2016 TRADEMARK YEAR IN REVIEW
2016 TRADEMARK YEAR IN REVIEW
Mark P. McKenna & Shelby Niemann*
This brief Essay reviews some of the most significant developments in trademark law during the past year. In most cases, we have interpreted “year” fairly liberally, particularly to highlight some longer-term trends. We focus on six areas: (1) the constitutionality of section 2(a) of the Lanham Act; (2) the Fourth Circuit’s Belmora decision and the availability of section 43(a) claims when the plaintiff has not used a mark in the United States; (3) the effect of B & B Hardware; (4) injunctive relief and the presumption of irreparable harm; (5) nominative fair use; and (6) initial interest confusion.
I. In re Tam
Section 2(a) of the Lanham Act continued to dominate trademark news this year, as it has the last couple of years. But this year an Asian rock band stole the spotlight from the Washington football team, giving us a Supreme Court battle.
In In re Tam, the Federal Circuit, sitting en banc, overturned the Trademark Office’s refusal to register the band name THE SLANTS, finding the disparagement bar on which the Trademark Office based the refusal to be unconstitutional. Expressly overruling In re McGinley, the Federal Circuit held that the prohibition on registering disparaging marks did not survive strict scrutiny because: (1) it is not content or viewpoint neutral; (2) it regulates the expressive aspects of a mark, not its function as commercial speech; (3) it is not government speech or a government subsidy; and (4) it would not pass the Central Hudson test even if it were a regulation of commercial speech.
Much has already been said about the difficult First Amendment questions raised by Tam, and surely much more will be said in the coming months. We highlight three points here: (1) the significance of denial of federal registration; (2) the problem with focusing on content and viewpoint neutrality; and (3) the question of whether commercial speech will continue to receive different treatment.
The consequences of a denial of federal registration may turn out to be central to the case. First, the Court’s understanding of the consequences of denial of registration is likely to determine the broad framing of the case. To the extent the Court sees the registration bar as an effective prohibition on speech, it is likely to subject the provision to strict scrutiny (whether it is characterized as content-based or viewpoint-based), and the constitutionality of the provision will be a function of the government’s interest in prohibiting registration of disparaging terms and whether section 2(a) is narrowly tailored to that interest.
If, on the other hand, the Court recognizes that denial of a registration does not prohibit any speech (i.e., does not impose any criminal or civil penalties for speech but merely denies benefits), then it is more likely to view the case through the lens of government program or government speech cases. In that case, the issue will be whether denial of the benefits of registration is significant enough that it amounts to government leveraging of its funding to control speech outside of the government program.
The conventional wisdom has long been that denial (or cancellation) of registration not only does not prevent use of a mark, it does not even prevent enforcement of the mark under federal law (particularly section 43(a) of the Lanham Act). Denial simply means that the mark will not receive any of the benefits of registration. The mark will not be presumed valid, nor will it be eligible for incontestable status. The mark owner will not get to claim nationwide rights by virtue of statutory constructive use. The government also will not seize imported goods bearing the claimed mark, since only goods bearing registered trademarks are subject to forfeiture. And none of the civil or criminal counterfeiting provisions will apply, as counterfeits are by definition those that make unauthorized use of registered marks. These consequences are not meaningless, of course, but they fall far short of unenforceability.
The Federal Circuit cast considerable doubt on this understanding of registration, however, in In re Tam. Echoing a small number of prior district court cases, the Federal Circuit strongly suggested that marks barred from registration under section 2 were unenforceable as a matter of federal law, and likely as a matter of state law as well. There are reasonable arguments to be made about how we should understand the relationship between registration and enforcement of unregistered rights, and this is not the only setting in which that issue has recently come to the fore. For our purposes it is enough to note that much more than the section 2(a) bars are at stake if the Supreme Court were to embrace the Federal Circuit’s thinking and accept that denial of registration entails lack of enforceability. For one thing, it is difficult to see how such a conclusion could be limited to the disparagement bar or even to section 2(a). For if the unregistrability of a mark establishes its unprotectability (because the exclusions in section 2 reflect public policies that also apply to unregistered marks), then it is not clear why one would not draw the same conclusion about all the bases for denial of registration. A conclusion that the disparagement bar is unconstitutional because denial of registration entails unprotectability could then have much broader effects on trademark law and practice, requiring a fundamental rethinking of the relationship between registration and protection of unregistered marks.
There is another way in which the Court will have to grapple with the extent to which the section 2(a) limitations can be distinguished from the rest of section 2, and perhaps even the rest of the Lanham Act, if it agrees with the Federal Circuit that the disparagement bar is unconstitutional. Section 2 prohibits registration of, among other things, generic and descriptive terms; marks that consist of or comprise the flag or coat of arms or other insignia of the United States; marks likely to be confused with prior marks; and primarily geographically deceptively misdescriptive marks. All of these prohibitions are content-based, so if the problem with the disparagement bar is that it is not content neutral, then all of section 2 is unconstitutional. And so, likely, are all of the important provisions of the Lanham Act, since infringement and dilution obviously depend on the content of that party’s mark.
It is possible that the Court could distinguish the disparagement bar on the ground that, like many of the other section 2(a) bars but unlike most of the remaining section 2 bars, the disparagement bar is not just content-based, it is viewpoint-based. Section 2(a) prohibits registration of THE SLANTS not only because of its content, but because the government disapproves of the disparaging message of the mark. Positive, non-disparaging references to people of Asian descent are allowed; only negative references are disallowed. It is worth noting, however, that as a matter of First Amendment doctrine, it is actually quite rare for a case to turn on the distinction between content-based and viewpoint-based regulations—a distinction the Supreme Court has acknowledged is “not a precise one.”
More to the point, whether one regards the disparagement bar as viewpoint-based depends on whether that is to be judged from the perspective of the speaker or the audience. That matters because it’s quite clear that the PTO refuses registration of marks it deems disparaging regardless of the viewpoint of the applicant. Indeed, the facts of In re Tam prove the point—Tam claimed that he should not be denied registration because he was trying to reclaim the term to strip it of its disparaging meaning. But that purpose was irrelevant to the mark’s registrability because disparagement is judged from the perspective of a “substantial composite” of the relevant group. As a result, section 2(a) can only be regarded as viewpoint based if the viewpoint of the audience rather than the viewpoint of the speaker controls. Perhaps it does, though the cases the Federal Circuit cited for that proposition are not clearly on point here.
Finally, there is the question of whether the government will get greater deference here because the registration bars regulated commercial speech, unlike the regulations at issue in most of the other cases involving content-based or viewpoint-based regulations. The Federal Circuit said that it did not get greater deference because the disparagement bar regulates the expressive aspects of a mark, not its function as commercial speech. According to the Federal Circuit:
[I]t is always a mark’s expressive character, not its ability to serve as a source identifier, that is the basis for the disparagement exclusion from registration. The disparagement provision must be assessed under First Amendment standards applicable to what it targets, which is not the commercial-speech function of the mark.
That approach is notable for several reasons. First, it is hard to imagine that the sort of disentanglement the Federal Circuit’s approach requires would be workable. If it would not, then courts will have to decide whether marks with expressive value should be treated as commercial or non-commercial speech. If the Federal Circuit is right that courts should treat these mixed character marks as non-commercial, then it is unclear why the marks are otherwise registrable. As Judge Reyna asked in his dissent:
[I]f the expressive content of the mark precludes regulation, on what authority may the government grant Mr. Tam the exclusive right to use this mark in commerce? Whatever standard of scrutiny protects the content of Mr. Tam’s trademark from government regulation, that same standard must necessarily be overcome by the government’s substantial interest in the orderly flow of commerce, or no trademark could issue.
Perhaps even more significantly, if courts are to begin focusing on the expressive content of marks and subjecting their regulation to strict scrutiny despite their commercial character, that could have truly revolutionary consequences in the infringement and dilution contexts. Use of a party’s mark in part for its expressive value—which, given the difficulty in disentangling the expressive from the commercial, might include a large number of uses—would trigger First Amendment scrutiny. And since protecting ordinary commercial interests typically is not enough to satisfy strict scrutiny, the consequences here could be enormous.
