Max Rack, Inc. v. Core Health & Fitness, LLC ( 2022 )


Menu:
  •                                RECOMMENDED FOR PUBLICATION
    Pursuant to Sixth Circuit I.O.P. 32.1(b)
    File Name: 22a0156p.06
    UNITED STATES COURT OF APPEALS
    FOR THE SIXTH CIRCUIT
    ┐
    MAX RACK, INC.,
    │
    Plaintiff-Appellee/Cross-Appellant,      │
    >        Nos. 20-3598/3600
    │
    v.                                                  │
    │
    CORE HEALTH & FITNESS, LLC,                                │
    Defendant-Appellant/Cross-Appellee.          │
    │
    ┘
    Appeal from the United States District Court for the Southern District of Ohio at Columbus.
    No. 2:16-cv-01015—Algenon L. Marbley, District Judge.
    Argued: July 28, 2021
    Decided and Filed: July 14, 2022
    Before: COLE, ROGERS, and MURPHY, Circuit Judges.
    _________________
    COUNSEL
    ARGUED: Robert Collings Little, BUCHALTER, Los Angeles, California, for
    Appellant/Cross-Appellee. Matthew J. Schonauer, SCHONAUER LAW LLC, Columbus, Ohio,
    for Appellee/Cross-Appellant. ON BRIEF: Robert Collings Little, C. Dennis Loomis,
    BUCHALTER, Los Angeles, California, for Appellant/Cross-Appellee. Matthew J. Schonauer,
    SCHONAUER LAW LLC, Columbus, Ohio, for Appellee/Cross-Appellant.
    MURPHY, J., delivered the opinion of the court in which ROGERS, J., joined, and
    COLE, J., joined in part. COLE, J. (pp. 35–37), delivered a separate opinion concurring in part
    and dissenting in part.
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 2
    _________________
    OPINION
    _________________
    MURPHY, Circuit Judge. Trademarks allow consumers to distinguish one product from
    another quickly and cheaply. But they could not perform this signaling function if a trademark
    owner’s competitors could freely use the owner’s mark to sell their (potentially inferior) goods.
    So the Lanham Act has long prohibited trademark infringement and given trademark owners a
    variety of remedies to combat it. See 
    15 U.S.C. §§ 1114
    (1), 1117(a), 1125(a)(1)(A). This case
    raises several challenging questions under the Lanham Act.
    Steve Skilken, the owner of Max Rack, Inc., invented a piece of gym equipment that he
    named the “Max Rack.” For years, his company sold Max Racks through a licensing agreement
    with Core Health & Fitness, LLC. When Max Rack’s last patent expired, however, Core Health
    decided to compete against Max Rack by selling an identical machine under a new name—the
    “Freedom Rack.” Max Rack alleged that Core Health committed two types of infringement
    during its transition to the Freedom Rack: it continued to sell “Max Racks” without
    authorization, and it attempted to sell Freedom Racks by free riding off the “Max Rack” name.
    A jury agreed, awarding Max Rack $1 million in damages and $250,000 in Core Health’s profits.
    The district court upheld the jury’s liability finding, doubled its profits award to $500,000, and
    granted attorney’s fees to Max Rack. But the court overturned Max Rack’s damages award.
    Both sides have appealed, and we must address several liability and remedy issues. All told, we
    affirm in part and reverse in part.
    I
    A
    Many avid weightlifters prefer to use free weights (think of a bench press with a standard
    barbell) over weightlifting machines (think of a chest-press machine) because free weights allow
    for more natural motions and help weightlifters develop better balance. Yet free weights also
    pose greater dangers. Weightlifting machines are designed to ensure that weights will come
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.               Page 3
    safely to rest when lifters can no longer hold them. If, by contrast, lifters without a “spotter” can
    no longer hold free weights, the weights can come crashing down and cause serious injury.
    Inventors thus have long sought to develop equipment that combines the freedom of free
    weights with the safety of weightlifting machines. Take a classic “Smith Machine”:
    Article, R.40-10, PageID 695. It attaches the ends of an Olympic-size barbell to two vertical
    “guide rods.” Skilken Tr., R.99, Page ID 2202. The guide rods are themselves connected to two
    sturdy vertical posts that stand eight or so feet in height and six or so feet apart. The barbell
    moves up and down along the guide rods. Safety latches on both sides of this barbell can
    connect to any in a series of paired hooks (or catches, depending on the model) placed from top
    to bottom on the vertical posts. The posts hold the barbell on these hooks when it is not in use.
    Weightlifters can perform exercises like squats with this machine by lifting the barbell off the
    posts and twisting it with their wrists to separate the barbell’s safety latches from the hooks.
    When they complete a set (or can no longer hold the weight), they need only twist the barbell
    until its latches reconnect with the closest pair of hooks. The vertical posts will again hold the
    barbell’s weight.
    Steve Skilken thought that he could make a machine with similar safety features but that
    better simulated the use of free weights. Born and raised in Columbus, Skilken was an athlete at
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 4
    The Ohio State University and continued to exercise after his graduation. While lifting at a local
    gym, a friend introduced him to a Smith Machine. Skilken soon came up with an idea to
    improve on its design. A Smith Machine allows the barbell to move only vertically (up and
    down), not horizontally (forward or backward). But when lifters use free weights to do exercises
    like lunges, the barbell will often move both up and down and forward and backward.
    In the late 1980s and early 1990s, Skilken spent significant time designing a machine that
    allowed a barbell to move both ways. His efforts resulted in the “Max Rack.” The patent for
    this machine contained a picture of its design:
    Patent, R.93-5, PageID 1522.         In later years after significant product development, a
    commercially available “Max Rack” looked like this:
    Website, R.93-3, PageID 1519.
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.               Page 5
    Like the Smith Machine, Skilken’s design connects each side of a standard barbell to two
    vertical guide rods. Unlike the Smith Machine, the Max Rack has two sets of vertical posts, one
    in the front and the other a few feet behind. The Max Rack connects the vertical guide rods
    attached to the barbell to horizontal guide rods that run along the top and bottom of the front and
    back posts. The vertical rods can move forward and backward along these horizontal rods. This
    design permits the barbell that holds the weights to move not just up and down along the vertical
    rods, but also forward and backward along the horizontal ones. The Max Rack thus allows lifters
    to perform more exercises with the barbell and creates a closer feel to traditional free weights.
    In 1997, Skilken’s company (which he also named “Max Rack”) received its first patent
    on the machine and registered the “Max Rack” trademark. By 2002, however, the Max Rack
    remained a relatively unknown piece of gym equipment. Kirt Moritz joined the company to take
    over marketing efforts. Four years later, the company had spent over $1 million developing and
    promoting the product. Among its promotional efforts, Max Rack advertised the machine on the
    internet, in magazines and trade journals, through late-night infomercials, and at trade shows.
    These efforts began to pay off. Professional sports teams (like the Green Bay Packers)
    and universities (like Ohio State) added Max Racks to their weight rooms. The sales of Max
    Racks also continued to increase from year to year. The company sold only 25 Max Racks in
    1999. By 2005, its annual sales had increased to 84. During these years, it sold a total of 365
    Max Racks.
    In 2005, Star Trac Strength, Inc., approached Skilken about teaming up to grow Max
    Rack sales. Early the next year, Max Rack and Star Trac entered into a licensing agreement.
    Star Trac received the exclusive right to make and distribute the Max Rack. In exchange, it
    agreed to pay Max Rack a royalty of $120 per unit sold and to take over marketing efforts. The
    agreement would continue until the expiration of the last Max Rack patent, at which point either
    party could opt out.
    Financial troubles hit Star Trac four years later. These troubles led the two companies to
    sign an “addendum” to their agreement in 2010. Max Rack agreed to accept a several-month
    delay in the payment of some $44,000 in royalties. Star Trac agreed to pay the $120 royalty in
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.              Page 6
    perpetuity (not just until the last patent’s expiration). Soon thereafter, Core Health bought Star
    Trac’s assets, including its licensing agreements. As part of its due diligence, Core Health
    reviewed this addendum. It did not believe that courts would enforce the requirement to pay
    royalties forever because of caselaw barring patent holders from seeking royalties after their
    patent’s expiration. Core Health did not inform Max Rack of this belief.
    Max Rack’s relationship with Core Health prospered for several years. From 2006 to
    2015, Star Trac and Core Health sold over 5,300 Max Racks under the licensing agreement.
    Max Rack received over $640,000 in royalties.
    Things changed in October 2015. Core Health sent Max Rack a letter stating its intent to
    get out of the agreement. The last patent on the machine would expire on November 21. Core
    Health did not plan to use the “Max Rack” mark on, or pay royalties for, machines made after
    that date. The agreement did, however, include a six-month window for Core Health to sell off
    remaining “Max Rack” machines that Core Health had started manufacturing before the
    agreement’s expiration (as long as it paid royalties to Max Rack).
    B
    Going forward, Core Health planned to make an identical machine using a new
    trademark: the “Freedom Rack.” Max Rack challenged the way in which Core Health executed
    this change.     Max Rack had concerns both with Core Health’s marketing and with its
    manufacturing.
