VERISIGN, INC., Plaintiff-Appellant, v. XYZ.COM LLC; Daniel Negari, Defendants-Appellees, 848 F.3d 292

PAMELA HARRIS, Circuit Judge: Plaintiff Verisign, Inc. is in the business of selling internet domain names and operates the popular .com and .net top-level domains. In 2014, a competitor arrived on the scene: Defendant XYZ.COM, LLC (“XYZ”) launched “.xyz,” a new top-level domain, and began registering domain names ending in .xyz. As part of its marketing push, XYZ, along with its CEO Daniel Negari, made a series of statements touting the popularity of the .xyz domain and warning of a scarcity of desirable .com domain names. Verisign sued XYZ and Negari, alleging that those statements violated the Lanham Act’s false advertising provisions. The district court granted

summary judgment to XYZ, holding that Verisign could not establish the elements of a Lanham Act claim. We agree. As to XYZ’s self-promoting statements, most of which concern its registration numbers, we hold that Verisign failed to produce the required evidence that it suffered an actual injury as a direct result of XYZ’s conduct. Nor can Verisign establish, we hold, that XYZ’s statements about the availability of suitable .com domain names were false or misleading statements of fact, as required for Lanham Act liability. Accordingly, we affirm the district court’s grant of

summary judgment. I. A. We begin with a brief overview of the domain name industry. A domain name is the string of characters in an Internet address. In the domain name “usc-ourts.gov,” for instance, the “.gov” element of the name is referred to as a top-level domain. Verisign is the exclusive operator of both the .com and .net top-level domains. It operates the .com registry, selling domain names to registrars who in turn sell those names to end users. Over 100 million .com domain names have been registered.

Among these 100 million registered domain names are 96 percent of all dictionary words — from apple.com to zebra.com. The worldwide domain name system is overseen by the nonprofit Internet Corporation for Assigned Names and Numbers (“ICANN”). In the early 2000s, ICANN began to consider the introduction of new top-level domains, in order to enhance competition and consumer choice. In 2012, ICANN began to accept applications for new top-level domains, and by July 2015 it had approved nearly 700 of 1,930 applications.

XYZ, founded by Negari, was among the hundreds of applicants granted the right to operate a new top-level domain. XYZ launched “.xyz” in 2014. At the heart of this case are certain promotional efforts undertaken by XYZ to market its new .xyz domain. According to Verisign, a number of XYZ’s activities crossed the line into false or misleading advertising, violating the Lanham Act and injuring Verisign’s own business operations. The statements of which Verisign complains can be divided into two rough categories, which we describe in turn.

First is a series of affirmative statements about .xyz, promoting .xyz’s popularity and touting its high registration numbers. For instance, Negari wrote a post on XYZ’s blog — titled “.xyz, the most used new gTLD [generic top-level domain] — Boooooooom!”—stating that “.xyz has received the most registrations of all new gTLDs with 447,544 domains registered.” J.A. 358. An XYZ employee later told a domain-name reseller that XYZ had over 600,000 registrations. And XYZ also promoted its registration numbers in emails to individual registrars. , By August 2015, XYZ had secured over one million .xyz registrations.

Verisign does not contest that figure, nor the literal truth of the registration numbers advanced by XYZ. Indeed, it is not disputed that those numbers match statistics maintained in the “zone file,” a database tracking domain names registered in domain extensions. Rather, the crux of Verisign’s claim is that XYZ’s registration numbers are false or misleading because they include not only registrations bought and paid for by consumers — indicating actual consumer demand — but also 375,000 .xyz registrations given away for free just after XYZ’s 2014 launch, through an agreement between XYZ and Web.com.

By including those registrations in its numbers, Veri-sign contends, XYZ misrepresented actual consumer demand for its product. Negari also promoted the popularity of the .xyz domain in media interviews. During one such interview, Negari, along with other domain operators, spoke with National Public Radio (“NPR”) as part of a story on the expansion of the top-level domain market. Introducing Negari, the NPR reporter said, “You could try to become the next .com, the next, all-purpose ending, the thing that you can stick on the back of any business name.” J.A. 593 (emphasis added).