In his 2014 Year in Review, Professor McKenna suggested that, by focusing courts’ attention on the “zone of interests” protected under the Lanham Act and proximate causation, the Supreme Court’s Lexmark decision might make courts more skeptical of claims of harm in more expansive trademark infringement actions. The jury is still out on that prediction. But the Fourth Circuit’s decision in Belmora LLC v. Bayer Consumer Care AG suggests that Lexmark has had some influence on the way courts understand the scope of section 43(a), and particularly on the question of whether a plaintiff must have trademark rights in order to pursue a claim under that section.
Belmora involved one company, Bayer Consumer Care AG (BCC), that owned the trademark “FLANAX” in Mexico and had sold naproxen sodium pain relievers under that mark in Mexico (and other parts of Latin America) since the 1970s. Another company, Belmora LLC, owned the FLANAX trademark in the United States and had used the mark in the United States since 2004 for its own naproxen sodium pain relievers. BCC successfully petitioned to cancel Belmora’s registration for the FLANAX mark on the ground the mark was deceptive. Belmora appealed the Trademark Trial and Appeal Board’s (TTAB) decision to the Eastern District of Virginia. In the meantime, Bayer filed a separate complaint against Belmora for false association under section 43 of the Lanham Act. “BCC and its U.S. sister company Bayer Healthcare LLC (‘BHC,’ and collectively with BCC, ‘Bayer’) contend[ed] that Belmora used the FLANAX mark [in the United States] to deliberately deceive Mexican-American consumers into thinking they were purchasing BCC’s product.” After the two cases were consolidated, the district court reversed the TTAB’s cancellation order and dismissed the false association and false advertising claims. Bayer appealed.
As the Fourth Circuit put it, the case required consideration of “whether the Lanham Act permits the owner of a foreign trademark and its sister company to pursue false association, false advertising, and trademark cancellation claims against the owner of the same mark in the United States.” That’s a particularly difficult question because it’s not just an issue of whether a party that lacks trademark rights can assert a claim under section 43(a), but rather a question of whether it matters that the claimant is a foreign company that lacks U.S. trademark rights because it has never used the mark in the United States. That is, Belmora has important implications for the territoriality of trademark rights.
On the general question of whether a party that does not have trademark rights can assert a claim under section 43(a), the court noted that “the plain language of § 43(a) does not require that a plaintiff possess or have used a trademark in U.S. commerce as an element of the cause of action.” Indeed, in that respect, “§ 43(a) stands in sharp contrast to . . . § 32, which is titled as and expressly addresses ‘infringement.’”
That argument resonates with the traditional distinction between trademark infringement and unfair competition. Registration under pre-Lanham Act statutes was available only to “technical trademarks,” which were distinguished from mere “trade names” and other matter that did not indicate the source of a party’s goods. Limitations on registration were important because pre-Lanham Act statutes gave a federal cause of action only to owners of registered trademarks whose marks were used by others in interstate commerce. Thus, only the owners of technical trademarks that had taken the initiative to register their marks could bring federal claims. Those claims under federal statutes were denominated trademark infringement actions because by definition they involved the use of technical trademarks. Unfair competition claims, by contrast, were common-law claims available to those that did not own technical trademarks and therefore could not have received a federal registration or brought a federal claim.
This divided state of affairs continued through the early years of the Lanham Act, though not without growing concern among some of the Lanham Act’s advocates that the Supreme Court’s decision in Erie would lead to disuniformity in the law of unfair competition. Those advocates believed the new federal statute needed to provide a federal statutory basis for unfair competition claims, and some even claimed that the Lanham Act in fact provided such a basis. Edward Rogers, one of the Lanham Act’s primary drafters, focused particularly on section 44, which he argued could be interpreted broadly to create the federal law of unfair competition. A few courts initially accepted that view, but one had to read section 44 quite creatively to find in it a federal unfair competition cause of action. Most courts therefore turned their attention to section 43(a)—first to ground a federal unfair competition claim in cases not involving trademarks but other false statements, and then eventually to allow federal claims for infringement of unregistered trademarks and trade dress.
To fit these latter types of claims into the text of the statute, courts interpreted the term “origin” broadly to refer not only to geographic origin but also to origin of manufacture. The Supreme Court has on occasion shown some doubts about the legitimacy of that interpretation, but it has now clearly accepted the conclusion that section 43(a) provides a cause of action for infringement of unregistered marks.
Insofar as we see section 43(a) primarily as a replacement for common-law unfair competition claims, the Fourth Circuit’s interpretation in Belmora makes some sense. Unfair competition claims by definition involved unregistered, and most commonly unregistrable, subject matter, so it would be odd to read into section 43(a) a requirement that the plaintiff demonstrate ownership of a trademark. The problem with that approach is that federal courts have spent the better part of the last fifty years eviscerating the distinction between trademark infringement and unfair competition, primarily for the purpose of eliminating the additional proof requirements that once attended unfair competition actions. Indeed, unfair competition has been so assimilated into trademark law that courts routinely insist that there is no meaningful difference between the infringement of registered and unregistered marks. It is therefore not surprising that so many of the Fourth Circuit’s pre-Belmora cases treated the plaintiff’s use of a mark in U.S. commerce as a prerequisite for a section 43(a) claim.
Perhaps Belmora portends a return to the old days, when trademark infringement and unfair competition were more clearly distinct causes of action. We think that might actually be a welcome development, though we should think carefully about the requirements for proving unfair competition rather than simply treating it as an infringement claim without the requirement that the plaintiff have trademark rights. That would be a much better reflection of historical practice, which we perhaps disregarded without nearly enough thought.
But what is especially notable about Belmora is its failure to recognize the implications of its decision for the territoriality of trademark rights. Few concepts are more fundamental in trademark law than the notion that rights are territorial in nature. As the Ninth Circuit said in Grupo Gigante, “[p]riority of trademark rights in the United States depends solely upon priority of use in the United States, not on priority of use anywhere in the world. Earlier use in another country usually just does not count.”
It is also black-letter law that trademark rights arise in the United States through use, and that those rights exist only in the areas of use. Of course, the Lanham Act also makes it possible to establish rights in advance of use by filing an intent-to-use application or one of the types of applications specifically available to foreign applicants. But even those provisions ultimately require use in the United States as a condition of continued rights.
It is one thing for courts to make an exception to these bedrock rules by recognizing the well-known marks doctrine, under which the owner of a well-known trademark can prevent use of that mark even in countries in which it has not used or registered. That doctrine has, of course, divided courts in the United States. And while, in our view, the Second Circuit was right to reject the doctrine, but there are reasonable arguments to be made on both sides of that debate. But as least the well-known marks doctrine makes an exception to the territoriality principle only in a limited range of cases—those involving “well-known” marks. Belmora potentially upends that balance, seemingly making it possible for foreign mark owners to prevail not only when their marks are well-known, but whenever use of that mark in the United States causes confusion. That’s no longer a limited exception to territoriality—it’s a rejection of it.
III. B & B Hardware
In its 2015 B & B Hardware, Inc. v. Hargis Industries, Inc. decision, the Supreme Court resolved a split in the lower courts over whether TTAB decisions have preclusive effect in subsequent litigation in federal courts. B & B dealt specifically with the preclusive effect of likelihood of confusion findings in opposition proceedings. B & B opposed registration of Hargis’s application to register SEALTITE for “self-piercing and self-drilling metal screws for use in the manufacture of metal and post-frame buildings,” claiming the mark was likely to cause confusion in light of B & B’s prior registration of SEALTIGHT for “threaded or unthreaded metal fasteners and other related hardwar[e]; namely self-sealing nuts, bolts, screws, rivets and washers, all having a captive o-ring, for use in the aerospace industry.” The TTAB agreed with B & B and denied registration to Hargis.