    1. Max Rack contended that Core Health did not stop using the Max Rack trademark in
    its marketing quickly enough. Core Health’s former vice president of marketing could not
    remember taking any action between July and December 2015 to change to the Freedom Rack
    mark, even though switching a brand name can take up to two months. Emails suggest that Core
    Health waited until over a month after the agreement’s expiration to update its website and other
    marketing materials. In January 2016, a marketing team attempted to scrub all marketing
    materials of the old mark. This review took two weeks.
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.              Page 7
    But the employees did not catch everything. Max Rack’s Kirt Moritz claimed that, for
    years, he would search Google for “Max Rack” and get back “hits” in which this mark was used
    with the Freedom Rack. He kept a record of one search from mid-2016. This search returned a
    link to Core Health’s website with a caption underneath the link that used the “Max Rack” name.
    Core Health’s former vice president of marketing suggested that this link sent viewers to a sales
    page (but Max Rack did not introduce a copy of the page). He also noted that Core Health’s
    sales page for the Freedom Rack at one point contained an outdated picture of a machine with a
    partially covered “Max Rack” logo on the bottom bar.
    These issues were not limited to Core Health. Moritz’s Google searches revealed that
    some third-party resellers, including OC Fitness Source, advertised the Freedom Rack using the
    Max Rack name. When Moritz visited OC Fitness’s site, he saw that it used a picture of the Max
    Rack.
    Core Health’s former vice president of marketing admitted that someone in marketing
    should have done the types of Google searches that Moritz undertook. But the employee in
    charge of the changeover could not remember doing so. Max Rack’s counsel confronted this
    employee with some remaining “Max Rack” references on Core Health’s website during her
    deposition. She later testified that she fixed all references that she learned of and explained that
    they had resulted from “[h]uman error.” O’Brien Tr., R.101, PageID 2573.
    Other errors remained. As late as trial, Moritz testified, the owner’s manual for the
    Freedom Rack, which could be found on Core Health’s website, included a reference to “Max
    Rack.” One line in the manual’s installation instructions read: “UNPACK THE MAX RACK
    and ASSEMBLE UPPER WELDMENTS.” Manual, R.95-1, PageID 1813. The manual used
    the “Freedom Rack” mark in many other places, however, including on its title page.
    2. Max Rack also had production complaints.          Kevin Corbalis, Core Health’s vice
    president of product development, testified that, by December 2015, the overseas manufacturer
    had finished all remaining Max Racks and begun to make Freedom Racks.
    Yet company emails undermine this claim. They show Core Health scrambling to change
    its production from Max Racks to Freedom Racks into 2016. A December 23, 2015 email, for
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 8
    example, suggests that the company still needed to size the new “Freedom Rack” artwork to fit
    the machine. Email, R.94-11, PageID 1752. Five days later, another email reiterates: “Need
    name change, logo, MaxRack removed from placard ASAP so no more are produce[d].” Email,
    R.94-12, PageID 1754.
    Correspondence in 2016 muddies things further. Emails from January suggest that Core
    Health had still not assigned the “Freedom Rack” a product number or sent engineering changes
    to the manufacturer. By February, the company was still trying to figure out the Freedom Rack’s
    standard cost and to ensure that the “next production runs” were of Freedom Racks. Emails,
    R.95-5, PageID 1833–1834; R.94-18, PageID 1771. A year after the agreement’s expiration,
    Corbalis asked employees for the “last manufacture date of the Max Rack[.]” Email, R.95-4,
    PageID 1832. An employee responded that Core Health had not authorized the manufacturer to
    “scrap all remaining Max Rack labels” until May 2016. 
    Id.,
     PageID 1831.
    Core Health’s records show that, after the agreement expired in 2015, it sold 271
    machines made as Max Racks. The agreement gave Core Health six months to sell off any units
    that it had started making before the expiration date. It sold 238 within that window. But it sold
    another 24 later in violation of the agreement. Core Health placed the remaining 9 units in a
    unique category because Core Health’s records suggested that it changed their labels to
    “Freedom Rack” before shipping them. Until this litigation, Core Health did not pay Max Rack
    for any of these 271 sales.
    Today, the Freedom Rack is the only machine on the market like the Max Rack. Skilken
    has not tried to enter into another licensing agreement to make “Max Racks.”             And no
    competitors sell a similar product.
    C
    Max Rack sued Core Health (and two other parties that do not matter now). As relevant
    here, Max Rack alleged two claims under the federal Lanham Act: for infringing its Max Rack
    trademark and engaging in unfair competition. See 
    15 U.S.C. §§ 1114
    (1)(a), 1125(a)(1)(A). It
    also alleged three claims under Ohio’s Deceptive Trade Practices Act: for passing off the Max
    Rack machine as its own, causing a likelihood of confusion as to the source of Core Health’s
    Nos. 20-3598/3600        Max Rack, Inc. v. Core Health & Fitness, LLC, et al.           Page 9
    machines, and causing a likelihood of confusion as to Core Health’s affiliation with the Max
    Rack mark. See Ohio Rev. Code § 4165.02(A)(1)–(3).
    These claims proceeded to trial. The week before trial started, Core Health sent Max
    Rack a belated check for the 262 units that it sold and shipped as Max Racks after the
    agreement’s expiration. For the 238 that it sold within the six-month window allowed by the
    agreement, Core Health paid the royalty plus interest. For the 24 that it sold in breach of the
    agreement, it paid its profits plus interest.
    After trial, the jury ruled for Max Rack on all counts. It awarded Max Rack $1 million in
    damages and $250,000 in Core Health’s profits. It also found that Core Health’s infringement
    had been intentional.
    The parties filed competing post-trial motions. The district court ruled for each side in
    part. Max Rack, Inc. v. Core Health & Fitness, LLC, 
    2020 WL 2128614
    , at *12 (S.D. Ohio May
    5, 2020). The court ruled for Max Rack on most issues. It denied Core Health’s motion for
    judgment as a matter of law or for a new trial as to its liability. 
    Id.
     at *2–5. The court also
    upheld Max Rack’s $250,000 award tied to Core Health’s profits. 
    Id. at *5
    . It even enhanced
    this award to $500,000 because of Core Health’s conduct during discovery. 
    Id.
     at *6–7. The
    court next awarded Max Rack attorney’s fees. 
    Id.
     at *7–8. It lastly enjoined Core Health from
    using the Max Rack mark without acknowledging that it is registered to Max Rack, Inc. 
    Id. at *11
    .
    At the same time, the district court ruled for Core Health regarding Max Rack’s award of
    $1 million in damages. The court overturned this award because Max Rack presented no
    evidence that Core Health’s use of the Max Rack mark had actually confused any consumers. 
    Id.
    at *4–5. It later rejected Max Rack’s motion for reconsideration on this damages question. Max
    Rack, Inc. v. Core Health & Fitness, LLC, 
    2020 WL 4933920
    , at *5 (S.D. Ohio Aug. 24, 2020).
    II
    Both parties appealed. Their combined appeals require us to consider six questions:
    (1) Did Max Rack present sufficient evidence that Core Health violated the law? (2) Is Core
    Nos. 20-3598/3600        Max Rack, Inc. v. Core Health & Fitness, LLC, et al.               Page 10
    Health entitled to a new trial? (3) Should the district court have vacated the jury’s profits award?
    (4) Did the court properly double that profits award? (5) Did the district court mistakenly vacate
    the jury’s damages award? (6) And did it wrongly grant attorney’s fees to Max Rack?
    Question 1: Did Max Rack present sufficient evidence that Core Health violated the law?
    Core Health begins by arguing that the district court should have granted it judgment as a
    matter of law under Federal Rule of Civil Procedure 50 because no reasonable jury could have
    found for Max Rack. Recall that Max Rack presented to the jury two federal claims under the
    Lanham Act (for trademark infringement and unfair competition) and three state claims under
    Ohio’s Deceptive Trade Practices Act (for passing off goods, causing likelihood of confusion as
    to source, and causing likelihood of confusion as to affiliation). See 
    15 U.S.C. §§ 1114
    (1)(a),
    1125(a)(1)(A); Ohio Rev. Code § 4165.02(A)(1)–(3). Whether the jury could have reasonably
    found for Max Rack on any one of these claims turns on that claim’s specific legal elements.
    After all, Core Health would be entitled to a judgment in its favor on a claim if Max Rack failed
    to produce enough evidence on even a single element, whether or not it proved the other
    elements. See Celotex Corp. v. Catrett, 
    477 U.S. 317
    , 322–23 (1986); Anchor v. O’Toole, 
    94 F.3d 1014
    , 1023 (6th Cir. 1996).
    But the parties do not cite the statutory text for any of the five claims, identify their legal
    elements, or otherwise distinguish between them. To seek judgment as a matter of law, Core
    Health argues generically that its use of the Max Rack mark was not likely to cause confusion as
    a matter of law, impliedly suggesting that all five claims contain an identical “likelihood of
    confusion” element. And Max Rack’s response does not challenge Core Health’s (implied)
    suggestion that the five claims share this identical element.
    Do the parties rightly assume that the claims follow the same likelihood-of-confusion
    test?   Perhaps so for the federal claims.       The two federal statutes use similar “likely to
    cause confusion” language.       Compare 
    15 U.S.C. § 1114
    (1)(a), with 
    id.