The statement of which Verisign complains appeared after the interview, on the XYZ corporate blog: In a post titled “.xyz — the Next .com,” J.A. 596, XYZ promotes Negari’s NPR interview and states that “NPR’s [reporter] described .xyz as the next .com,” id. (emphasis added) — diverging, Verisign argues, from the actual formulation (“You could try to become the next .com”) used by NPR. The post ends with a link to the full NPR interview. Id. In addition to promoting the XYZ domain -with false statements, Verisign alleges, XYZ falsely disparaged Verisign’s competing .com domain.

Here, Verisign points to a series of statements by XYZ and Negari calling into question the availability of suitable .com domain names. First, during his interview with NPR, Negari, referring to the .com domain, said, “All of the good real estate is taken. The only thing that’s left is something with a dash or maybe three dashes and a couple numbers in it.” J.A. 593.

A few months later, a Negari blog post on XYZ’s website included this line: “Did you know that 99% of all registrar searches today result in a ‘domain taken’ page?” J.A. 628. (Negari echoed that claim in a subsequent radio interview, saying that “nine out of ten .com searches show up as unavailable.” J.A. 666.) Finally, XYZ posted a 35-second advertisement to video-sharing website YouTube, comparing a new Audi with a .xyz license plate to a dilapidated Honda with a .com plate, and saying, “With over 120 million .corns registered today, it’s impossible to find the domain name that you want.” J.A. 527.

All of those statements, Verisign claims, are false or misleading. B. In December 2014, Verisign filed suit against XYZ and Negari, claiming that its new competitor had violated the Lanham Act’s false advertising provisions. 15 U.S.C. § 1125(a)(1)(B). As is customary in Lanham Act suits, Verisign undertook a consumer survey to support its claims.

Verisign’s survey tested consumer reaction to XYZ’s self-promoting statements about its registration numbers and, according to Verisign, demonstrated that consumers understood XYZ’s references to “registrations” to signify actual consumer “purchases,” rather than .xyz registrations given away for free. After extensive discovery, XYZ moved for

summary judgment,- and the district court granted its motion. In an accompanying opinion, the court comprehensively rejected Verisign’s claim, finding that Veri-sign failed to establish a violation of the Lanham Act on multiple grounds, any one of which would have been sufficient justification for an award of

summary judgment to XYZ. First, on the threshold question of falsity, the district court held that Verisign could not show that any of the statements complained of was false or misleading. As to XYZ’s promotion of its registration numbers, the court found, those numbers were “verifiably true,” J.A. 838 — based on accurate statistics from the “zone file,” J.A. 839 — and “nothing in the record supports the notion” that XYZ’s statements were “false,” J.A. 838. The court also concluded that XYZ’s blog-post promotion of Negari’s NPR interview — including “the next .com” statement — was not false. J.A. 837.

Similarly, the district court held, Veri-sign could not establish that XYZ’s statements regarding .com availability were false or misleading. Statements like “all of the good real estate is taken” and “it’s impossible to find the domain name that you want,” the district court explained, reflect subjective opinion or “puffery,” neither of which is actionable under the Lan-ham Act. J.A. 833.

And while Negari’s assertion that 99 percent of registrar searches result in a “domain taken” message is a verifiable statement of fact, the district court concluded, Verisign had not shown that statement to be false; on the contrary, Verisign’s own data demonstrated that .com names “are largely unavailable.” J.A. 837. Because Verisign had “fail[ed] to meet the first element” of a Lanham Act claim, J.A. 839 — a false or misleading statement — the district court could have awarded

summary judgment without further analysis. Instead-, it went on to identify other, independent defects in Verisign’s case. Under the Lanham Act, even a false statement is not actionable unless it is “material,” or likely to influence a consumer decision. Here, the district court held, Verisign had presented “no evidence to show that consumers were influenced by the statements” in question, having chosen not to test materiality in its consumer survey. J.A. 840.

Nor, the district court held, could Verisign show that any of XYZ’s statements actually deceived consumers, as required under the Lanham Act, pointing again to shortcomings in Ver-isign’s survey evidence. Finally, and providing yet another independent justification for its decision, the district court held that Verisign had failed to establish the requisite causal connection between the allegedly false statements and any damages actually suffered by Verisign.

Verisign’s first theory of damages — that XYZ’s false advertising had diverted sales from Verisign’s domains to .xyz — rested on the analysis of expert Lauren Kindler. But the district court excluded that testimony on the ground that Kindler’s methods and conclusions were unreliable, primarily because Kindler relied on data “point[ing] only to correlation, not causation.” J.A. 843.