While the opposition proceeding was pending, B & B sued Hargis in federal court for trademark infringement. While the civil action was pending, the TTAB ruled in B & B’s favor, after which B & B argued that “Hargis could not contest likelihood of confusion because of the preclusive effect of the TTAB decision.” The district court rejected that argument, noting that the TTAB is not an Article III court, and the jury proceeded to find that there was no likelihood of confusion. On appeal, the Eighth Circuit accepted “for the sake of argument that agency decisions can ground issue preclusion,” but it nevertheless affirmed the judgment in Hargis’s favor. It did so on the grounds that the TTAB uses different likelihood of confusion factors than do courts in the Eighth Circuit; the TTAB placed too much emphasis on the appearance and sounds of the marks; and the burdens of persuasion were different in the two proceedings.
The Supreme Court reversed, clarifying that the decisions of administrative agency tribunals generally, and the TTAB specifically, can have preclusive effect. Regarding TTAB proceedings, the Court held that, “[s]o long as the other ordinary elements of issue preclusion are met, when the usages adjudicated by the TTAB are materially the same as those before the district court, issue preclusion should apply.” More specifically the Court held that likelihood of confusion findings made by the TTAB can have preclusive effect because the likelihood of confusion standard is the same in the TTAB and in district court infringement proceedings.
But the fact that TTAB determinations can have preclusive effect does not mean they always will. While the same standard nominally applies in both settings, the Court understood that the TTAB often will apply that standard very differently than would a court in civil litigation. As the Court recognized, “[t]he Board typically analyzes the marks, goods, and channels of trade only as set forth in the application and in the opposer’s registration, regardless of whether the actual usage of the marks by either party differs.” And because preclusion applies only when the usages adjudicated by the TTAB are “materially the same as those before the district court,” the Court acknowledged that it would not apply in “some or even many cases”—those in which the usages at issue in district court materially differ from those considered by the TTAB.
Many observers predicted that the practical effect of B & B would be minimal, since it would be a rare case in which the TTAB actually considered uses that were materially the same as those considered by the district court in subsequent litigation. And those predictions have largely proven accurate in the likelihood of confusion context. Indeed, aside from the Eighth Circuit’s decision on remand in B & B itself, courts have refused to give TTAB likelihood of confusion determinations preclusive effect.
Courts have given TTAB decisions preclusive effect in other contexts, however. In Ashe v. PNC Financial Services Group, Inc., the court gave preclusive effect to the TTAB’s priority determination, noting that it had “previously found that the issue of priority determined in an earlier proceeding before the Trademark Board ‘is identical to priority of use in an infringement suit.’” The Eastern District of Virginia also gave preclusive effect to a TTAB decision involving fraud on the PTO. According to the court, “[t]he issue of whether Defendant committed fraud on the USPTO [was] identical to the question considered by the TTAB in the prior proceeding.” And all of the predicate conditions for preclusion applied: “the determination of fraud was critical and necessary to the TTAB’s final decision,” “[t]he TTAB’s decision [was] considered final and valid,” and “Defendant was represented by counsel before the TTAB, and had a full and fair opportunity to litigate the issue of fraud in the prior proceeding.”
In Vaad L’Hafotzas Sichos, Inc. v. Kehot Publication Society, the court suggested, though only in dicta, that a TTAB determination of trademark validity would have preclusive effect. That case dealt primarily with likelihood of confusion, but in a prior opinion involving the same parties, the court held that Merkos was the owner of the Kehot logo, “which [was] indeed a trademark.” The court noted that no new evidence was presented in the later action “to countermand the Court’s determination that there was substantial evidence to support that conclusion.” Even if such evidence had been presented, the court continued, “it [was] likely the TTAB’s decision [rejecting Vaad’s opposition] would have had preclusive effect.” Ultimately, the court did not believe it was necessary to decide the preclusion issue, however, “because the Court’s disposition of the case would remain unchanged.”
Courts have otherwise read B & B relatively narrowly, in particular by concluding that B & B applies only to inter partes proceedings and not to ex parte registration decisions. In In re Cordua Restaurants, Inc., the Federal Circuit affirmed the TTAB’s affirmance of the Trademark Office’s refusal to register a stylized version of the word CHURRASCOS.
The Trademark Office rejected the application on the ground that the mark was generic, despite the fact that the applicant owned prior registration of the same word mark in standard characters. The Board affirmed that refusal, and the Federal Circuit specifically held that “Cordua’s existing registration of the CHURRASCOS word mark (the ’321 Registration) d[id] not preclude a finding that the stylized form of the mark is generic.”
The Federal Circuit also rejected Cordua’s argument that “the examiner’s determination that the CHURRASCOS word mark is trademark-eligible is binding on this case as a matter of issue preclusion.” According to the court, while TTAB decisions in inter partes opposition proceedings may have preclusive effect, “there [was] no suggestion in B & B Hardware that an examiner’s decision to register a mark or to refuse registration satisfies the traditional requirements of issue preclusion.” Thus, the Trademark Office’s prior decision to register the standard character mark was not preclusive against the Office in the context of the application to register the stylized version of the mark.
Courts have become more deeply divided over the last few years regarding the requirements for injunctive relief in trademark cases. Until relatively recently, most courts maintained that likelihood of confusion presumptively caused irreparable harm. That old rule was created at a time when trademark infringement focused on cases of competing goods, and it probably made good sense in that context. Use of a mark by direct competitors implicated both producer and consumer interests, threatening both to divert sales from the mark owner and to defraud consumers. As trademark law has expanded over the last several decades, however, it is substantially less clear how the various activities trademark law reaches harm either mark owners or consumers. And while courts have tended to accept modern harm stories as a general matter, some courts’ doubts about their validity influenced the ways they think about irreparable harm and the availability of injunctive relief.
In eBay Inc. v. MercExchange, L.L.C., the Supreme Court rejected a presumption of injunctive relief in patent cases, holding that the typical equitable principles “apply with equal force to disputes arising under the Patent Act.” In order to obtain an injunction:
A plaintiff must demonstrate: (1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction.
eBay has had some meaningful effect in certain categories of patent cases, particularly those involving non-practicing entities, and every court that has considered the question has recognized eBay’s applicability in trademark law as well. But courts have reached different conclusions about the extent of eBay’s effect in trademark cases.
In Herb Reed Enterprises, LLC v. Florida Entertainment Management, Inc., the Ninth Circuit recognized eBay’s applicability to trademark cases, holding that just as “[n]othing in the Patent Act indicate[d] that Congress intended such a departure” from standard equitable principles in patent cases, neither did anything in the Lanham Act indicate such a departure in trademark cases. According to the Ninth Circuit, irreparable harm cannot be presumed; a plaintiff must provide evidence of irreparable harm, and it must do so with more than “cursory and conclusory” statements. “Gone are the days when once the plaintiff in an infringement action has established a likelihood of confusion, it is ordinarily presumed that the plaintiff will suffer irreparable harm if injunctive relief does not issue.”
District courts in the Ninth Circuit have tended to read Herb Reed broadly, and several have denied injunctive relief in cases in which irreparable injury once would have been presumed. In Café Foundation, Inc. v. Seeley, for example, the court emphasized that “past consumer confusion does not demonstrate a likelihood of future harm.” While “evidence of loss of control over business reputation and damage to goodwill can support a finding of irreparable harm,” the claim that the plaintiff “could suffer harm in the form of damaged goodwill is nothing more than an unsupported and conclusory statements [sic] regarding harm [Plaintiff] might suffer.” Similarly, the court denied preliminary injunctive relief in Hakkasan LV, LLC v. Miller (a cybersquatting case) because the plaintiff could not demonstrate irreparable harm from the defendant’s registration of domain names that were similar to the plaintiff’s. In particular, plaintiff could not “even quantify the number of internet users who ha[d] mistakenly used Miller’s services and not Hakkasan’s, or who w[ould] refuse to visit Hakkasan’s Las Vegas nightclub due to confusion between Chakkasan.comD and Chakkasanlv.comD and Miller’s Contested Domain Names.” Nor was there any “evidence that [the defendant] ha[d] taken any steps to compete with [plaintiff’s] business beyond registering the Contested Domain Names and offering them for sale.”