     § 1125(a)(1)(A);
    see 4 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 23:1 (5th ed.),
    Westlaw (database updated Mar. 2022). And an oft-cited treatise identifies as their primary
    difference that § 1114(1)(a) permits only owners of federally registered marks to bring
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 11
    infringement claims, whereas § 1125(a) allows owners of unregistered marks to assert these
    claims under the “unfair competition” label. 1 McCarthy, supra, § 4:6; 5 id. §§ 27:12, 27:14; cf.
    Daddy’s Junky Music Stores, Inc. v. Big Daddy’s Fam. Music Ctr., 
    109 F.3d 275
    , 280, 288 (6th
    Cir. 1997). (The treatise cautions that, when an owner of a registered mark asserts a trademark-
    infringement claim, the court should not present to the jury duplicative trademark-infringement
    claims under § 1114(1)(a) and § 1125(a) because it might confuse jurors. 5 McCarthy, supra,
    § 27:14.)
    Things are more complicated for the state claims. A recent Ohio Supreme Court decision
    has told courts not to follow federal precedent reflexively if Ohio’s Deceptive Trade Practices
    Act uses language different from the Lanham Act. See Wooster Floral & Gifts, L.L.C. v. Green
    Thumb Floral & Garden Ctr., Inc., 
    172 N.E.3d 60
    , 66–69 (Ohio 2020). But again, the parties do
    not even cite the statutory language governing these state claims. Ultimately, therefore, we will
    not identify the elements for any of the claims in this case. The parties’ briefing choices allow us
    simply to assume that all five of Max Rack’s claims required it to prove the same likelihood-of-
    confusion element that governs an infringement claim under § 1114(1)(a). Their briefing choices
    likewise allow us to assume that Max Rack proved all other (unspecified) elements.
    In a typical trademark-infringement case, a company using one mark (say, “5-hour
    ENERGY”) alleges that another company has chosen a deceptively similar mark (say, “6 Hour
    POWER”) that will lead consumers to believe goods or services bearing the marks are affiliated
    with each other. Innovation Ventures, LLC v. N.V.E., Inc., 
    694 F.3d 723
    , 731 (6th Cir. 2012);
    see also, e.g., AWGI, LLC v. Atlas Trucking Co., 
    998 F.3d 258
    , 268 (6th Cir. 2021) (“Atlas”
    versus “Atlas Trucking” and “Atlas Logistics”); Daddy’s, 
    109 F.3d at
    278–79 (“Daddy’s Junky
    Music Stores” and “Daddy’s” versus “Big Daddy’s Family Music Center”). In that context, we
    have evaluated the totality of the circumstances to decide whether a new mark would generate a
    “likelihood of confusion” as to its affiliation with the established mark. See Homeowners Grp.,
    Inc. v. Home Mktg. Specialists, Inc., 
    931 F.2d 1100
    , 1106–07 (6th Cir. 1991). We have also
    asked eight recurring questions to help ensure consistent results across cases under this consider-
    everything approach: How well-known is the plaintiff’s mark? How related are the parties’
    goods? How similar are the marks? Have consumers actually been confused? Do the parties
    Nos. 20-3598/3600        Max Rack, Inc. v. Core Health & Fitness, LLC, et al.           Page 12
    use similar marketing channels? How much care do consumers exercise when buying these
    products? Did the defendant intentionally choose its mark because of the plaintiff’s? And are
    the companies’ product lines likely to expand? See Frisch’s Rests., Inc. v. Elby’s Big Boy of
    Steubenville, Inc., 
    670 F.2d 642
    , 648 (6th Cir. 1982).
    But this case is not a typical one. Max Rack does not allege that Core Health’s “Freedom
    Rack” mark is deceptively similar to its “Max Rack” mark. It has no objection to that new mark.
    Rather, Max Rack alleges that Core Health continued to use the “Max Rack” mark after the
    expiration of the agreement that gave it the right to do so.
    A separate body of law has developed for this distinct claim that a holdover licensee has
    continued to use a licensor’s mark after their agreement expired. See 4 McCarthy, supra,
    § 25:31 (citing cases).     In this context, courts have jettisoned the usual totality-of-the-
    circumstances test in favor of a more categorical rule: “proof of continued, unauthorized use of
    an original trademark by one whose license to use the trademark had been terminated is
    sufficient to establish ‘likelihood of confusion.’” U.S. Structures, Inc. v. J.P. Structures, Inc.,
    
    130 F.3d 1185
    , 1190 (6th Cir. 1997); ITT Indus., Inc. v. Wastecorp. Inc., 87 F. App’x 287, 293
    (3d Cir. 2004); Gorenstein Enters., Inc. v. Quality Care-USA, Inc., 
    874 F.2d 431
    , 435 (7th Cir.
    1989); Burger King Corp. v. Mason, 
    710 F.2d 1480
    , 1492–93 (11th Cir. 1983); Pro. Golfers
    Ass’n of Am. v. Bankers Life & Cas. Co., 
    514 F.2d 665
    , 670 (5th Cir. 1975); L & L Wings, Inc. v.
    Marco-Destin, Inc., 
    676 F. Supp. 2d 179
    , 188 (S.D.N.Y. 2009). This rule makes good sense
    when a holdover licensee blatantly infringes the licensor’s mark by using that mark in the same
    way as it did before the agreement expired. Without any need to tick through factors, it should
    be obvious that consumers will believe, for example, that a pizza restaurant remains affiliated
    with the “Little Caesars” chain when a terminated franchisee continues to operate the pizzeria as
    a Little Caesars. See Little Caesars Enters., Inc. v. Miramar Quick Serv. Rest. Corp., 
    2020 WL 4516289
    , at *3 (6th Cir. 2020) (order).
    How do these principles play out here? Before applying them to Max Rack’s claims, we
    must begin with uncertainty over our standard for evaluating the jury’s verdict. Circuit courts
    have disagreed over the proper standard of review for the ultimate question whether a likelihood
    of confusion exists. See 4 McCarthy, supra, §§ 23:71, 23:73. Traditionally, we have viewed this
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 13
    question as one of law that we have considered de novo. Progressive Distrib. Servs., Inc. v.
    United Parcel Serv., Inc., 
    856 F.3d 416
    , 427 (6th Cir. 2017); Homeowners Grp., 
    931 F.2d at 1107
    ; Frisch’s, 
    670 F.2d at 651
    . But a recent Supreme Court case suggests that the question
    could at least sometimes raise a fact question for the jury that we would review deferentially.
    See Hana Fin., Inc. v. Hana Bank, 
    574 U.S. 418
    , 422 (2015). Regardless, Max Rack produced
    enough evidence to establish a likelihood of confusion even under our traditional standard. See
    Homeowners Grp., 
    931 F.2d at 1107
    .
    Max Rack presented two theories to the jury: (1) that Core Health sold some Max Racks
    without authorization, and (2) that Core Health used the Max Rack mark when selling Freedom
    Racks to confuse consumers into thinking that the Freedom Rack was affiliated with the Max
    Rack brand. The district court relied on the second theory, but we think it easier to start with the
    first.
    Theory One: Continued Sales of Max Racks. The jury reasonably could have found that
    Core Health violated the licensing agreement by selling Max Racks after that agreement expired.
    This violation would trigger our rule that trademark infringement exists if a holdover licensee
    continues to use a licensor’s mark on its goods without authorization. See U.S. Structures,
    
    130 F.3d at 1190
    . Most notably, Core Health admits that it sold 24 Max Racks outside the six-
    month window that the agreement gave it to liquidate inventory. These unauthorized sales alone
    supported the jury’s finding that Core Health intentionally infringed the Max Rack mark by
    leading the public “to think that the continuing user [Core Health] is still connected with the
    trademark owner [Max Rack].” 4 McCarthy, supra, § 25:31; see also Bill Blass, Ltd. v. SAZ
    Corp., 
    751 F.2d 152
    , 154–55 (3d Cir. 1984). In fact, Core Health paid Max Rack its profits on
    these units before trial. That decision may have reduced its monetary liability, but it did not
    erase its violation.
    In addition, the jury reasonably (if just barely) could have found that Core Health began
    to manufacture Max Racks after the agreement expired in November 2015. Core Health agrees
    that it sold 238 Max Racks after the expiration date (but before the end of the six-month
    window). Although the agreement allowed Core Health to sell those units if they were in
    production as of the expiration date, the jury could have found that Core Health started making
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.            Page 14
    them only after the agreement’s termination. Core Health presented no records identifying the
    dates that its manufacturer started producing the units sold after the agreement expired. And
    Max Rack presented plenty of circumstantial evidence that the units were placed in production
    too late. Among other things, emails from over a month after the expiration suggested that Core
    Health’s manufacturer needed “MaxRack removed from placard ASAP so no more are
    produce[d].” Email, R.94-12, PageID 1754. In February 2016, moreover, Core Health was still
    trying to ensure that the “next production runs” would be of Freedom Racks. Emails, R.95-5,
    PageID 1833–1834. And Core Health did not authorize its manufacturer to “scrap” the Max
    Rack labels for a few more months. Emails, R.95-4, PageID 1831. Yet any sales of later-
    produced Max Racks would have violated the agreement and so likewise infringed Max Rack’s
    trademark. See U.S. Structures, 
    130 F.3d at 1190
    .