As to Verisign’s second theory of damages — that XY'Z’s statements -disparaging the .com domain had lessened the goodwill associated with Verisign’s product — the court concluded that Verisign had failed to produce any evidence of lessened goodwill apart from its own subjective belief that its reputation had been harmed. Verisign timely noted this appeal. II. A. We review de novo the district court’s grant of

summary judgment, viewing the evidence in the light most favorable to Verisign, the nonmoving party, and drawing all reasonable inferences in Verisign’s favor. Design Res., Inc. v. Leather Indus. of Am., 789 F.3d 495, 500 (4th Cir. 2015). To show that

summary judgment should not be awarded, Verisign “must provide more than a scintilla of evidence— and not merely conclusory allegations or speculation — upon which a jury could properly find in its favor.” Id. (internal quotation marks and citation omitted). We review the district court’s evidentiary ruling regarding Verisign’s expert testimony for abuse of discretion. Meyers v. Lamer, 743 F.3d 908, 915 (4th Cir. 2014). B.

The Lanham Act prohibits the “false or misleading description of fact, or false or misleading representation of fact, which ... in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person’s goods, services, or commercial activities[J” 15 U.S.C. § 1125(a)(1).

Accordingly, a plaintiff asserting a false advertising claim must establish each of five elements, showing that: (1) the defendant made a false or misleading description of fact or representation of fact in a commercial advertisement about his own or another’s product; (2) the misrepresentation is material, in that it is likely to influence the purchasing decision; (3) the misrepresentation actually deceives or has the tendency to deceive a substantial segment of its audience; (4) the defendant placed the false or misleading statement in interstate commerce; and (5) the plaintiff has been or is likely to be injured as a result of the misrepresentation, either by direct diversion of sales or by a lessening of goodwill associated with its products..

Design Resources, 789 F.3d at 501 (quoting PBM Prods., LLC v. Mead Johnson & Co., 639 F.3d 111, 120 (4th Cir. 2011)). “[FJailure to establish any one” of these five elements is “fatal” to a plaintiffs claim. Id. And importantly, Verisign must be able to point to at least one challenged statement that satisfies all five Lanham Act requirements; as the parties agree, a Lanham Act claimant may not mix and match statements, with some satisfying one Lanham Act element and some satisfying others.

See Johnson & Johnson Vision Care, Inc. v. 1-800 Contacts, Inc., 299 F.3d 1242, 1248 (11th Cir. 2002) (rejecting argument that challenged statements should be evaluated “in concert” to determine whether any or all violated Lanham Act). Here, we agree with the district court that Verisign failed to present evidence sufficient to show that any of XYZ’s challenged statements violated the Lanham Act.

We address in turn the two categories of statements at issue: self-promotional statements about .xyz’s registration numbers and popularity, and disparaging statements about the availability of Veri-sign’s .com domain names. 1. The crux of Verisign’s claim is that XYZ used deceptive statements about its registration numbers and popularity to manufacture an unwarranted “gold rush” for .xyz domains.

By touting registration numbers that included free registrations and claiming that NPR had dubbed it the “next .com,” Verisign argues, XYZ created a false picture of consumer demand that was intended to drive up .xyz sales to registrars and website owners. XYZ, for its part, insists that its registration numbers were true; that it had no obligation to separate out registrations arising from a legitimate commercial promotion; and that as the district court found, Verisign’s consumer survey evidence failed to establish that consumers were misled by XYZ’s numbers.

The blog post stating that it had been deemed the “next .com” by NPR, XYZ argues, is either true when viewed in context of the NPR story or harmless exaggeration, non-actionable as “puffery.” As described above, the district court sided thoroughly with XYZ, holding that XYZ’s statements were not false as a matter of law and that Verisign also had' failed to establish materiality and deception, two additional and independent elements of a Lanham Act claim. Without casting any doubt on those holdings by the district court, we find that we need not review them here.

We agree with the district court that Verisign failed to establish yet another Lanham Act element — that it suffered an injury flowing directly from the challenged statements — and affirm the grant of

summary judgment to XYZ on that ground alone. To recover damages under the Lanham Act, Verisign must show not only false advertising by XYZ, but also that XYZ’s statements caused Verisign actual damages. See PBM Products, 639 F.3d at 122 (summary judgment properly awarded to defendant in Lanham Act case because plaintiff could not prove that allegedly false statements caused any damages); Xoom, Inc. v.