The Third Circuit has followed the Ninth Circuit’s lead, emphasizing the need to “demonstrate that irreparable harm is likely” in Lanham Act cases. And some district courts in the Third Circuit have denied injunctive relief due to lack of irreparable harm. In 7-Eleven, Inc. v. Sodhi, for example, the court refused to grant injunctive relief even though the plaintiff proved likelihood of success on the merits of its trademark infringement claim because plaintiff could not show that irreparable harm was likely.
But district courts in the Third Circuit do not seem universally to have given up on the idea that evidence of consumer confusion establishes irreparable harm. In Astrazeneca AB v. Camber Pharmaceuticals, Inc., for example, the court found that the defendant’s use of the color purple for its generic pills would “create (and [was] intended to create) the false impression that its generic esomeprazole magnesium capsules are identical to Nexium®, not merely bioequivalent, and may be an ‘authorized generic,’ that is, a generic drug made or authorized by the brand name company.” On that basis, the court concluded that the defendant’s conduct put “at risk [Astrazeneca]’s reputation” and that [Astrazeneca] had therefore “demonstrated a likelihood of irreparable harm.”
Some other courts have affirmed eBay’s application to trademark law without clearly ruling on the question of whether likelihood of consumer confusion will continue to establish irreparable harm. In JL Powell Clothing LLC v. Powell, for example, the First Circuit remanded the case to the district court, noting that “[t]here ha[d] also been no addressing of the plaintiffs’ likelihood of success on its related intellectual property claim to the trademark ‘J.L. Powell’—and so no express finding on customer confusion (or the difficulty of establishing the damage from any such confusion)—that might support a finding of irreparable harm.” The parenthetical is notable for the ambiguity it adds regarding eBay’s effect, and it echoes earlier First Circuit cases that acknowledged the possibility that eBay had called into doubt the viability of a presumption of irreparable harm but that avoided ruling directly on the question.
The Eleventh Circuit is quite unclear. In Hoop Culture, Inc. v. GAP Inc., the court acknowledged that eBay had called into question the circuit’s previous practice of presuming irreparable harm once the plaintiff in a trademark case established a likelihood of success on the merits. The court further emphasized that a plaintiff seeking an injunction must show that irreparable harm is “likely, not merely possible.” Yet the court expressly disclaimed making any “express holding about the effect of eBay on this specific case or [the] circuit’s presumption of irreparable harm more generally.” And it left open the possibility that irreparable harm might be presumed in some cases: “All of this is not to say that a presumption of irreparable harm or something like it will never be an appropriate exercise of the district court’s equitable discretion.” In the particular case, however, the court found that “the district court’s findings of facts were sufficient to rebut any presumption of irreparable harm that may have applied,” so an injunction was unwarranted regardless of eBay’s effect.
Neither the Second nor the Eighth Circuit has yet addressed the issue, but district courts in those circuits seem to be moving towards abolishing the presumption. In Ann Clark, Ltd. v. R & M International, Corp., for example, the District of Vermont noted that “the presumption of irreparable injury [when likelihood of confusion is shown] is no longer in effect.” And in JDR Industries, Inc. v. McDowell, the court stated, “[t]he Court does not presume irreparable harm based solely on its finding of likely (or actual) confusion.”
In other circuits, eBay seems not to have changed very much. Thus far, no court in the Fourth, Fifth, or Sixth Circuits has denied injunctive relief on the ground that the plaintiff could not establish irreparable harm despite evidence of confusion. The Fifth Circuit has even explicitly reaffirmed the presumption of irreparable harm: “All that must be proven to establish liability and the need for an injunction against infringement is the likelihood of confusion—injury is presumed.”
V. Nominative Fair Use
The concept of nominative fair use is primarily associated with the Ninth Circuit, which developed the terminology in New Kids on the Block. Though no circuit has affirmatively rejected the doctrine, most circuits have not expressly accepted it either. And since the Third Circuit’s decision in Century 21 Real Estate Corp. v. Lendingtree, Inc., the law has been divided in terms of which party bears the burden on the nominative fair use question and the relationship of nominative fair use to the likelihood of confusion factors. The Second Circuit’s recent decision in International Information Systems Security Certification Consortium, Inc. v. Security University, LLC deepened that division.
In its New Kids decision, the Ninth Circuit held that use of another’s mark is not actionable when it meets the following three conditions:
First, the product or service in question must be one not readily identifiable without use of the trademark; second, only so much of the mark or marks may be used as is reasonably necessary to identify the product or service; and third, the user must do nothing that would, in conjunction with the mark, suggest sponsorship or endorsement by the trademark holder.
Though it seems clear that the New Kids court meant the third factor to refer to additional conduct beyond mere use of the mark that would suggest sponsorship or affiliation, courts applying New Kids often reduced the third factor to a confusion inquiry. Indeed, the Ninth Circuit held that the New Kids factors replaced the likelihood of confusion factors and that a defendant arguing nominative fair use had the burden of proving that the use would not cause confusion.
Lendingtree rejected the Ninth Circuit’s approach of substituting the nominative fair use analysis for evaluation of the likelihood of confusion factors. It did so in part because the Ninth Circuit’s approach had the effect of shifting to the defendant the burden of proving lack of confusion, which the Third Circuit believed was in tension with the Supreme Court’s decision in KP Permanent. The Third Circuit therefore emphasized that, in a nominative fair use case, the plaintiff retained the burden of first proving that confusion was likely. But the court also acknowledged that many of the confusion factors would not be well-suited to a nominative fair use case, and it therefore focused on only four of the factors as relevant in such a case.
In Toyota Motor Sales, U.S.A., Inc. v. Tabari, the Ninth Circuit maintained that the New Kids factors replace the likelihood of confusion factors in a nominative fair use case, and it faulted the district court for treating nominative fair use as an affirmative defense to be evaluated after the plaintiff established likelihood of confusion under the Sleekcraft factors. But to avoid the burden-shifting problem, Tabari made clear that, because the New Kids factors replace the likelihood of confusion factors, the plaintiff bears the burden of proving that the defendant’s use is not nominative fair use.
The Second Circuit deviated from both of those approaches in International Information Systems, specifically rejecting Century 21 v. Lendingtree and holding that nominative fair use is not an affirmative defense to a trademark infringement claim. According to the Second Circuit, the statutory fair use defense does not encompass nominative fair use because nominative use is “not the use of a name, term, or device otherwise than as a mark which is descriptive of and used merely to describe the goods or services of the alleged infringer.” Nominative use instead “involves using the mark at issue as a mark to specifically invoke the mark-holder’s mark, rather than its use, other than as a mark, to describe the alleged infringer’s goods or services.” And because Congress did not make such use a defense to trademark infringement, nominative fair use cannot be deemed an affirmative defense.
Given that (non)status, the court held, the nominative fair use factors cannot supplant the likelihood of confusion factors. Instead, district courts are to consider the Third Circuit and Ninth Circuit’s nominative fair use factors in addition to the Polaroid factors. Indeed, the Second Circuit specifically reiterated that, when considering the nominative fair use factors, courts must not “consider only source confusion, but rather must consider confusion regarding affiliation, sponsorship, or endorsement by the mark holder.”
VII. Initial Interest Confusion
The Ninth Circuit’s recent decision in Multi Time Machine, Inc. v. Amazon.com, Inc., reflects the continuation of initial interest confusion’s slow death. For the last several years, courts in jurisdictions that had previously embraced initial interest confusion have been cutting the doctrine back, primarily by emphasizing labeling and context and refusing to assume that mere searching indicates confusion. In its 1-800 Contacts decision, for example, the Tenth Circuit rejected an initial interest confusion claim based on use of “1-800 CONTACTS” and related terms as AdWords (advertisement program offered by Google) keywords. According to the court, even initial interest confusion was unlikely because the advertisements shown on the search results page were clearly labeled as such, and those ads unambiguously identified the source of the ads. The Ninth Circuit also had previously emphasized the labeling of ads in its Network Automation decision. There the court held that use of the plaintiff’s mark as a keyword was not likely to cause confusion in large part because of “the labeling and appearance of the advertisements and the surrounding context on the screen displaying the results page.”