    Lastly, Max Rack cites the testimony of Core Health’s finance director regarding 9 more
    units sold after the agreement expired. The director testified that those units “were purchased as
    Max Rack[s]” but that Core Health changed the logos to “Freedom Rack” before shipping the
    units to customers. Winegardner Tr., R.100, PageID 2461. As Max Rack notes, if those
    customers were sold a “Max Rack,” they surely would have been confused when opening the
    box and finding a “Freedom Rack” inside. Core Health responds that Max Rack misconstrued
    this testimony. According to Core Health, the finance director suggested that Core Health
    bought the units as “Max Racks” from the manufacturer and rebranded them as “Freedom
    Racks” before selling them. To support this claim, however, Core Health points to only more
    opaque testimony: the director later testified that the units “were purchased as Max Rack and
    sold as Freedom Rack.” 
    Id.,
     PageID 2564. Purchased by whom? His testimony shows the perils
    of the passive voice. Because Core Health failed to clear up the factual ambiguity at trial, the
    jury could have treated these sales as infringing. All told, then, the jury reasonably could have
    found that Core Health infringed the Max Rack mark by selling up to 271 Max Racks without
    authorization.
    Theory Two: Use of “Max Rack” Name to Sell Freedom Racks. The district court did not
    rely on this evidence of continuing Max Rack sales. It instead held that the jury could have
    found that Core Health, when selling Freedom Racks, continued to use the name “Max Rack” so
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 15
    as to lead consumers to believe that the machines were affiliated with the Max Rack brand. Yet
    the court cited three pieces of relatively insubstantial evidence for this conclusion: OC Fitness, a
    third-party reseller, used the Max Rack mark when selling Freedom Racks on its website; Kirt
    Moritz conducted a Google search for “Max Rack” and got a “hit” to an unidentified page on
    Core Health’s website; and a stray “Max Rack” reference remained in the Freedom Rack
    owner’s manual. See Max Rack, 
    2020 WL 2128614
    , at *2–3.
    Did this evidence suffice? The district court committed legal error by relying on the first
    piece: that a downstream reseller (OC Fitness) used the Max Rack mark when advertising
    Freedom Racks. Max Rack does not identify any basis for imputing this separate entity’s
    conduct to Core Health. We can think of one possibility: the Supreme Court has held that a
    trademark owner can hold a competing manufacturer liable for “contributory infringement” if the
    competitor intentionally convinces a third-party distributor to infringe the owner’s mark or if the
    competitor continues to sell to the distributor after learning of the infringing conduct. See
    Inwood Lab’ys., Inc. v. Ives Lab’ys., Inc., 
    456 U.S. 844
    , 853–54, 854 n.13 (1982); 4 McCarthy,
    supra, § 25:18. Here, however, Core Health warned its regular resellers of the pending name
    change. And Max Rack points to no evidence that Core Health knew of OC Fitness’s use of the
    Max Rack mark. So Max Rack does not even present a contributory-infringement theory on
    appeal, and the district court erred by relying on OC Fitness’s alleged misconduct to hold Core
    Health liable.
    That said, although we find the question close, we agree that Core Health’s own
    continued uses of the Max Rack mark could have created this likelihood of confusion when
    considered collectively. At the outset, keep in mind a critical distinction in the Lanham Act
    between proving a violation and proving a proper remedy. See Web Printing Controls Co., v.
    Oxy-Dry Corp., 
    906 F.2d 1202
    , 1204–05 (7th Cir. 1990). To prove a violation, Max Rack
    needed to show only that Core Health’s use of the Max Rack name would likely confuse
    consumers over whether the Freedom Rack was affiliated with the owner of the Max Rack mark
    in some way. Max Rack did not need to show actual confusion or past injury. See id.; see also
    Balance Dynamics Corp. v. Schmitt Indus., Inc., 
    204 F.3d 683
    , 689 (6th Cir. 2000). We should
    Nos. 20-3598/3600     Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 16
    not conflate the minimum showing required to prove a violation with the more rigorous showing
    required to obtain damages. See 4 McCarthy, supra, § 23:12.
    In addition, there is no dispute that Core Health did continue to use the Max Rack name,
    even if accidentally, in association with its Freedom Rack sales. As one example, the employee
    who oversaw the change to the Freedom Rack in Core Health’s marketing materials conceded
    that some parts of the website mistakenly referred to the Max Rack up through the time of her
    deposition. O’Brien Tr., R.101, PageID 2572–73. As another example, Core Health’s former
    vice president of marketing suggested that the company’s ads of the Freedom Rack on its
    website used an outdated photo of a Max Rack with a partially concealed “Max Rack” logo on a
    black bar:
    Website, R.93-7, PageID 1542; Dilts Tr., R.101, PageID 2496. Kirt Moritz at Max Rack
    similarly testified that his 2016 search for “Max Rack” on Google returned a “hit” to Core
    Health’s Star Trac website referencing “Max Rack” in a snippet under the web address:
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.            Page 17
    Ex. 112, R.94-2, PageID 1584. Core Health’s former vice president of marketing suggested that
    this link directed viewers to a sales page. Lastly, at the time of trial, Moritz identified an
    extraneous “Max Rack” reference still in the Freedom Rack owner’s manual on Core Health’s
    website. Manual, R.95-1, PageID 1813.
    Core Health rightly responds that these examples show minor uses of the Max Rack mark
    and perhaps prove only that it implemented the change carelessly. But this response—that Core
    Health’s uses of the Max Rack mark were harmless—goes more to whether Max Rack
    established damages than to whether it established a violation. See Web Printing, 
    906 F.2d at 1204
    . Core Health needed to ensure that its materials did not perpetuate the “perception that” its
    Freedom Racks were affiliated with the Max Rack brand. Frisch’s Rest., Inc. v. Shoney’s Inc.,
    
    759 F.2d 1261
    , 1270 (6th Cir. 1985) (citation omitted). These references could create that
    perception—whether we consider the question under our likelihood-of-confusion factors (for
    similar marks) or our categorical rule (for holdover licensees). The goods were the same, the
    companies were potential competitors, and the mark (Max Rack) was identical. And Core
    Health (a former licensee) continued to use the Max Rack mark after the expiration of the
    licensing agreement. Indeed, if we accepted Core Health’s argument that these uses of “Max
    Rack” did not violate the Lanham Act because they were trivial, Core Health could continue with
    them indefinitely. Max Rack would remain powerless to stop Core Health from, say, using the
    picture of a Max Rack to sell its Freedom Racks.
    Important, too, is Core Health’s failure to identify a procompetitive justification for its
    “Max Rack” references. Such a justification often exists when a competitor uses another’s mark
    to give consumers truthful information about competing goods.           That information allows
    consumers to identify their preferred product with lower search costs. See August Storck K.G. v.
    Nabisco, Inc., 
    59 F.3d 616
    , 618 (7th Cir. 1995); 4 McCarthy, supra, §§ 23:11, 25:52; William
    McGeveran & Mark P. McKenna, Confusion Isn’t Everything, 
    89 Notre Dame L. Rev. 253
    , 284
    (2013). Suppose, for example, that a new entrant decides to reproduce a well-known type of
    mineral water. The new entrant may truthfully tell consumers that it sells water similar to the
    well-known brand at a lower price. See Saxlehner v. Wagner, 
    216 U.S. 375
    , 379–81 (1910); cf.
    Conopco, Inc. v. May Dep’t Stores Co., 
    46 F.3d 1556
    , 1565 (Fed. Cir. 1994). As Justice Holmes
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 18
    explained, the new entrant is “not trying to get the good will of the name, but the good will of the
    goods.” Saxlehner, 
    216 U.S. at
    380–81. This type of “fair use” of a mark seeks to enlighten
    rather than confuse. See 4 McCarthy, supra, § 25:52. And we should not apply trademark law in
    a way that effectively gives owners of established products monopoly power over those products
    (something within patent law’s domain). See Saxlehner, 
    216 U.S. at 380
    .
    In this case, then, Max Rack could not rely on trademark law to prevent Core Health from
    truthfully telling consumers that the Freedom Rack is identical to the Max Rack or that Core
    Health used to make the Max Rack for Max Rack, Inc. Yet Core Health did not assert at trial
    that its challenged uses of the Max Rack mark served any information-supplying purpose. That
    is for good reason. A consumer who saw the references would come away thinking that the
    Freedom Rack’s producer was affiliated with the Max Rack’s owner, not that they were
    potentially competing against each other. Core Health instead asserted that the references were
    “accidents.”   Whether innocent or not, however, the references still could have created a
    likelihood of confusion. Cf. Louis Vuitton S.A. v. Lee, 
    875 F.2d 584
    , 589 (7th Cir. 1989).
    Question 2: Is Core Health entitled to a new trial?
    Even if Core Health cannot obtain judgment as a matter of law under Rule 50, it next
    argues that the district court at least should have granted it a new trial under Rule 59. But the
    district court properly exercised its discretion by denying Core Health’s new-trial request.
    A district court may grant a new trial “for any reason for which a new trial has heretofore
    been granted in an action at law in federal court[.]” Fed. R. Civ. P. 59(a)(1)(A). We have
    interpreted this rule to permit a new trial on many grounds. See Mosby-Meachem v. Memphis
    Light, Gas & Water Div., 
    883 F.3d 595
    , 606 (6th Cir. 2018); see also 11 Charles Alan Wright
    et al., Federal Practice & Procedure § 2805, at 68–73 (3d ed. 2012). The jury’s verdict might
    conflict with the weight of the evidence presented at trial. Compare Innovation Ventures, LLC v.
