjoys a full record including highly detailed descriptions of Plaintiffs’ and Defendants’ customers’ experience and ample evidence of (1) actual confusion including transcripts of potential customers who called Defendants’ call centers and believed they were calling Plaintiffs, and (2) statements by Defendants’ principals describing the actual confusion as evidence that their own advertising was working. See Kemp, 398 F.3d at 1058 (evidence of actual confusion, while not required, is strong evidence of a likelihood of confusion); SquirtCo, 628 F.3d at 1091 (“Likewise, actual confusion is not essential to a finding of trademark infringement, although it is positive proof of likelihood of confusion.”).

Against this backdrop, we conclude a jury question existed as to the issue of consumer sophistication and summary judgment on the theory of initial-interest confusion was error. For the same reasons, and based on Insty*Bit, we conclude that limiting the infringement instruction to require confusion at the time of purchase was error. Finally, given the strength of the Plaintiffs’ evidence on the issue of confusion, we cannot conclude that the summary judgment and instructional errors were harmless. See Fed. R. Civ. P. 61. In so ruling, we make no comment as to how a finding of confusion at times other than the moment of purchase might affect the analysis of remedies and the determination of damages.

-17- B. False Advertising.

Plaintiffs moved for judgment as a matter of law on the false advertising claims as to certain statements for which Plaintiffs argued Defendant Craig Miller admitted literal falsity. The district court denied the motion. On appeal, Plaintiffs renew their arguments. In addition, in the conditional cross appeal, Defendants acting as conditional cross appellants challenge the formulation of the jury instructions. Defendants argue the district court improperly shifted the burden of proof by applying an erroneous presumption as to the elements of the false advertising claims.

On de novo review, we find no error in the district court’s denial of Plaintiffs’ motion for judgment as a matter of law and submission of the issue of falsity to the jury. Pittari v. Am. Eagle Airlines, Inc., 468 F.3d 1056, 1061 (8th Cir. 2006) (standard of review). The test for literal falsity is “rigorous,” statements must be analyzed in their broader context, and there are any number of reasons why the jury might have chosen to discount Miller’s testimony. Buetow v. A.L.S. Enters., Inc., 650 F.3d 1178, 1185 (8th Cir. 2011) (reversing a summary ruling as to a finding of literal falsity). We agree with the district court’s prudent choice to place the factually complex question in the hands of the jury.

Regarding the conditional cross appeal, Defendants ask the court to reverse and remand based on the false advertising jury instructions if the court “remands the case for any reason in Select Comfort’s nine-issue appeal.” Because we reverse and remand as to the infringement claim, we address the conditional cross appeal. See Murphy v. FedEx Nat’l LTL, Inc., 618 F.3d 893, 901 (8th Cir. 2010) (addressing conditional cross appeal due to remand on other issues).

“We review a district court’s formulation of jury instructions for an abuse of discretion and its interpretation of law de novo.” United States v. Farah, 899 F.3d

-18- 608, 614 (8th Cir. 2018). The instructions for the false advertising claim identified the elements as: (1) a false statement about Defendants, Defendants’ products, Plaintiffs, or Plaintiffs’ products in an advertisement; (2) such statement deceived or tended to deceive a substantial portion of its audience; (3) the statement was material in that it was likely to influence a purchasing decision; and (4) Plaintiffs were or are likely to be injured as a result. The instructions also provided, over Defendants’ objection, that the jury could presume materiality (element 3) if the Defendants made: (1) a literally false statement; (2) a false statement relating to the inherent quality or characteristic of a product; or (3) a deliberately false or misleading statement that was comparative or implicated a competitor or its product.

In their opening brief as cross-appellants, Defendants do not take issue with the second or third “triggers” for the presumption. Rather, Defendants challenge the presumption of materiality based on the first trigger: a finding that a statement was literally false. By limiting their challenge, Defendants appear to recognize that the second two triggers are essentially definitions for materiality that describe types of statements reasonable persons would recognize as likely to influence a purchasing decision. In fact, other circuits have reached this conclusion in the context of false advertising claims. See, e.g., Cashmere & Camel Hair Mfrs. Inst. v. Saks Fifth Ave., 284 F.3d 302, 312 n.10 (1st Cir. 2002) (labeling a statement as material because it was related to an inherent quality or characteristic of the product); Nat’l Basketball Ass’n v. Motorola, Inc., 105 F.3d 841, 855 (2d Cir. 1997) (stating that a requirement that a misrepresentation address an “inherent quality or characteristic of the product . . . is essentially one of materiality, a term explicitly used in other circuits.” (cleaned up)).