These cases did not entirely close the door on initial interest confusion, as both of them implied that a claim could still be pursued against the user of a keyword whose advertisements were not sufficiently clearly labeled, even if any confusion that was caused by the ads might be dispelled when users clicked on them. But the increasing focus on the context of the search results page (rather than assumptions about users’ intent when entering search terms) was moving initial interest confusion closer to point-of-sale confusion and away from the mere use of keywords.
MTM v. Amazon doubled down on that contextual approach, noting that the labeling and appearance of the products for sale on Amazon’s web page was the most important factor in assessing initial interest confusion. In that case, MTM sued Amazon for trademark infringement for generating a search results page in response to searches for “mtm special ops”—MTM’s registered trademark for watches. Amazon, which did not sell MTM watches, displayed the following page:
The Ninth Circuit recognized that the multi-factor likelihood of confusion test was not “particularly apt” in a case like this.
[T]he Sleekcraft test was developed for a different problem—i.e., for analyzing whether two competing brands’ marks are sufficiently similar to cause consumer confusion. Although the present case involves brands that compete with MTM, such as Luminox, Chase–Durer, TAWATEC, and Modus, MTM does not contend that the marks for these competing brands are similar to its trademarks.
Unlike a standard trademark case, MTM argued that the search page, which displayed the “mtm special ops” search term, was likely to confuse consumers because it did not explicitly inform searchers that Amazon does not carry MTM watches. Thus, the court focused primarily on the nature of the goods at issue (here, expensive watches) and the type of consumer.
Like the Tenth Circuit in 1-800 Contacts, the court in MTM v. Amazon was convinced that the labeling and appearance of the search results page was the most important factor in the case. As the Ninth Circuit had previously indicated, labeling can eliminate the likelihood of initial interest confusion in cases involving Internet search terms. And here the labeling was clear: “Because Amazon clearly labels each of the products for sale by brand name and model number accompanied by a photograph of the item, it is unreasonable to suppose that the reasonably prudent consumer accustomed to shopping online would be confused about the source of the goods.”
Indeed, the labeling was sufficient to prevent even initial interest confusion:
MTM argues that initial interest confusion might occur because Amazon lists the search term used—here the trademarked phrase “mtm special ops”—three times at the top of the search page. MTM argues that because Amazon lists the search term “mtm special ops” at the top of the page, a consumer might conclude that the products displayed are types of MTM watches. But, merely looking at Amazon’s search results page shows that such consumer confusion is highly unlikely. None of these watches is labeled with the word “MTM” or the phrase “Special Ops,” let alone the specific phrase “MTM Special Ops.” Further, some of the products listed are not even watches. The sixth result is a book entitled “Survive!: The Disaster, Crisis and Emergency Handbook by Jerry Ahem.” The tenth result is a book entitled “The Moses Expedition: A Novel by Juan Gómez-Jurado.” No reasonably prudent consumer, accustomed to shopping online or not, would assume that a book entitled “The Moses Expedition” is a type of MTM watch or is in any way affiliated with MTM watches. Likewise, no reasonably prudent consumer accustomed to shopping online would view Amazon’s search results page and conclude that the products offered are MTM watches.
The court also specifically rejected MTM’s argument that, in order to eliminate the likelihood of confusion, Amazon needed to “change its search results page so that it explain[ed] to customers that it d[id] not offer MTM watches for sale before suggesting alternative watches to the customer.” In the court’s view, “[t]he search results page makes clear to anyone who can read English that Amazon carries only the brands that are clearly and explicitly listed on the web page. The search results page is unambiguous—not unlike when someone walks into a diner, asks for a Coke, and is told ‘No Coke. Pepsi.’”
© 2017 Mark P. McKenna and Shelby Niemann. Individuals and nonprofit institutions may reproduce and distribute copies of this Essay in any format, at or below cost, for educational purposes, so long as each copy identifies the author, provides a citation to the Notre Dame Law Review Online, and includes this provision in the copyright notice.
* Mark P. McKenna is Associate Dean for Faculty Development, Professor of Law, and Notre Dame Presidential Fellow, Notre Dame Law School. Shelby Niemann is a J.D. Candidate at Notre Dame Law School, Class of 2018.
 The Supreme Court granted certiorari in Lee v. Tam on September 29, 2016. See In re Tam, 808 F.3d 1321 (Fed. Cir. 2015), cert. granted Lee v. Tam, 137 S. Ct. 30 (U.S. Sept. 19, 2016) (No. 15-1293).
 In re Tam, 808 F.3d at 1327, 1358. The Federal Circuit took the case sua sponte en banc after a panel of the court affirmed the Trademark Office’s rejection of the mark but cast doubt on the constitutionality of section 2(a). See In re Tam, 600 F. App’x 775 (Fed. Cir. 2015) (per curiam) (ordering that the case be heard en banc); In re Tam, 785 F.3d 567 (Fed. Cir. 2015).
 In re Tam, 808 F.3d at 1330 n.1; see In re McGinley, 660 F.2d 481, 484 (C.C.P.A. 1981) (holding section 2(a) constitutional on the ground that denial of registration does not proscribe any conduct or suppress any tangible form of expression).
 See Walker v. Sons of Confederate Veterans, 135 S. Ct. 2239, 2246 (2015) (holding that Texas specialty license plate designs were government speech, and the state board therefore did not violate a nonprofit organization’s free speech rights by denying its application for a design with a Confederate flag); see also Nat’l Endowment for the Arts v. Finley, 524 U.S. 569, 588–89 (1998); Rust v. Sullivan, 500 U.S. 173, 194 (1991).
 See, e.g., Agency for Int’l Dev. v. All. for Open Soc’y Int’l, Inc., 133 S. Ct. 2321, 2330, 2332 (2013) (invalidating the federal government’s attempt to limit a funding program for efforts to stop the spread of HIV and AIDS to organizations that have explicit policies against prostitution and sex trafficking on the ground that “[b]y demanding that funding recipients adopt—as their own—the Government’s view on an issue of public concern,” the government had impermissibly leveraged its funds to influence speech “outside the scope of the federally funded program” (quoting Rust, 500 U.S. at 197)).
[A]ny word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which . . . is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person.
15 U.S.C. § 1125(a)(1) (2012).
 In re Tam, 808 F.3d 1321, 1344 (Fed. Cir. 2015) (suggesting that “it is unclear whether Mr. Tam could actually enforce any common law rights to a disparaging mark,” and calling the government’s suggestion that Tam could fall back on common law rights “illusionary”). The earlier panel decision had been less equivocal, stating flatly that unregistrable marks are not protectable under federal or state law. See In re Tam, 785 F.3d 567, 576 (Fed Cir. 2015) (“Not only is a disparaging trademark denied federal registration, but it cannot be protected by its owner by virtue of a § 43(a) unfair competition claim.” (citing 15 U.S.C. § 1125(a))); see also id. at 577 (“[N]ot only are the benefits of federal registration unavailable to Mr. Tam, so too are the benefits of trademark registration in nearly all states.”).
 Rebecca Tushnet makes this argument in her excellent piece on registration. See Rebecca Tushnet, Registering Disagreement: Registration in Modern American Trademark Law, 130 Harv. L. Rev. 867 (2017).
 The Federal Circuit expressly limited its decision to the constitutionality of the disparagement provision, though it is hard to imagine the rest of section 2(a) would survive under the court’s analysis. See In re Tam, 808 F.3d at 1330 n.1 (“We limit our holding in this case to the constitutionality of the § 2(a) disparagement provision. Recognizing, however, that other portions of § 2 may likewise constitute government regulation of expression based on message, such as the exclusions of immoral or scandalous marks, we leave to future panels the consideration of the § 2 provisions other than the disparagement provision at issue here.”).
 Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 831 (1995). The distinction seems to matter only in cases involving public forums, so unless the Court is prepared to hold that the registration system creates a limited public forum, this distinction would seem out of place here.
 The Federal Circuit stated that “[t]he legal significance of viewpoint discrimination is the same whether the government disapproves of the message or claims that some part of the populace will disapprove of the message.” Id. at 1336. According to the Federal Circuit, “[t]his point is recognized in the Supreme Court’s long-standing condemnation of government impositions on speech based on adverse reactions among the public.” Id. (first citing Snyder v. Phelps, 562 U.S. 443, 460–61 (2011); then citing R.A.V. v. City of St. Paul, 505 U.S. 377 (1992); and then citing Texas v. Johnson, 491 U.S. 397, 414 (1989)). Of those cases, R.A.V. is the most analogous. That case involved an ordinance that prohibited the display of a symbol that one knows or has reason to know “arouses anger, alarm or resentment in others on the basis of race, color, creed, religion or gender.” R.A.V., 505 U.S. at 380 (quoting St. Paul, Minn., Code of Ordinances § 292.02 (1990)). The Court found the ordinance unconstitutional because it was content-based, but its focus was on the fact that the ordinance prohibited speech that produced an effect (arousing anger, alarm or resentment) only when the effect was due to race, color, creed, religion, or gender. Id. at 391 (“Displays containing abusive invective, no matter how vicious or severe, are permissible unless they are addressed to one of the specified disfavored topics.”). But note that the Court juxtaposed those kinds of regulations of “fighting words” with more categorical prohibitions that are more analogous to section 2(a):
Displays containing some words—odious racial epithets, for example—would be prohibited to proponents of all views. But “fighting words” that do not themselves invoke race, color, creed, religion, or gender—aspersions upon a person’s mother, for example—would seemingly be usable ad libitum in the placards of those arguing in favor of racial, color, etc., tolerance and equality, but could not be used by those speakers’ opponents.
Id. Contra Boos v. Barry, 485 U.S. 312, 315, 321 (1988) (holding that a law prohibiting display of any sign within five hundred feet of a foreign embassy, if the sign would tend to bring that foreign government into “public odium” or “disrepute,” was content-based but not viewpoint-based).
 See, e.g., Riley v. Nat’l Fed’n of the Blind of N.C., Inc., 487 U.S. 781, 796 (1988) (“[W]here, as here, the component parts of a single speech are inextricably intertwined, we cannot parcel out the speech, applying one test to one phrase and another test to another phrase. Such an endeavor would be both artificial and impractical.”).
 The 2014 Trademark Year in Review, like this Year in Review, was originally prepared for the Houston Intellectual Property Law Association and University of Houston Institute for Intellectual Property & Information Law’s annual Fall Institute. See Mark P. McKenna, Trademark Year in Review (Feb. 5, 2015) (unpublished manuscript), https://ssrn.com/abstract=2645863.
 Arbitrary or fanciful terms as applied to particular products were considered technical trademarks, and they were protected in actions for trademark infringement. See Restatement (Third) of Unfair Competition § 9 cmt. g (Am. Law Inst. 1995); see also 1 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 4:4 (4th ed. 2016) (defining technical trademarks as marks that were “fanciful, arbitrary, distinctive, non-descriptive in any sense and not a personal name” (first citing Milton Handler & Charles Pickett, Trade-marks and Trade Names—An Analysis and Synthesis: I, 30 Colum. L. Rev. 168, 169 (1930); and then citing G.W. Cole Co. v. Am. Cement & Oil Co., 130 F. 703 (7th Cir. 1904))). Trade names included surnames, geographic terms, and descriptive terms—what we would now refer to as non-inherently distinctive designations. McCarthy, supra, § 4:5. Federal registration also was limited to marks used in interstate commerce, or commerce with foreign nations or Indian tribes. See Trademark Act of 1905, ch. 592, § 1, 33 Stat. 724, 724 (“[T]he owner of a trade-mark used in commerce with foreign nations, or among the several States, or with Indian tribes . . . may obtain registration. . . .”).
 See Trademark Act of 1905, § 16 (creating cause of action for infringement of registered marks); see also id. § 17 (granting federal courts jurisdiction in cases involving use of a registered mark in interstate commerce). The requirement that the defendant’s use affect interstate commerce was, formerly, a real limitation. See U.S. Printing & Lithograph Co. v. Griggs, Cooper & Co., 279 U.S. 156, 157–59 (1929) (refusing the plaintiff’s claim under the 1905 Act that the defendant infringed its registered “Home Brand” trademark, which the plaintiff used in “certain named States of the northwest” by printing and selling labels containing the word “Home,” on the ground that the defendant’s activities took place only within the limits of the state of Ohio and did not interfere with interstate commerce).
 McKenna, supra note 25, at 291. Not every mark eligible to be registered was in fact registered. Consequently, although all federal claims were trademark infringement claims, not all trademark infringement claims were federal claims. There were also common law trademark infringement claims, brought to vindicate exclusive rights in unregistered trademarks, though many courts conflated those claims with unfair competition claims—seeing trademark infringement claims as one category consisting of claims under the federal statute, and common law unfair competition claims as the other category consisting of any claim not based on a registered mark. Id. at 290–96.
 See Edward S. Rogers, Introduction to Daphne Robert, The New Trade-Mark Manual: A Handbook on Protection of Trade-Marks in Interstate Commerce, at xvi–xxi (1947) (arguing that section 44(h) of the statute should be understood to give courts the ability to develop a uniform body of federal unfair competition law).
 See Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 29–30 (2003) (“Although a case can be made that a proper reading of § 43(a), as originally enacted, would treat the word ‘origin’ as referring only ‘to the geographic location in which the goods originated,’ the Courts of Appeals considering the issue, beginning with the Sixth Circuit, unanimously concluded that it ‘does not merely refer to geographical origin, but also to origin of source or manufacture,’ thereby creating a federal cause of action for traditional trademark infringement of unregistered marks.” (footnote omitted) (first quoting Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 777 (1992) (Stevens, J., concurring); and then quoting Federal-Mogul-Bower Bearings, Inc. v. Azoff, 313 F.2d 405, 408 (1963))).
 See Two Pesos, 505 U.S. at 768 (“[I]t is common ground that § 43(a) protects qualifying unregistered trademarks and that the general principles qualifying a mark for registration under § 2 of the Lanham Act are for the most part applicable in determining whether an unregistered mark is entitled to protection under § 43(a).” (citations omitted)).
 See Belmora LLC v. Bayer Consumer Care AG, 819 F.3d 697, 709 (4th Cir. 2016) (citing Lamparello v. Falwell, 420 F.3d 309, 313 (4th Cir. 2005) (“Both infringement [under § 32] and false designation of origin [under § 43(a)] have [the same] five elements.” (alterations in original))); see also Int’l Bancorp, LLC v. Societe des Bains de Mer et du Cercle des Estrangers a Monaco, 329 F.3d 359, 362 n.2 (4th Cir. 2003) (“[T]he tests for trademark infringement and unfair competition . . . are identical.”); People for the Ethical Treatment of Animals v. Doughney, 263 F.3d 359, 364 (4th Cir. 2001) (using same test for both causes of action); Lone Star Steakhouse & Saloon v. Alpha of Va., Inc., 43 F.3d 922, 930 (4th Cir. 1995) (“[T]o prevail under §§ 32(1) and 43(a) of the Lanham Act for trademark infringement and unfair competition, respectively, a complainant must demonstrate that it has a valid, protectible [sic] trademark . . . .”).