    N2G Distrib., Inc., 
    763 F.3d 524
    , 534–38 (6th Cir. 2014), with Strickland v. Owens Corning,
    
    142 F.3d 353
    , 357–58 (6th Cir. 1998). Or improper arguments might have caused the verdict to
    rest more on prejudice against the losing party than on the merits of the winning party’s case.
    Compare Tompkins v. Crown Corr, Inc., 
    726 F.3d 830
    , 835–36 (6th Cir. 2013), with City of
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.              Page 19
    Cleveland v. Peter Kiewit Sons’ Co., 
    624 F.2d 749
    , 755–60 (6th Cir. 1980). The excessive
    nature of a monetary award might also justify a new trial (or at least a remittitur) if it shows that
    the verdict arose out of passion or mistake. Compare Roush v. KFC Nat’l Mgmt. Co., 
    10 F.3d 392
    , 397–400 (6th Cir. 1993), with Farber v. Massillon Bd. of Educ., 
    917 F.2d 1391
    , 1395–96
    (6th Cir. 1990).
    Core Health relied on several of these grounds here. It claimed that the weight of the
    evidence did not support the verdict, and that the verdict instead resulted from Max Rack’s
    efforts to bias the jury against Core Health with irrelevant evidence. To support this bias claim,
    Core Health relied on references at trial to the “addendum” to the licensing agreement that Max
    Rack entered into with Star Trac in 2010. Under this addendum, Max Rack allowed Star Trac to
    make belated royalty payments in exchange for a right to a perpetual royalty. When Core Health
    took over Star Trac’s contracts, it decided that a court would not enforce the requirement to pay
    royalties forever because of caselaw that barred patent holders from seeking royalties after their
    patent’s expiration. See Kimble v. Marvel Ent., LLC, 
    576 U.S. 446
    , 449 (2015). Max Rack does
    not dispute Core Health’s understanding of the law. In fact, Max Rack did not even assert a
    breach-of-contract claim based on Core Health’s decision to violate the addendum and stop
    paying royalties.
    According to Core Health, however, Max Rack’s witnesses repeatedly inflamed the jury
    with claims that Core Health unfairly broke its promise in the addendum. Steve Skilken’s
    testimony offers a good example. When counsel asked him if he considered the licensing
    agreement to have been “successful,” he volunteered: “There’s an addendum that supersedes
    this.” Skilken Tr., R.99, PageID 2220. He proceeded to describe the addendum’s origins. When
    Star Trac “ran out of money,” Skilken explained, Max Rack could have licensed its machine to
    other parties. 
    Id.,
     PageID 2221. But Skilken would “hate to do that to anybody,” so he kindly
    deferred Star Trac’s payments in exchange for the perpetual royalty. 
    Id.
     When the patent neared
    its expiration six years later, Skilken complained, Core Health refused even to have a
    “discussion” about abiding by its end of the bargain. 
    Id.,
     PageID 2234–36. His testimony was
    littered with similar remarks. See, e.g., 
    id.,
     PageID 2249–50.
    Nos. 20-3598/3600         Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 20
    Max Rack also invoked the addendum during the cross-examination of Core Health’s
    general counsel. The general counsel admitted that he had concluded that a court would not
    enforce the addendum when he first reviewed it in 2010. Brown Tr., R.101, PageID 2592–93.
    Max Rack’s counsel then characterized the general counsel’s failure to inform Max Rack of this
    view as “fraudulent.” 
    Id.,
     PageID 2593, 2602.
    To top it off, Core Health identifies objective factors indicating that this largely irrelevant
    testimony had its intended effect. During deliberations, the jury submitted questions about the
    addendum: “Is [the addendum] a lawful agreement to take into account given that there was no
    lawyer present?” and “Is there any law that prevents royalties to be collected on an expired
    patent, federal or state law?” Tr., R.102, PageID 2687. Core Health also notes that, as the
    district court found, Max Rack’s $1 million damages award had no basis in the evidence.
    For several reasons, though, the district court did not commit reversible error in denying
    Core Health’s new-trial motion. Those reasons begin with our standard of review. We review
    the court’s decision under a deferential abuse-of-discretion standard. See Mosby-Meachem, 883
    F.3d at 602; Bach v. First Union Nat’l Bank, 149 F. App’x 354, 367–68 (6th Cir. 2005). Absent
    a legal error, this standard generally bars us from overturning a court’s discretionary denial of a
    new-trial motion unless the record leaves us with a “definite and firm conviction” that the court
    “committed a clear error of judgment[.]” Holmes v. City of Massillon, 
    78 F.3d 1041
    , 1045 (6th
    Cir. 1996) (citation omitted); see also CFE Racing Prods., Inc. v. BMF Wheels, Inc., 
    793 F.3d 571
    , 584 (6th Cir. 2015). We owe special deference to the district court for claims (like Core
    Health’s) that improper arguments tainted the jury’s decisionmaking. See CFE Racing, 793 F.3d
    at 590. Unlike us, that court had the ability to evaluate the allegedly improper conduct (and its
    prejudicial impact) by witnessing the conduct firsthand, not by reviewing a transcript of it
    secondhand. See Balsley v. LFP, Inc., 
    691 F.3d 747
    , 762 (6th Cir. 2012) (citing Peter Kiewit
    Sons’, 
    624 F.2d at 756
    ).
    We have no definite and firm conviction of an error. As an initial matter, our caselaw
    recognizes that a district court’s instructions can fix potential harm caused by prejudicial
    arguments or evidence. See Holmes, 
    78 F.3d at
    1046–47; see also, e.g., CFE Racing, 793 F.3d at
    589; Troyer v. T.John.E. Prods., Inc., 526 F. App’x 522, 525–26 (6th Cir. 2013). The court’s
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 21
    instructions in this case served that purpose. Its responses to the questions that the jury posed
    during deliberations explained that the jurors need not consider the addendum’s legality and that
    Core Health had correctly interpreted the law. The court indicated that “[t]he lawfulness of [the
    addendum] is not relevant to the decision that you have to make” and that “there is state and
    federal law that prohibits any agreement to extend royalties to a patent past the patent
    expiration.” Tr., R.102, PageID 2693–94. At one point in Skilken’s testimony about the
    addendum, moreover, the court instructed the jury to disregard his statements. See Skilken Tr.,
    R.99, PageID 2227–28. We presume that the jury followed these instructions when reaching its
    verdict. See Troyer, 526 F. App’x at 525.
    In addition, the jury’s liability finding (that Core Health engaged in conduct that created a
    likelihood of confusion) does not conflict with the weight of the evidence. Although Rule 59’s
    weight-of-the-evidence standard differs from Rule 50’s sufficiency-of-the-evidence standard, we
    reach this result based on the same evidence and reasoning on which we relied when affirming
    the denial of judgment as a matter of law. See, e.g., Mosby-Meachem, 883 F.3d at 606–07; J.C.
    Wyckoff & Assocs. v. Standard Fire Ins. Co., 
    936 F.2d 1474
    , 1488 (6th Cir. 1991). Furthermore,
    since our court treats the ultimate likelihood-of-confusion question as a legal one that we review
    de novo, see Homeowners Grp., 
    931 F.2d at 1107
    , a new trial on that question would make even
    less sense. That is especially so because Core Health does not dispute that at least some of its
    actions (the sale of 24 Max Racks outside the six-month window) amounted to infringement.
    See U.S. Structures, 
    130 F.3d at 1190
    . It instead challenges the scope of its liability and the size
    of the awards. Yet why should these claims entail a complete redo?
    Lastly, the jury’s $1 million damages award provides no basis for a new trial either. To
    be sure, the district court agreed that the evidence did not support this award when overturning it.
    See Max Rack, 
    2020 WL 2128614
    , at *4–5. And a court may find an entirely new trial proper if
    prejudicial arguments on one issue so inflamed the jury as to have “spillover” effects on other
    issues. Peter Kiewit Sons’, 
    624 F.2d at 759
    ; see Dossett v. First State Bank, 
    399 F.3d 940
    , 946–
    47 (8th Cir. 2005). But the district court could reasonably conclude that the award did not prove
    these spillover effects. The award could have been based just as much on a legal mistake as on
    undue prejudice. As Core Health notes, Max Rack’s witnesses testified that Max Rack spent $1
    Nos. 20-3598/3600          Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 22
    million bringing the product to market, and Max Rack’s counsel referred to those efforts in
    closing argument. Max Rack does not now claim to be entitled to recover for the efforts. Yet
    this mistake would not prove a vindictive verdict, and courts have routinely granted new trials
    solely on damages (or remittiturs) for similar mistakes. See 11 Wright et al., supra, § 2814, at
    198, § 2815, at 208; see, e.g., Skalka v. Fernald Env’t Restoration Mgmt. Corp., 
    178 F.3d 414
    ,
    425 (6th Cir. 1999). The district court did not abuse its discretion by holding that its decision to
    vacate the award sufficed to correct the error.
    Question 3: Did the district court commit error by failing to vacate the jury’s profits award?
    Unable to rebut the jury’s liability findings, Core Health turns to challenging its monetary
    remedies. Core Health initially argues that the district court wrongly refused to overturn the
    jury’s $250,000 award to Max Rack for Core Health’s profits. But this sum had a reasonable
    basis in the evidence when read against the Lanham Act’s burden-shifting approach to proving
    profits.