Imageline, Inc., 323 F.3d 279, 286 (4th Cir. 2003) (Lanham Act plaintiff must “prove actual damages and a causal link between those damages and the Lanham Act violation”), abrogated on other grounds by Reed Elsevier, Inc. v. Muchnick, 559 U.S. 154, 130 S.Ct. 1237, 176 L.Ed.2d 18 (2010). This is not a minor or technical element of a Lanham Act claim; indeed, as the Supreme Court has explained, it is the core requirement that a plaintiff “show economic or reputational injury flowing directly from the deception wrought by the defendant’s advertising” that assures

Article III standing in Lanham Act cases. Lexmark Int’l, Inc. v. Static Control Components, Inc., — U.S. -, 134 S.Ct. 1377, 1391, 188 L.Ed.2d 392 (2014). Accordingly, we have made clear that the indispensable fifth element of a Lanham Act claim is that “the plaintiff has been or is likely to be injured as a result of the [alleged] misrepresentation, either by direct diversion of sales or by a lessening of goodwill associated with its product.” Design Resources, 789 F.3d at 501 (quoting PBM Products, 639 F.3d at 120).

With respect to XYZ’s self-promoting statements, Verisign proceeded under the first of these damages theories, claiming that XYZ’s fake “gold rush” directly diverted sales from Verisign’s .net registry to XYZ’s new .xyz domain. For support, Verisign rested on its expert, Lauren Kindler, who indeed opined that XYZ’s statements sapped Verisign’s profits by redirecting registrations away from Veri-sign’s .net registry. Relying on Verisign’s claim that XYZ’s false statements were the reason consumers signed up for .xyz domains, Kindler also suggested that Ver-isign was entitled to disgorgement of XYZ’s profits.

Exercising its gatekeeping function under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), the district court excluded Kindler’s report, leaving Verisign with nothing to show injury in the form of diversion of sales. Acting as gatekeepers, district courts are charged with ensuring that expert testimony of all kinds is relevant and reliable. See Cooper v. Smith & Nephew, Inc., 259 F.3d 194, 199 (4th Cir. 2001) (citing Daubert); see also Kumho Tire Co., Ltd. v.

Carmichael, 526 U.S. 137, 147, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) (district court’s “basic gatekeeping obligation” applies to all expert testimony). Here, citing Daubert, the district court concluded that Kindler’s “methods ... are questionable” and that her “conclusions are not reliable,” primarily because Kindler’s analysis failed to distinguish between “correlation” and “causation.” J.A. 843. We agree with the district court. Kindler concluded that Verisign lost approximately $527,000 in profits as a result of .xyz registrations diverted from .net.

To arrive at that figure, she began by identifying a decline in registrations on the .net domain between June 2014 and May 2015 — the time period during which XYZ made the allegedly false statements at issue. She then attributed to XYZ a percentage of those lost registrations, based on .xyz’s market share of the new top-level domain market. Multiplying that percentage by what would have been Verisign’s profits on the lost registrations, she arrived at her figure for lost profits.

And Kindler followed a similar approach to calculating potential disgorgement damages, starting with the total registration fees XYZ collected during the relevant time period and subtracting various costs to find the total profits XYZ earned as a result of its alleged misrepresentation. That analysis suffers from what we have identified as a “fatal flaw” in calculating Lanham Act damages: It assumes rather than demonstrates that every .xyz registration during the relevant time period was the result of XYZ’s allegedly false statements.

See PBM Products, 639 F.3d at 122 (“The fatal flaw in [plaintiffs] economic information was that its expert assumed that every sale [defendant] made was attributable to the [allegedly false statement].”) Kindler’s report does show that Verisign experienced a drop in .net registrations after .xyz — along with other new top-level domains — became available, and while XYZ was making the statements of which Verisign complains.

What it' does not show, however, is anything other than a temporal link between XYZ’s statements and the drop-off in .net registrations— “correlation,” in the district court’s words, but not “causation.” J.A. 843. And that is not enough under the Lanham Act, where the plaintiff bears the burden of proving a “causal link between actual damages and [the defendant’s] actions.” Xoom, 323 F.3d at 286; see Lexmark, 134 S.Ct. at 1391 (plaintiff must show injury “flowing directly” from defendant’s advertising).