 Grupo Gigante SA De CV v. Dallo & Co., 391 F.3d 1088, 1093 (9th Cir. 2004) (alteration in original) (footnote omitted) (quoting 5 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 29:2 (4th ed. 2002)); see also Buti v. Perosa, S.R.L., 139 F.3d 98, 103–05 (2d Cir. 1998); Person’s Co. v. Christman, 900 F.2d 1565, 1569–70 (Fed. Cir. 1990); Fuji Photo Film Co. v. Shinohara Shoji Kabushiki Kaisha, 754 F.2d 591, 599 (5th Cir. 1985).
 American courts have tended to refer to the doctrine as the “famous marks” doctrine, though it is known internationally as the “well-known marks” doctrine, originating as it did with an addition to Article 6bis of the Paris Convention. Paris Convention for the Protection of Industrial Property, art. 6bis, Mar. 20, 1883, as revised at Stockholm, July 14, 1967, 21 U.S.T. 1583, 828 U.N.T.S. 305 (requiring member states “ex officio if their legislation so permits, or at the request of an interested party, to refuse or to cancel the registration, and to prohibit the use, of a trademark which constitutes a reproduction, an imitation, or a translation, liable to create confusion, of a mark considered by the competent authority of the country of registration or use to be well known in that country as being already the mark of a person entitled to the benefits of this Convention and used for identical or similar goods” (emphasis added)).
 Compare Grupo Gigante, 391 F.3d at 1094 (holding that there is a “famous-mark exception” to the territoriality principle), with ITC Ltd. v. Punchgini, Inc., 482 F.3d 135, 159–65 (2d Cir. 2007) (refusing to recognize a famous marks doctrine as a matter of federal law); see also ITC Ltd. v. Punchgini, Inc., 880 N.E.2d 852, 859 (N.Y. 2007) (answering a certified question from the Second Circuit regarding a New York law and holding that, though New York does not recognize a famous marks doctrine as such, “when a business, through renown in New York, possesses goodwill constituting property or a commercial advantage in this state, that goodwill is protected from misappropriation under New York unfair competition law”).
 Grupo Gigante, 391 F.3d at 1098 (holding that, for the famous-mark exception to apply, secondary meaning is not enough, because “[i]n addition, where the mark has not before been used in the American market, the court must be satisfied, by a preponderance of the evidence, that a substantial percentage of consumers in the relevant American market is familiar with the foreign mark”).
 There’s also an interesting, and apparently unrecognized, tension here with the Federal Circuit’s suggestion in Tam that unregistrable marks are not protectable. Marks that are never used in interstate commerce in the United States are clearly not registrable, so under the Federal Circuit’s formulation, they would not be eligible for protection under section 43(a), even if merely unregistered marks would be.
 Id. Because of the unique setting of registration proceedings, Hargis bore the burden of persuasion at the TTAB, but infringement plaintiffs like B & B always bear the burden of persuasion in civil litigation. See KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 118 (2004) (“Section 1115(b) places a burden of proving likelihood of confusion (that is, infringement) on the party charging infringement even when relying on an incontestable registration.”).
 Id. at 1307 (citing Brief for the United States as Amicus Curiae Supporting Petitioner at 23, B & B Hardware, Inc. v. Hargis Indus., Inc., 123 S. Ct. 1293 (2015) (No. 13-352)); see Cunningham v. Laser Golf Corp., 222 F.3d 943, 950 (Fed. Cir. 2000) (refusing to consider evidence that the goods the parties sold in the marketplace were not identical, because “the Board must look to the registrations themselves to determine the scope of the goods covered by the contested mark” (citations omitted)); Octocom Sys., Inc. v. Hous. Comput. Servs., Inc., 918 F.2d 937, 942 (Fed. Cir. 1990) (“The authority is legion that the question of registrability” is decided in the registration context on the basis of the parties’ registration submissions “regardless of what the record may reveal as to the particular nature of an applicant’s goods, the particular channels of trade or the class of purchasers to which sales of the goods are directed.” (citations omitted)); see also U.S. Patent & Trademark Office, supra note 17, § 1207.01 (stating that the focus of ex parte examination is on use described in application). For a more complete description of the differences between PTO proceedings and infringement analysis in district court, see Tushnet, supra note 15, at 892–99.
 See, e.g., CSL Silicones Inc. v. Midsun Grp. Inc., 170 F.3d 304, 319 (D. Conn. 2016) (not giving TTAB likelihood of confusion finding preclusive effect because civil litigation would involve a different set of transactional facts); see also Paleteria La Michoacana, Inc. v. Productos Lacteos Tocumbo S.A. DE C.V., 188 F. Supp. 3d 22, 37 (D.D.C. 2016) (citing Kappos v. Hyatt, 132 S. Ct. 1690, 1701 (2012) (holding that district courts reviewing TTAB decisions make findings of fact de novo, and that it was unable to “meaningfully defer” to TTAB factual findings in this particular case “because the TTAB, in many respects, considered a ‘different set of facts’ [regarding likelihood of confusion] than has been presented here”)).
 See Progressive Distrib. Servs., Inc. v. United Parcel Servs., Inc., 186 F. Supp. 3d 741, 749 (W.D. Mich. 2016), appeal docketed, No. 16-1830 (6th Cir. June 17, 2016) (“Because the USPTO’s rejection of UPS’s mark was a low-level determination, and there is no indication in the record that the examining attorney reviewed the evidence presented to this Court in the instant case, this Court declines to give any weight to the PTO examiner’s likelihood of confusion determination.”).
 Id. (citations omitted); see also Progressive Distrib. Servs., Inc., 186 F. Supp. 3d at 749 (declining, without even citing B & B, to give any weight to the Trademark Examiner’s likelihood of confusion determination, because “the USPTO’s rejection of UPS’s mark was a low-level determination, and there is no indication in the record that the examining attorney reviewed the evidence presented to this Court in the instant case”).
 In re Cordua, 823 F.3d at 607. On the merits, that holding is difficult to square with the Trademark Office’s own rule that standard character marks include all stylizations. See U.S. Patent & Trademark Office, supra note 19, § 1207.01(c)(iii) (2017) (“A registrant is entitled to all depictions of a standard character mark regardless of the font style, size, or color, and not merely ‘reasonable manners’ of depicting such mark.” (citations omitted)). If the standard character mark is not generic, it is hard to imagine how a more distinctive version of that mark could be generic. But of course, it’s hard to square the standard rule with the idea that stylized versions of previously-registered standard character marks are registrable. If those standard character marks really include all stylizations, then a stylized version of that mark is not a different mark worthy of separate registration. For a discussion of the difficulty of standard character marks see Tushnet, supra note 15, at 882–84.
 See, e.g., McDonald’s Corp. v. Robertson, 147 F.3d 1301, 1310 (11th Cir. 1998); Pappan Enters., Inc. v. Hardee’s Food Sys., Inc., 143 F.3d 800, 805 (3d Cir. 1998); Tally-Ho, Inc. v. Coast Cmty. Coll. Dist., 889 F.2d 1018, 1029 (11th Cir. 1989); Processed Plastic Co. v. Warner Commc’ns, Inc., 675 F.2d 852, 858 (7th Cir. 1982).
 See Christopher B. Seaman, Permanent Injunctions in Patent Litigation after eBay: An Empirical Study, 101 Iowa L. Rev. 1949, 1949 (2016) (finding that “eBay has effectively created a bifurcated regime for patent remedies, as operating companies who compete against an infringer still obtain permanent injunctions in the vast majority of cases that are successfully litigated to judgment,” but non-practicing entities almost always are denied injunctive relief).
 Commentators are also somewhat divided about how eBay should be interpreted in the trademark context. See Mark A. Lemley, Did eBay Irreparably Injure Trademark Law?, 92 Notre Dame L. Rev. (forthcoming 2017), https://ssrn.com/abstract=2808677; Rebecca Tushnet, What’s the Harm of Trademark Infringement, 49 Akron L. Rev. 627 (2016).
 736 F.3d 1239, 1249 (9th Cir. 2013) (first alteration in original) (quoting eBay, 547 U.S. at 391–92); see also San Miguel Pure Foods Co. v. Ramar Int’l Corp., 625 F. App’x 322, 327 (9th Cir. 2015); Titaness Light Shop, LLC v. Sunlight Supply, Inc. 585 F. App’x 390, 391 (9th Cir. 2014).