A finding that a statement is literally false, in contrast, does not appear to suggest in any direct manner that the statement is material. A literally false statement could address any number of topics. As such, a finding of literal falsity, standing alone, does not necessarily make a statement more or less likely to influence a

-19- purchasing decision. Of course, depending on the nature of the falsehood and the topic it addresses, a jury might conclude a false statement is material. But the reasoning leading to such a conclusion depends on additional facts beyond mere falsity. In any event, an inference of a statement’s materiality based merely upon its falsity is neither so clear nor direct that it might support a burden-shifting presumption in a plaintiff’s favor.

In defense of the instruction, Plaintiffs (as cross-appellees) point to our case in Porous Media Corp. v. Pall Corp., 110 F.3d 1329 (8th Cir. 1997). In Porous Media, we identified the elements of a claim for false advertising as: (1) a misrepresentation as to “the nature, characteristics or qualities” of a defendant’s products “alone or in comparison with” a plaintiff’s products; (2) actual deception or a tendency to deceive “a substantial segment of their audience”; (3) materiality, meaning the misrepresentations were “likely to influence buying decisions”; (4) an interstate commerce nexus; and (5) injury in the form of a “direct diversion of sales” or a “lessening of [plaintiff’s] goodwill.” Id. at 1332. We then approved a rebuttable presumption of deception (element 2) and damages (element 5) upon a showing that: (1) such a misrepresentation about products was made deliberately (with knowledge of its falsity); and (2) the statement was made “as an important part of its marketing efforts.” Id. at 1332–33.

Porous Media, standing alone, does not support the proposition that a jury may presume materiality. Rather, Porous Media involved presumptions under different circumstances and as to different elements. The court in Porous Media was not asked to address, and did not address, a question regarding a presumption of materiality. District courts in our circuit have examined Porous Media to assess the availability of presumptions as to different elements of a false-advertising claim and concluded Porous Media does not support an inference of materiality. See 3M Innovative Props. Co. v. Dupont Dow Elastomers, LLC, 361 F. Supp. 2d 958, 971 (D. Minn. 2005) (“[T]he Eighth Circuit’s opinion in Porous Media, which upheld the use in a false advertising case of some presumptions upon proof of intentional conduct, referred

-20- only to the deception element and not to materiality.”); see also, Aviva Sports, Inc. v. Fingerhut Direct Mrkt’g, Inc., 829 F. Supp. 2d 802, 813 (D. Minn. 2011) (“Even where literal falsity is established, materiality is not presumed.”). At least two circuit courts have reached the same conclusion. See Johnson & Johnson Vision Care, Inc. v. 1–800 Contacts, Inc., 299 F.3d 1242, 1250 (11th Cir. 2002) (rejecting a presumption of materiality based on a finding of literal falsity and stating, “The materiality requirement is based on the premise that not all deceptions affect consumer decisions.”); Cashmere & Camel Hair Mfrs. Inst. v. Saks Fifth Ave., 284 F.3d 302, 312 n. 10 (1st Cir. 2002) (“[M]ateriality focuses on whether the false or misleading statement is likely to make a difference to purchasers. Thus, even when a statement is literally false or has been made with the intent to deceive, materiality must be demonstrated in order to show that the misrepresentation had some influence on consumers.” (citing McCarthy § 27:35 (4th ed. 2001))).

At least one circuit arguably disagrees and appears to permit an inference of materiality. See, e.g., Pizza Hut, Inc. v. Papa John's Int’l, Inc., 227 F.3d 489, 497 (5th Cir. 2000) (citing Am. Council of Cert’d Podiatric Phys. & Surgeons v. Am. Bd. of Podiatric Surgery, Inc., 185 F.3d 606, 614 (6th Cir.1999) and Avila v. Rubin, 84 F.3d 222, 227 (7th Cir.1996)). The Fifth Circuit’s discussion of materiality, however, is not clear and does not provide an explanation as to why a finding of literal falsity implies materiality. Further, the cases from other circuits as cited in Pizza Hut do not clearly address a presumption of materiality. Rather, they appear to focus on the element of consumer deception. In fact, we agree with the Eleventh Circuit which noted that the Fifth Circuit appeared to “conflate[] the element of consumer deception with the element of materiality” thus “blurr[ing] the boundary between the two elements.” Johnson & Johnson, 299 F.3d at 1250–51 (criticizing the Fifth Circuit’s approach in Pizza Hut, 227 F.3d at 497).

Ultimately, we conclude it was error to instruct the jury in a manner that shifted the burden of proof on the materiality element based on a finding of literal falsity.

-21- And, based on the specific jury forms returned in this case, we do not find the error to be harmless as to those claims where Plaintiffs prevailed.6 Accordingly, we reverse and remand for a new trial on the seven false advertising claims on which Plaintiffs prevailed.