To be sure, Kindler does suggest that XYZ’s profits during the time period in question were “at least in part due to the alleged false and misleading advertising,” and opines that “participants [in the domain-name market] view registration numbers as a measure of demand,” so that “inflated registration figures [like XYZ’s] are likely to entice market participants.” S.A. 904. But Kindler’s analysis does not quantify any such effect, beyond establishing a temporal connection.

And a conclusory repetition of Verisign’s “gold rush” theory of the case does little, if anything, to advance the proximate-cause showing demanded by Lexmark and PBM Products. Finally, we note that it is perhaps not surprising that Verisign would experience difficulty in proving damages directly caused by XYZ’s self-promoting statements. The record suggests that XYZ’s boasts about its registration numbers and NPR interview were distributed to a narrow audience, comprised mostly of readers of XYZ’s blog and a small percentage of registrars.

Indeed, XYZ argues that its statements were disseminated to so small a group that they do not even constitute “commercial advertising or promotion” covered by the Lanham Act. See 15 U.S.C. § 1125(a)(1)(B) (prohibiting misrepresentations in “commercial advertising or promotion”); Fashion Boutique of Short Hills, Inc. v. Fendi USA, Inc., 314 F.3d 48, 56 (2d Cir. 2002) (statement qualifies as “commercial advertising or promotion” only if it is “disseminated sufficiently to the relevant purchasing public”). We need not decide that question here.

We simply'observe that XYZ’s statements appear to have had limited potential to influence the domain-name market, particularly at a time when hundreds of new top-level domains were clamoring for attention in a newly competitive market. In sum, the district court did not abuse its discretion in excluding Kindler’s report, and we agree that Verisign has not established that XYZ’s statements regarding registration numbers or the NPR interview were causally linked to damages in the form of diverted registrations.

For that reason alone, Verisign cannot prevail on its Lanham Act claim with respect to XYZ’s self-promoting statements. 2. We turn next to the second category of statements at issue: statements by XYZ that call into question the availability of desirable .com domain names.

The thrust of Verisign’s claim here is that XYZ falsely disparaged Verisign’s product— the .com domain — by telling the consuming public that Verisign’s inventory of .com names was so depleted that only cumbersome or unsightly names remained: XYZ’s YouTube advertisement warned that “[w]ith over 120 million .corns registered today, it’s impossible to find the domain name that you want,” J.A. 527, and Negari said in an interview that “[a]ll of the good [.com] real estate is taken,” leaving “only .. 1 something with a dash or maybe three dashes and a couple of numbers,” J.A. 593.

And XYZ bolstered those claims, Verisign alleges, with “phony statistics” asserting that “99% of all registrar searches today result in a ‘domain taken’ page[J” J.A. 628. XYZ, naturally, takes a different view. According to XYZ, claims as to whether “all of the good [.com] real estate is taken,” or it is “impossible to find the domain name that you want,” are either statements of opinion or harmless “puffery,” or exaggeration, neither of which is actionable under the Lanham Act.

As for the 99 percent statistic, XYZ insists, even Veri-sign concedes that it is literally true, and there is no evidence that it tended to mislead consumers. Here again, the district court sided with XYZ, concluding that Verisign could not satisfy the very first element of a Lanham Act claim by showing that any of the statements in question was a false or misleading factual statement. We agree, and therefore need not consider the district court’s alternative grounds for ruling against Verisign with respect to this group of statements.

As we have explained, the threshold element of a Lanham Act claim is a showing that the defendant “made a false or misleading description of fact or representation of fact” about his own product or, as relevant here, the product of a competitor. Design Resources, 789 F.3d at 501. That showing has two distinct components: “[T]he contested statement must be false, and it must be a representation of fact.” Id.

As to falsity, a plaintiff may show either literal falsity — a statement that is “false on its face” — or that a statement, while literally true, is misleading, in that it is “likely to mislead and to confuse consumers given the merchandising context.” Scotts Co. v. United Indust. Corp., 315 F.3d 264, 273 (4th Cir. 2002).