 See Cafe Found., Inc. v. Seeley, No. 16-cv-0628, 2016 WL 1258624, at *7 (N.D. Cal. Mar. 31, 2016); Hakkasan LV, LLC v. Miller, No. 2:15-cv-0290, 2016 WL 1064473, at *4 (D. Nev. Mar. 17, 2016); Arcsoft, Inc. v. Cyberlink Corp., 153 F. Supp. 3d 1057, 1071 (N.D. Cal. 2015); Alto Velo Racing Club v. Rouleur Sports Grp., LLC, No. 5:15-cv-2144, 2015 WL 5462055, at *3 (N.D. Cal. Sept. 17, 2015); Sec. Alarm Fin. Enters., L.P. v. Alarm Prot. Tech., LLC, No. 3:13-cv-0102, 2015 WL 3459415, at *2 (D. Alaska June 1, 2015); Cutting Edge Solutions, LLC v. Sustainable Low Maint. Grass, LLC, No. 14-cv-2770, 2014 WL 5361548, at *7 (N.D. Cal. Oct. 20, 2014); Wells Fargo & Co. v. ABD Ins. & Fin. Servs., Inc., No. C 12-3856, 2014 WL 4312021, at *8 (N.D. Cal. Aug. 28, 2014); AFD China Intellectual Prop. Law (USA) Office, Inc. v. AFD China Intellectual Prop. Law Office, No. 3:09-cv-1509, 2014 WL 2619644, at *6 (D. Or. June 12, 2014).
 Ferring Pharm., Inc. v. Watson Pharm., Inc., 765 F.3d 205, 206 (3d Cir. 2014). Ferring involved a false advertising claim under section 43(a)(1)(B), 15 U.S.C. § 1125(a)(1)(B) (2012), but the court emphasized eBay’s applicability to all Lanham Act cases. Id. at 214.
 No. 13-3715, 2016 WL 541135, at *5 (D.N.J. Feb. 9, 2016) (noting that “Plaintiff relie[d] primarily on two forms of harm: harm to reputation as evidenced by customer complaints and harm to the property stemming from Defendants’ unauthorized occupation of the Stores,” and finding that, while these types of harm might support injunctive relief, “Plaintiff ha[d] not sufficiently demonstrated that these forms of harm ha[d] occurred or [were] likely to occur”); see also Congoo, LLC v. Revcontent LLC, No. 16-401, 2016 WL 1547171, at *4 (D.N.J. Apr. 15, 2016); Dunkin Donuts Franchising LLC v. Claudia III, LLC, No. 14-2293, 2014 WL 3900569, at *5 (E.D. Pa. Aug. 11, 2014); Buzz Bee Toys, Inc. v. Swimways Corp., 20 F. Supp. 3d 483, 510 (D.N.J. 2014).
 For a thorough survey of courts’ approaches to eBay, see Peter J. Karol, Trademark’s eBay Problem, 26 Fordham Intell. Prop. Media & Ent. L.J. 625 (2016). Karol puts circuits (other than the Ninth and Third) into three groups: “Noncommittal or Equivocating Circuits (First, Tenth, and Eleventh Circuits),” “Circuits Endorsing Use of a Presumption after eBay (Fourth, Fifth, and Sixth Circuits),” and “Circuits that Have Not Directly Addressed the Applicability of eBay to Trademark Law but Where District Courts Have Attempted to Do So (Second, Seventh, and Eighth Circuits).” Id. at 637–51.
 See Swarovski Aktiengesellschaft v. Bldg. #19, Inc., 704 F.3d 44, 54 (1st Cir. 2013) (per curiam) (avoiding the question of whether irreparable harm can be presumed from confusion, but noting that “[w]hether or not the presumption of irreparable harm remains viable in this context, it is difficult to see how irreparable harm could be established without a finding of confusion”); Voice of the Arab World, Inc. v. MDTV Med. News Now, Inc., 645 F.3d 26, 31 (1st Cir. 2011) (“[I]t is unnecessary to decide at this time whether the rule relied upon by the district court (i.e., irreparable harm is presumed upon a finding of likelihood of success on the merits of a trademark infringement claim) is consistent with such principles, because—even if we assume without deciding that such rule is good law—we still find that the district court abused its discretion in applying such a presumption here.”).
 Id. The Eleventh Circuit had previously held that eBay was applicable to trademark cases but “decline[d] to decide whether the district court was correct in its holding that the nature of the trademark infringement gives rise to a presumption of irreparable injury.” N. Am. Med. Corp., 522 F.3d at 1228.
 Id. The Tenth Circuit has similarly avoided the eBay question, acknowledging the issue once but declining to address it. Lorillard Tobacco Co. v. Engida, 213 F. App’x 654, 657 (10th Cir. 2007) (“We need not consider how eBay may apply in this context, however, because in any event Lorillard has not shown that any harm Lorillard would suffer in the absence of an injunction outweighed the potential harm to I and G if an injunction were granted.”).
 No. 1:14-CV-143, 2014 WL 7392026, at *3 (D. Vt. Dec. 29, 2014) (alteration in original) (quoting Marks Org., Inc. v. Joles, 784 F. Supp. 2d 322, 334 (S.D.N.Y.2011)). Ann Clark suggested that Salinger v. Colting, 607 F.3d 68 (2d Cir. 2010), was the reason for the change. Salinger was a copyright case in which the court explicitly rejected its former rule that irreparable harm is presumed when a plaintiff proves infringement. Salinger, 607 F.3d at 74–75. It rejected that presumption on the ground it was inconsistent with eBay. Id. at 79.
 121 F. Supp. 3d 872, 891 (D. Neb. 2015) (first citing Ferring Pharm., Inc. v. Watson Pharm., Inc., 765 F.3d 205, 206 (3d Cir. 2014); and then citing Herb Reed Enters., LLC v. Fla. Entm’t Mgmt., Inc., 736 F.3d 1239 (9th Cir. 2013)); see also Plasti Dip Int’l Inc. v. Rust-Oleum Brands Co., No. 14-1831, 2014 WL 7183789, at *7 (D. Minn. Dec. 16, 2014).
 New Kids on the Block v. News Am. Publ’g, Inc., 971 F.2d 302, 308–09 (9th Cir. 1992). Nominative fair use’s only appearance in the Lanham Act is in section 43(c)(3)(A), exclusions from dilution liability. 15 U.S.C. § 1125(c)(3)(A) (2012).
 Id. at 222–23; see KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 121–22 (2004) (“Since the burden of proving likelihood of confusion rests with the plaintiff, and the fair use defendant has no free-standing need to show confusion unlikely, it follows (contrary to the Court of Appeals’s view) that some possibility of consumer confusion must be compatible with fair use, and so it is.”).
 Id. at 1245 (“Perhaps in the abstract, one who searches for a particular business with a strong mark and sees an entry on the results page will naturally infer that the entry is for that business. But that inference is an unnatural one when the entry is clearly labeled as an advertisement and clearly identifies the source, which has a name quite different from the business being searched for.”).
Consumers who go onto Amazon.com and search for the term “mtm special ops” are directed to a search results page. On the search results page, the search query used—here, “mtm special ops”—is displayed twice: in the search query box and directly below the search query box in what is termed a “breadcrumb.” The breadcrumb displays the original query, “mtm special ops,” in quotation marks to provide a trail for the consumer to follow back to the original search. Directly below the breadcrumb, is a “Related Searches” field, which provides the consumer with alternative search queries in case the consumer is dissatisfied with the results of the original search. Here, the Related Search that is suggested to the consumer is: “mtm special ops watch.” Directly below the “Related Searches” field is a gray bar containing the text “Showing 10 Results.” Then, directly below the gray bar is Amazon’s product listings. The gray bar separates the product listings from the breadcrumb and the “Related Searches” field.