    If a defendant infringes a trademark in violation of §§ 1114(1)(a) or 1125(a)(1)(A), the
    Lanham Act offers several remedies. See Romag Fasteners, Inc. v. Fossil, Inc., 
    140 S. Ct. 1492
    ,
    1494–95 (2020). (The parties do not suggest that Ohio law follows different rules for any of the
    remedial issues that they raise, so we need not consider that law separately.) The Act indicates
    that “the plaintiff shall be entitled, . . . subject to the principles of equity, to recover (1)
    defendant’s profits, (2) any damages sustained by the plaintiff, and (3) the costs of the action.”
    
    15 U.S.C. § 1117
    (a). When deciding whether a plaintiff can recover a “defendant’s profits,”
    courts have looked to the “principles of equity” because an accounting of profits has traditionally
    been viewed as an equitable remedy. 5 McCarthy, supra, § 30:61. Courts have identified
    several equitable factors to help guide the inquiry, including whether the defendant acted
    willfully and whether other remedies would suffice. See Romag, 140 S. Ct. at 1497; Kars 4 Kids
    Inc. v. Am. Can!, 
    8 F.4th 209
    , 223 (3d Cir. 2021); Quick Techs., Inc. v. Sage Grp. PLC, 
    313 F.3d 338
    , 349 (5th Cir. 2002).
    Here, however, Core Health does not challenge the district court’s abstract conclusion
    that equity allowed for a profits award. It instead challenges the court’s conclusion that the
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 23
    evidence sufficed to show that it earned any money through infringing conduct. The Act adopts
    a burden-shifting approach to proving a defendant’s profits: “In assessing profits the plaintiff
    shall be required to prove defendant’s sales only; defendant must prove all elements of cost or
    deduction claimed.” 
    15 U.S.C. § 1117
    (a). This text placed a modest burden on Max Rack to
    prove Core Health’s “sales.” It then flipped the burden to Core Health to identify any deductions
    from this number for such things as production costs. See Wynn Oil Co. v. Am. Way Serv. Corp.,
    
    943 F.2d 595
    , 606 (6th Cir. 1991); 5 McCarthy, supra, § 30:65; see also Mishawaka Rubber &
    Woolen Mfg. Co. v. S.S. Kresge Co., 
    316 U.S. 203
    , 206–07 (1942).
    Core Health argues that Max Rack did not meet its initial burden to prove Core Health’s
    “sales.” At the outset, though, the parties debate the standard of review. Core Health all but asks
    us to grant it judgment as a matter of law under Rule 50, suggesting that we should apply a
    sufficiency-of-the-evidence test and review the district court’s denial of relief de novo.
    See Mosby-Meachem, 883 F.3d at 602. Max Rack responds that we should review the district
    court’s denial of relief for an abuse of discretion because Core Health’s post-trial motion referred
    to the profits award only when seeking a new trial under Rule 59. See id. To make matters more
    complex, it is not clear that the jury should have had the final say on this equitable remedy
    because several cases have noted that parties lack a right to a jury trial for an accounting of
    profits. See Ferrari S.P.A. v. Roberts, 
    944 F.2d 1235
    , 1248 (6th Cir. 1991); 6 McCarthy, supra,
    § 32:124. And when a trademark owner seeks both legal and equitable relief, some courts have
    tasked the jury with making only an “advisory” profits finding. Cf. Variety Stores, Inc. v.
    Walmart Inc., 852 F. App’x 711, 722 & n.5 (4th Cir. 2021); Fed. R. Civ. P. 39(c). This fact, too,
    could affect our standard of review. See Visible Sys. Corp. v. Unisys Corp., 
    551 F.3d 65
    , 79 n.10
    (1st Cir. 2008).
    Rather than resolve any standard-of-review questions, we will simply accept Core
    Health’s framing that we should review the district court’s refusal to overturn the profits award
    de novo using a sufficiency-of-the-evidence test. Even under that test, enough evidence justified
    the award.
    Turning to the merits, Core Health argues that the district court wrongly held that the
    stray references to “Max Rack” on its website could have tainted its Freedom Rack sales. See
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.            Page 24
    Max Rack, 
    2020 WL 2128614
    , at *5. Core Health has a point. What does the Act mean when it
    provides that a plaintiff must prove a “defendant’s sales”? 
    15 U.S.C. § 1117
    (a). Which sales?
    Core Health, for example, makes many weightlifting machines. It would seem dubious to claim
    that Max Rack could satisfy its initial burden merely by introducing Core Health’s total
    companywide sales data and thereby shift to Core Health the burden to disprove that its
    infringement affected every dollar of revenue. As Judge Posner explained, a plaintiff likely
    cannot place an infringer’s “corporate income tax return in the record and rest [its] case for an
    award of infringer’s profits.” Taylor v. Meirick, 
    712 F.2d 1112
    , 1122 (7th Cir. 1983). Instead, a
    plaintiff likely must show some connection between the identified “sales” and the alleged
    infringement. For example, in the related copyright context (which uses a similar burden-
    shifting approach), we have held that a plaintiff’s proposed “gross revenue number” (i.e., the
    defendant’s sales figure) “must have a reasonable relationship to the infringing activity.”
    Balsley, 691 F.3d at 768–69.
    We need not decide, however, whether the stray references to “Max Rack” on Core
    Health’s website tainted all Freedom Rack sales and allowed Max Rack to seek Core Health’s
    profits for the sales. Core Health does not dispute that the sale of any unit “bearing the
    infringing mark” would have a sufficient connection between the infringement and the sale.
    Wynn, 
    943 F.2d at 606
     (quoting Mishawaka, 
    316 U.S. at 206
    ). And Core Health ignores that the
    profits award could have been tied to Max Rack’s alternative infringement theory: that Core
    Health sold over 200 Max Racks without authorization. If the jury accepted this theory, an
    obvious connection existed between sales and infringement.           Core Health’s own figures
    suggested that it earned $486,088 in revenue and $188,787 in profits on the 218 Max Racks sold
    in 2016. Spreadsheet, R.94-16, PageID 1767. Max Rack also poked holes in the validity of Core
    Health’s cost figures, suggesting that its inflated numbers hid its true profits.     The jury’s
    $250,000 profits award thus could have rested on a distinct infringement theory that Core Health
    overlooks. That fact suffices to reject Core Health’s challenge to the award.
    Question 4: Did the district court properly double the jury’s profits award?
    Even if we reject its challenge to the profits award, Core Health next contends, the
    district court still erred by doubling that award to $500,000. This time, we agree. We review the
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 25
    court’s enhancement under the deferential abuse-of-discretion standard. Cf. La Quinta Corp. v.
    Heartland Props. LLC, 
    603 F.3d 327
    , 342 (6th Cir. 2010); Kars 4 Kids, 8 F.4th at 223–24. But
    the court abused its discretion because its justification for this enhancement rested on a
    misunderstanding of the governing burden-shifting law. Cf. Skydive Ariz., Inc. v. Quattrocchi,
    
    673 F.3d 1105
    , 1114–15 (9th Cir. 2012).
    Recognizing that a fact finder might sometimes have trouble identifying the precise
    amount of profits generated by a defendant’s infringement, Congress gave district courts the
    discretionary power to adjust a profits award up or down. The Lanham Act provides: “If the
    court shall find that the amount of the recovery based on profits is either inadequate or excessive
    the court may in its discretion enter judgment for such sum as the court shall find to be just,
    according to the circumstances of the case.” 
    15 U.S.C. § 1117
    (a). Yet the Act imposes a clear
    limit on this discretionary power: “Such sum . . . shall constitute compensation and not a
    penalty.” 
    Id.
    This language distinguishes between two types of adjustments. See La Quinta, 
    603 F.3d at
    342–43; 5 McCarthy, supra, § 30:91; see also, e.g., Kars 4 Kids, 8 F.4th at 224–25; Ga.-Pac.
    Consumer Prods. LP v. von Drehle Corp., 
    781 F.3d 710
    , 717–19 (4th Cir. 2015); Getty
    Petroleum Corp. v. Bartco Petroleum Corp., 
    858 F.2d 103
    , 109–13 (2d Cir. 1988). On the one
    hand, the court may increase a profits award for a compensatory reason, such as a concern that
    the award does not encompass the defendant’s full profits. See La Quinta, 
    603 F.3d at 343
    .
    Perhaps the defendant received intangible benefits as a result of its infringing conduct. See
    Merck Eprova AG v. Gnosis S.p.A., 
    760 F.3d 247
    , 262–63 (2d Cir. 2014). Or perhaps the
    defendant engaged in discovery “stonewalling” that prevented the plaintiff from identifying its
    total infringing sales. Cf. Bos. Pro. Hockey Ass’n, Inc. v. Dallas Cap & Emblem Mfg., Inc., 
    597 F.2d 71
    , 77 (5th Cir. 1979).
    On the other hand, the court may not increase a jury’s profits award for a punitive reason.
    The record might show such an improper purpose, for example, if the court highlighted the
    defendant’s bad faith as the basis for the increase. Cf. Skydive Ariz., 
    673 F.3d at 1115
    ; Jurgens v.