But in order to be “false” in any way cognizable under the Lanham Act, a statement must also be one of fact — that is, a “specific and measurable claim, capable of being proved false or of being reasonably interpreted as a statement of objective fact.” Design Resources, 789 F.3d at 502 (quoting Pizza Hut, Inc. v. Papa John’s Int’l, Inc., 227 F.3d 489, 496 (5th Cir. 2000)). The requirement that an actionable statement be one of fact necessarily excludes certain other kinds of statements from liability. See Am. Italian Pasta Co. v.

New World Pasta Co., 371 F.3d 387, 390 (8th Cir. 2004) (outside the category of actionable statements of fact “a category of non-actionable statements exists”). Statements of opinion, rather than fact, cannot be the basis of a Lanham Act claim. See id. at 391; Design Resources, 789 F.3d at 502. And the Act does not reach what is known as “puffery” — “exaggerated advertising, blustering, and boasting,” or “vague” and “general claim[s] of superiority” — on the theory that no reasonable person would rely on such superlatives. Pizza Hut, 227 F.3d at 496 (quoting 4 J.

Thomas McCarthy, McCarthy on Trademark and Unfair Competition § 27.38 (4th ed. 1996)); see Am. Italian Pasta Co., 371 F.3d at 391 (“[p]uffery and statements of fact are mutually exclusive”). Like the district court, we think that XYZ’s statements concerning the availability of desirable .com names constitute opinion or puffery, not statements of fact on which reasonable consumers could rely. XYZ’s YouTube video claim that it is “impossible to find the domain name that you want,” J.A. 527, we conclude, cannot be interpreted as a verifiable statement of objective fact. Cf. Design Resources, 789 F.3d at 502.

That is in part thanks to the indefinite nature of the referenced “you”: Whether an anonymous “you” can find the domain name of his or her choosing is not something that can be proven true or false. See Am. Italian Pasta Co., 371 F.3d at 391 & n.5 (phrase “America’s Favorite Pasta” cannot be construed as verifiable statement of fact in part because no “specific person or [ ] identifiable group” is indicated as favoring the product).

Instead, taken as a whole — with the exaggerated “impossible” and the implied supposition about what “you” might want — the statement conveys an opinion about consumer preferences, a blustery assertion of the subjective value (or lack thereof) of available .com names that qualifies as puffery, or some combination of the two. So, too, .with Negari’s claim during his interview that “[a]ll of the good real estate is taken.” J.A. 593. As the district court explained, what counts as a “good” domain name is a matter of opinion, not fact: “[Wjhether a domain is a ‘good’ domain is subjective to the registrant.” J.A. 837.

Verisign leans heavily on Negari’s next sentence — “The only thing that’s left is something with a dash or maybe three dashes and a couple of numbers in it,” J.A. 593— and insists that it is literally false, because there are at least some available .com names that do not include dashes and numbers. But we read the two sentences together, see Castrol Inc. v.

Pennzoil Co., 987 F.2d 939, 946 (3d Cir. 1993) (“court must view the face of the statement in its entirety”), and in context, we think that the overall message must be construed as one of subjective opinion: The available .com names are not “good” because they involve dashes and numbers. And while Negari may have exaggerated when he said that the “only [.com names] left” are of the dashes and numbers variety, that is precisely the kind of puffery or bluster on which no reasonable consumer would rely. See id. at 945 (describing puf-fery as “exaggeration or overstatement”).

Particularly when it comes to spoken statements like Negari’s, which may be offered more casually than their written counterparts, we must take care not to label as “literally false” what really is no more than a colloquial exaggeration, readily understood as such. See Schering-Plough Healthcare Prods., Inc. v. Schwarz Pharma, Inc., 586 F.3d 500, 512-13 (7th Cir. 2009) (doctrine of “literal falsity” must take account of fact that “Americans ... say things like T am literally out of my mind.’ ”).

Finally, there are XYZ’s statistics: the claim that “99% of all registrar searches today result in a ‘domain taken’ page,” J.A. 628, repeated by Negari on a local radio show in a slightly different form, J.A. 666 (“nine out of ten .com searches show up as unavailable”). The district court agreed with Verisign that these statements qualify as statements of verifiable fact; the problem for Verisign, the district court held, is that they are verifiably true.

As the district court explained, Verisign’s own data shows that out of approximately two billion requests it receives each month to register a .com name, fewer than three million — less than one percent — actually are registered. And indeed, Verisign’s expert on domain name availability conceded the accuracy of the data behind the 99 percent statistic. Verisign’s real argument is that even if literally true, the 99 percent figure qualifies as “false” for Lanham Act purposes because it is misleading.