C. Other Issues

Several additional issues merit mention. First, after the close of discovery and after the summary judgment rulings, Defendants withdrew their counterclaim alleging that “SLEEP NUMBER” had become generic and had been abandoned. Defendants then moved to amend their pleadings to add a counterclaim alleging “SLEEP NUMBER” was void because it was generic ab initio, not merely that it had become generic over time as urged at summary judgment. The district court denied the motion to amend noting: the issue of whether a mark was void ab initio depended on evidence predating or at the time that Plaintiffs began using the mark; the parties had not made such evidence a focus of discovery; Defendants had not provided adequate notice that such a theory was being pursued in their counterclaim; discovery had specifically focused on a different period of time; and discovery had closed.

Defendants challenge this ruling on appeal as an abuse of discretion. Plaintiffs disagree and also argue an ab initio genericness defense is not available to attack a registered, incontestible mark. We need not address the legal question of which defenses the Lanham Act permits as against incontestible marks. Instead, we find no abuse of the district court’s substantial discretion in refusing to permit amendment of the counterclaim after the close of discovery and on the eve of trial. See Kozlov v. Ass’d Wholesale Grocers, Inc., 818 F.3d 380, 394–95 (8th Cir. 2016). Here, the

6 The only appeal taken on the eight claims the jury found for Defendants is on the issue of the directed verdict. As discussed above, we reject that appeal, and therefore the judgment as to those eight claims remains unaffected by this opinion.

-22- proposed amendment would have required additional discovery even though the moving party had participated in discovery, without objection, in a manner that did not address the issue. See Thompson-El v. Jones, 876 F.2d 66, 67–68 (8th Cir. 1989).

Second, during trial, Plaintiffs objected to Defendants’ expert’s testimony when the expert’s testimony regarding survey evidence strayed from commentary as to the survey results and instead purported to describe for the jury what is meant by “the essence of confusion.” According to Plaintiffs, this testimony amounted to impermissible expert testimony describing a legal standard. Because we are reversing as to the infringement claims, we need not address this question in depth. We merely note that an expert’s testimony as to the structure and meaning of survey evidence or other factual matters generally should not usurp the court’s role in defining the law for the jury. See Peterson v. City of Plymouth, 60 F.3d 469, 475 (8th Cir. 1995) (holding that it was an abuse of discretion to allow testimony that “was not a fact-based opinion, but a statement of legal conclusion” because the “legal conclusions were for the court to make”). Of course, slight deviations from this general rule are unlikely to result in harmful error if identified and explained for the jury, but a factual expert should not opine as to meaning of a legal standard. Fed. R. Civ. P. 61.

Third, Plaintiffs also objected to the Defendants use of an exhibit at trial—an adjustable air bed that was not from the production years at issue in the case. The district court allowed the exhibit for the purpose of educating the jury generally as to the parties’ products and the nature of adjustable air beds. We find no abuse of the district court’s substantial discretion in admission of the demonstration bed. Bradshaw v. FEE Transp. Servs., 715 F.3d 1104, 1108–09 (8th Cir. 2013) (visual aids permissible). Any infirmities as to the demonstration go to the weight rather than the admissibility of the evidence.

-23- Fourth, regarding jury instructions, Plaintiffs argue the district court misapplied the burden of proof on Defendant-Appellees’ cross claim seeking a declaration that Plaintiffs held no trademark rights in “NUMBER BED”. In particular, Plaintiffs argue the district court erroneously placed the burden on Plaintiffs (as the cross-claim defendants) to prove that they possessed enforceable trademark rights in the phrase “NUMBER BED.” We reject Plaintiffs’ argument. Although a declaratory judgment plaintiff in most contexts bears the burden of proof, declaratory judgment plays a somewhat different role in trademark and patent disputes. The filing of a declaratory judgment action in a trademark or patent suit may meaningfully affect venue, but the burden to prove the existence of enforceable intellectual property rights stays with the alleged owner of the rights, whether that party is the declaratory judgment plaintiff or defendant. See Medtronic, Inc. v. Mirowski Family Ventures, LLC, 571 U.S. 191, 194 (2014) (holding that, in a patent infringement declaratory judgment action filed by an alleged infringer, the defendant patent holder bears the burden of proving infringement); see also, e.g., KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 117–18 (2004) (describing a trademark declaratory judgment defendant as the plaintiff in an analysis of burdens of proof concerning likelihood of confusion and fair use); Cross Commerce Media, Inc. v. Collective, Inc., 841 F.3d 155, 166 (2d Cir. 2016) (purported rights owner, though declaratory judgment defendant, bears the burden of proving protected rights in a mark).

Other issues the parties raise on appeal amount to alleged trial errors we need not address in light of our remand.

III. Conclusion

We reverse and vacate the judgment as to the infringement and false advertising claims. We leave undisturbed those portions of the judgment dealing with the dilution claims and the alleged “NUMBER BED” trademark. We otherwise remand for further proceedings consistent with this opinion. ______________________________

-24-