    McKasy, 
    927 F.2d 1552
    , 1564 (Fed. Cir. 1991). Or such an improper purpose might exist if the
    court increased the profits to penalize the defendant for discovery violations, something that
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.            Page 26
    other laws and court rules are better equipped to handle. Cf. Bos. Pro. Hockey Ass’n, 
    597 F.2d at 77
    .
    In this case, the district court got off to the right start by noting that the Act bars a
    punitive enhancement and by basing its increase on Core Health’s evasive discovery conduct.
    See Max Rack, 
    2020 WL 2128614
    , at *6–7. Max Rack had asked Core Health to provide all
    documents that would show the costs that Core Health incurred in making the machines. But
    Core Health allegedly failed to provide sufficient cost data, which led Max Rack to move to
    exclude any nondisclosed evidence from trial as a sanction. The court granted this motion in
    limine. After the trial, the court then relied on Core Health’s failure to disclose its cost data a
    second time by using it to enhance Max Rack’s profits award. 
    Id.
     It reasoned that Core Health’s
    cost-related discovery failures “could have undoubtedly hindered [Max Rack’s] ability to
    calculate accurately [Core Health’s] infringing profits.” 
    Id. at *7
    .
    This logic got things backward under the Lanham Act. Given the Act’s burden-shifting
    approach, the district court’s first order granting Max Rack’s motion in limine increased the
    chances of overcompensation (not undercompensation) to Max Rack. The Act required Max
    Rack “to prove [Core Health’s] sales only[.]” 
    15 U.S.C. § 1117
    (a). Max Rack had no difficulty
    doing so. It established this sales figure with a report from Core Health’s finance director
    summarizing Core Health’s revenue from its sales.          During closing argument, Max Rack
    conceded that it was “willing to accept [Core Health’s] numbers for revenue.” Tr., R.102,
    PageID 2627. It did not suggest that Core Health had hidden sales.
    At this point, the Act flipped the burden to Core Health to prove “all elements of cost or
    deduction claimed.” 
    15 U.S.C. § 1117
    (a). What would happen if Core Health did nothing to
    prove its costs? Core Health’s total sales (the amount that Max Rack agreed on) would become
    the “profits” that Max Rack could recover. 
    Id.
     But this sales figure would greatly exceed Core
    Health’s real-world profits (revenue minus expenses) because it would not account for its real-
    world expenses. The district court’s order on the motion in limine thus made it more likely that
    the jury would treat an inflated sales figure as Core Health’s profits, because the order hindered
    Core Health’s ability to prove its costs. As other courts have recognized when discussing this
    type of discovery sanction, the order punished Core Health by barring it from introducing
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 27
    relevant evidence to reduce its liability. Cf. Update Art, Inc. v. Modiin Publ’g, Ltd., 
    843 F.2d 67
    ,
    70, 72 (2d Cir. 1988); Blackman v. Hustler Mag., Inc., 
    800 F.2d 1160
    , 1163–64 (D.C. Cir. 1986).
    Under this legal regime, we can see no compensation-based rationale in the district
    court’s decision to rely on Core Health’s discovery conduct a second time by using it to increase
    Max Rack’s profits award. Core Health’s conduct did not make it more difficult for Max Rack
    to satisfy its burden to prove Core Health’s sales. Given the first discovery sanction, Core
    Health’s conduct instead made it more difficult for Core Health to prove its own costs—and so
    more likely that the jury would treat Core Health’s sales figure as its “profits” figure. Max Rack,
    moreover, cannot now complain that more cost evidence from Core Health would have revealed
    a lower cost figure (so a higher profits figure) because Max Rack itself moved to exclude any
    further cost evidence. The award was likely already inflated, so the decision to double it cannot
    be described as remedial. That decision effectively amounted to a second sanction for Core
    Health’s failure to produce its cost evidence. But such a “penalty” runs afoul of § 1117(a). See
    Skydive Ariz., 
    673 F.3d at 1115
    .
    That the missing discovery concerned Core Health’s costs also distinguishes the lone
    Fifth Circuit case on which the district court relied. See Bos. Pro. Hockey Ass’n, 
    597 F.2d at 77
    .
    In that case (which overturned an enhanced award), the court discussed a hypothetical defendant
    that refused to turn over sales data. See 
    id.
     Such stonewalling very well could prevent a plaintiff
    from proving the defendant’s total sales (and thus allow the defendant to hide its profits). But
    Max Rack had no problem identifying Core Health’s sales, and any stonewalling on costs would
    have harmed Core Health. So the district court abused its discretion by granting enhanced profits
    when its first sanction sufficed to remedy Core Health’s discovery shortcomings.
    Question 5: Did the district court mistakenly vacate the jury’s $1 million damages award?
    Max Rack argues that the district court erred in the other direction—by overturning the
    jury’s $1 million damages award. Although we review the court’s decision de novo, Balance
    Dynamics, 
    204 F.3d at 688
    , we agree that Max Rack failed to establish any recoverable damages.
    Apart from a “defendant’s profits,” the Lanham Act permits recovery of “any damages
    sustained by the plaintiff” from the infringement. 
    15 U.S.C. § 1117
    (a). That is, a plaintiff may
    Nos. 20-3598/3600        Max Rack, Inc. v. Core Health & Fitness, LLC, et al.                Page 28
    seek “pecuniary compensation” for the “loss, detriment, or injury” caused by the defendant’s
    infringing use of the plaintiff’s mark (or a deceptively similar one). Black’s Law Dictionary 466
    (4th ed. 1951); Webster’s New International Dictionary 664 (2d ed. 1934). To obtain damages,
    then, the plaintiff must do more than prove the “likelihood of confusion” that suffices to establish
    a defendant’s liability (and authorize an injunction). See Audi AG v. D’Amato, 
    469 F.3d 534
    ,
    542 (6th Cir. 2006). Such a likelihood may (or may not) generate the type of “injury” that the
    plaintiff must prove to receive this type of monetary relief.           See id.; 5 McCarthy, supra,
    §§ 30:2.50, 30:74.
    What kinds of injuries may a plaintiff recover for? Because the Lanham Act uses a
    common-law term (“damages”), we have looked to the common law of torts to help answer this
    question. See La Quinta, 
    603 F.3d at 342
    ; Broan Mfg. Co., Inc. v. Associated Distribs., Inc., 
    923 F.2d 1232
    , 1235 (6th Cir. 1991). Compensatory damages at common law generally sought to
    place plaintiffs in the “substantially equivalent” position that they would have been in if no tort
    had occurred. Restatement (Second) of Torts § 903 cmt. a (Am. L. Inst. 1979). A recoverable
    “loss” thus can take a variety of forms. A trademark owner might seek to recover profits that it
    has lost because a competitor used an infringing mark to poach sales to customers. See Balance
    Dynamics, 
    204 F.3d at
    690–91; 5 McCarthy, supra, § 30:79. If a holdover licensee continues to
    use a mark, the trademark owner might also seek the royalties that it would have earned on the
    licensee’s illicit sales under the licensing agreement. See La Quinta, 
    603 F.3d at
    341 & n.10;
    5 McCarthy, supra, § 30:86. Apart from these “lost profits,” a trademark owner might further
    seek to recover for the “lost goodwill” that arose when consumers bought the infringer’s inferior
    product and soured on the owner’s brand as a result. Balance Dynamics, 
    204 F.3d at
    690–91. Or
    a trademark owner might seek to recover the “damage control costs” that it incurred to reduce the
    harm from the infringer’s conduct—say, by spending money on advertisements clarifying that
    the owner has no affiliation with the infringer. See 
    id.
     at 691–92.
    To recover for these different losses, a plaintiff might need to prove different things.
    Before awarding “marketplace” damages like lost profits, for example, courts generally require a
    plaintiff to show that an infringing mark has actually confused some consumers about the mark’s
    affiliation with the plaintiff’s good (not just that it is likely to do so). See 
    id. at 691
    ; 5 McCarthy,
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.           Page 29
    supra, § 30:74 (citing cases). Because juries might have trouble quantifying these losses, this
    actual-confusion evidence helps ensure that they at least exist. See Balance Dynamics, 
    204 F.3d at 691
    . But other losses do not require this showing. Damage-control costs, for example, seek to
    ensure that actual confusion does not arise. So it would make little sense to require proof of
    actual confusion to recover these costs. See 
    id.
     Nor have we referred to an actual-confusion
    element when requiring a holdover licensee to pay the royalty (an objective measure) on
    infringing sales. See La Quinta, 
    603 F.3d at
    341–45. Lastly, if a plaintiff has obtained an award
    of the defendant’s profits (which is a proxy for what the plaintiff would have earned on the
    sales), the plaintiff must tie separate damages to a distinct harm and cannot seek a double
    recovery. See Nintendo of Am., Inc. v. Dragon Pac. Int’l, 
    40 F.3d 1007
    , 1011 (9th Cir. 1994); 5
    McCarthy, supra, § 30:73; cf. La Quinta, 
    603 F.3d at 345
    .