See Scotts Co., 315 F.3d at 273 (to show falsity, plaintiff may establish either that statement is literally false or that it is misleading). Specifically, Verisign insists that the 99 percent number is a “naive” metric of unmet demand because it includes automated searches undertaken as part of a highly competitive fight among registrars for high-demand, previously registered domain names. Appellant Br. at 35.

Even on its face, this argument is perplexing: It is not clear why the fact that registrars battle fiercely and on a daily basis over a limited supply of desirable .com names would make it misleading to use a statistic — literally true — that conveys the difficulties associated with registration of a preferred .com name. But even if we were to credit the logic of this argument, Verisign’s position could not prevail. When a Lanham Act plaintiff relies on literal falsity, no evidence of consumer confusion is required. See Scotts Co., 315 F.3d at 274.

But where, as here, a plaintiff argues that a statement is literally true but nevertheless misleading, then it is incumbent on the plaintiff to present “extrinsic evidence of consumer confusion,” generally in the form of consumer survey evidence. Id. at 274, 276; Design Resources, 789 F.3d at 501; see also, e.g., PBM Products, 639 F.3d at 120 (“A court may find on its own that a statement is literally false, but, absent a literal falsehood, may find that a statement is impliedly misleading only if presented with evidence of actual consumer deception.”) (internal quotation marks and citation omitted).

Verisign cannot meet that burden here. That a literally true statement “tend[s] to mislead or confuse consumers” is “most often proved by consumer survey data.” Scotts Co., 315 F.3d at 276 (internal quotation marks and citation omitted); see also Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharm. Co., 290 F.3d 578, 588 (3d Cir. 2002) (plaintiff “should have been required to prove through a consumer survey that the [statement] actually misled or had a tendency to mislead consumers”).

But while Verisign did conduct a consumer survey, it tested only consumer reaction to XYZ’s self-promoting statements about its .xyz registration numbers, and not consumer reaction to XYZ’s claims regarding .com availability. As a result, Verisign can offer no consumer survey evidence bearing on the 99 percent statistic cited by XYZ and whether it had the misleading effect posited by Verisign. Nor did Verisign advance any other evidence showing that the literally true 99 percent figure misled or tended to mislead consumers.

And while we have considered whether an intent to deceive might be enough to trigger a presumption of consumer confusion, see Scotts Co., 315 F.3d at 281-82, Verisign could not rely on any such presumption here, having failed to produce evidence that XYZ intended to mislead consumers by sharing a statistic supported by Verisign’s own data. In short, like the district court, we conclude that none of XYZ’s challenged statements about .com availability is a false or misleading statement of fact, as required to meet the first element of a Lanham Act claim. For that reason alone, the district court properly awarded

summary judgment to XYZ with respect to these statements. III. For the foregoing reasons, we affirm the district court’s grant of

summary judgment to defendants XYZ and Negari. AFFIRMED . Under the agreement, Web.com gave a $3 million dollar advertising credit to XYZ, and, in exchange, XYZ gave Web.com $3 million dollars’ worth of .xyz domain names (375,000 names). Those names were then given by Web.com to its subscribers for free trial periods. . Although the challenged statements refer to the .com domain, Verisign’s lost profit calculation rests on a decline in sales associated with its .net domain, on the theory that .net and .xyz are closer substitutes.

Verisign’s reliance on lost .net registrations may also be explained by the fact that .com registrations continued to increase after XYZ’s statements. . With respect to XYZ’s statements regarding .com availability, Verisign advances a separate theory of injury, alleging that those statements caused a "lessening of goodwill associated with its products." See Design Resources, 789 F.3d at 501 (identifying both diversion of sales and lessening of goodwill as cognizable Lanham Act injuries).

As noted above, the district court rejected that theory, too, finding that Verisign had failed to produce sufficient evidence of lost goodwill. Because we hold below that XYZ's disparaging statements do not violate the Lanham Act for other reasons, we have no occasion to review that determination. . Verisign has taken the position that in assessing the dissemination and impact of XYZ’s statements, the relevant market is not the general public but rather the registrars who purchase domain names and then resell them to end users.

It strikes us as especially unlikely that these savvy industry players would construe XYZ's claims about .com availability as factual statements and rely on them accordingly.