    These rules show that Max Rack failed to present enough evidence to recover damages
    under either of its infringement theories. Under Max Rack’s first theory, Core Health sold some
    271 Max Racks without authorization. Admittedly, the jury could have found that Max Rack
    suffered marketplace losses—for example, the royalty that it would have earned—if it believed
    that Core Health violated the agreement by selling those units. Before trial, however, Core
    Health paid Max Rack the profits Core Health earned on 24 of the units and its standard royalty
    for 238 more (all with interest). Max Rack does not claim that this payment failed to adequately
    remedy any marketplace injury. That leaves only the 9 machines that the jury could have found
    were sold to consumers as “Max Racks” but shipped as “Freedom Racks.” Although Core
    Health did not pay Max Rack for these sales, Max Rack’s profits award would have remedied
    any harm from them. A further award would have amounted to a double recovery.                See
    Nintendo, 
    40 F.3d at 1010
    .
    Max Rack’s damages claim fares no better when we turn to its second infringement
    theory: that Core Health sold an unknown number of Freedom Racks by keeping “Max Rack”
    references on its website. Max Rack all but concedes that it lacks evidence of actual consumer
    confusion from Core Health’s minor uses of the Max Rack name, which our cases suggest it
    needed to show. See Balance Dynamics, 
    204 F.3d at 691
    . Besides, Max Rack also fails to point
    us to any other evidence that these stray uses of its name caused it harm. Max Rack had not even
    Nos. 20-3598/3600      Max Rack, Inc. v. Core Health & Fitness, LLC, et al.           Page 30
    created a new “Max Rack” product by the time of trial, so it lost no sales from Core Health’s
    uses. Max Rack would have made no sales regardless of Core Health’s conduct. The Freedom
    Rack was also identical to the Max Rack and made by the same manufacturer. So Max Rack
    identifies no “goodwill” injury that could have resulted from an inferior product’s use of its
    name. Contra Broan Mfg., 
    923 F.2d at
    1238–39.
    Max Rack counters that the district court overturned its award solely on the ground that
    the company failed to prove actual confusion. Max Rack, 
    2020 WL 2128614
    , at *4–5. It argues
    that we should do away with any categorical actual-confusion requirement for proving
    marketplace damages because the Supreme Court in Romag recently did away with a similar
    categorical rule. See 140 S. Ct. at 1494–97. In the past, we had suggested that a plaintiff must
    prove that a defendant willfully infringed the plaintiff’s mark to obtain the defendant’s profits
    under § 1117(a). See Frisch’s Rests., Inc. v. Elby’s Big Boy of Steubenville, 
    849 F.2d 1012
    ,
    1015–16 (6th Cir. 1988). Romag rejected this categorical requirement as inconsistent with
    § 1117(a)’s text. 140 S. Ct. at 1494–95. The Court instead treated a defendant’s intent as only a
    “highly important” factor for a profits award. Id. at 1497. Max Rack asks us to treat actual
    confusion in the same way because § 1117(a) does not list actual confusion in the text either.
    But we need not decide how Romag affects the traditional actual-confusion prerequisite for
    marketplace damages. Even if we treated actual confusion as an “important” factor, id., Max
    Rack identifies nothing to show that Core Health’s stray references to “Max Rack” caused it
    harm. No matter the governing legal standard, Max Rack’s speculation of harm would not
    suffice to justify a damages award. See La Quinta, 
    603 F.3d at 342
    .
    Max Rack notes, however, that even our existing caselaw did not require it to prove
    actual confusion for some damages, such as the damage-control costs that we have mentioned.
    Balance Dynamics, 
    204 F.3d at
    689–91. But Max Rack makes no claim that it spent even a
    dollar seeking to mitigate the damages caused by Core Health’s conduct. And Max Rack does
    not claim that it suffered any other type of loss for which actual consumer confusion is
    irrelevant.
    Max Rack thus turns to out-of-circuit precedent.      It argues that some courts have
    presumed actual confusion (and shifted the burden to a defendant to disprove it) when the
    Nos. 20-3598/3600         Max Rack, Inc. v. Core Health & Fitness, LLC, et al.        Page 31
    defendant has intentionally deceived the public by using a deceptive mark or comparative false
    advertising. See Porous Media Corp. v. Pall Corp., 
    110 F.3d 1329
    , 1336 (8th Cir. 1997); Res.
    Devs., Inc. v. Statue of Liberty-Ellis Island Found., Inc., 
    926 F.2d 134
    , 140 (2d Cir. 1991).
    Highlighting the jury’s intentional-infringement finding in this case, Max Rack claims that the
    presumption of actual confusion should apply here. But the jury’s intent finding rests most
    reasonably on Core Health’s continued sales of Max Racks in violation of the licensing
    agreement, and Max Rack already received full compensation for those sales. No reasonable
    jury, by contrast, could have found that Core Health’s few extraneous references to the Max
    Rack mark in its Freedom Rack advertisements were intentional (rather than accidental). See
    Res. Devs., 
    926 F.2d at
    140–41. So we need not consider how this intent element should affect
    damage awards.
    Max Rack ends with a procedural claim. It criticizes the district court for granting Core
    Health judgment as a matter of law on the damages award because Core Health had only sought
    a new trial under Rule 59. True, a court that finds a verdict excessive under Rule 59 generally
    must give the plaintiff a choice between a remittitur and a new trial. See Farber, 
    917 F.2d at
    1395–96; 11 Wright et al., supra, § 2815, at 208–10. Max Rack did not receive that choice: the
    court effectively granted a remittitur (down to zero) without a new-trial option. But courts have
    also held that they may reduce or eliminate a monetary award without permitting a new trial
    when the award rests on a legal error (not on the discretionary ground that the verdict was
    excessive). See Roush, 10 F.3d at 400; Westchester Fire Ins. Co. v. Hanley, 
    284 F.2d 409
    , 418
    (6th Cir. 1960); Niagara Fire Ins. Co. v. Bryan & Hewgley, Inc., 
    195 F.2d 154
    , 157 (6th Cir.
    1952); see also Saccameno v. U.S. Bank Nat’l Ass’n, 
    943 F.3d 1071
    , 1091–92 (7th Cir. 2019);
    Johansen v. Combustion Eng’g, Inc., 
    170 F.3d 1320
    , 1330–31 (11th Cir. 1999). Here, the district
    court found a legal error—that no evidence undergirded Max Rack’s claimed injury. So it could
    grant judgment outright.
    Question 6: Did the district court wrongly grant Max Rack an award of attorney’s fees?
    That leaves Core Health’s challenge to Max Rack’s award of some $381,808.32 in
    attorney’s fees. We review this award—like the district court’s award of double profits—under a
    deferential abuse-of-discretion standard. See Eagles, Ltd. v. Am. Eagle Found., 
    356 F.3d 724
    ,
    Nos. 20-3598/3600         Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 32
    728 (6th Cir. 2004); see also Highmark Inc. v. Allcare Health Mgmt. Sys., Inc., 
    572 U.S. 559
    ,
    563 (2014). We again conclude that the court abused its discretion.
    The Lanham Act gives a district court discretion to compel the losing party to pay the
    winning party’s attorney’s fees, but only in certain cases. It provides: “The court in exceptional
    cases may award reasonable attorney fees to the prevailing party.” 
    15 U.S.C. § 1117
    (a). The use
    of the word “exceptional” limits the district court’s discretion by reserving attorney’s fees for
    “rare” or “unusual” cases. Webster’s, supra, at 889; 5 Oxford English Dictionary 498–99 (2d ed.
    1989); Octane Fitness, LLC v. ICON Health & Fitness, Inc., 
    572 U.S. 545
    , 553–54 (2014). The
    Act thus departs from, say, the Copyright Act, which permits fee awards without any
    exceptional-circumstances limit. See Kirtsaeng v. John Wiley & Sons, Inc., 
    579 U.S. 197
    , 202
    (2016).
    When interpreting a similar limit on attorney’s fees in the patent laws, the Supreme Court
    has explained that a case might qualify as “rare” or “unusual” for different reasons. See Octane,
    572 U.S. at 554. The plaintiff might have an unusually strong (or unusually weak) case on the
    merits. See id. Blatant trademark infringement might justify an award for the plaintiff, whereas
    a frivolous infringement claim might justify one for the defendant. See id. Or the losing party
    might have litigated the case in an unreasonable manner—for example, by requesting costly
    discovery to coerce a settlement despite the weakness of its claims. See id. In the end, however,
    the statute leaves it for the district court to decide under the totality of the circumstances whether
    the case before it has the “rare” qualities that distinguish it from a typical case. See id. Circuit
    courts have followed this same approach for the Lanham Act’s similar language. See Evoqua
    Water Techs., LLC v. M.W. Watermark, LLC, 
    940 F.3d 222
    , 235 (6th Cir. 2019); see also LHO
    Chi. River, L.L.C. v. Perillo, 
    942 F.3d 384
    , 388 (7th Cir. 2019) (citing cases); 5 McCarthy,
    supra, § 30:99 n.9 (same).
    Judged under this framework, nothing about this case looks unusual. To begin with, Max
    Rack did not have a noticeably stronger “litigating position.” Octane, 572 U.S. at 554. That
    point is made clear by considering its two infringement theories. The evidence for its first theory
    (that Core Health’s post-agreement Max Rack sales violated the agreement) could be described
    as “thin.” See Visible Sys. Corp., 
    551 F.3d at 82
    . Except for the limited number of units that
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.                Page 33
    Core Health sold outside the selloff window (and for which it paid compensation before trial),
    Core Health’s liability rested on the claim that Core Health began to make those units after the
    agreement expired. But that claim was supported only by circumstantial evidence falling close
    to the hazy border dividing what a jury can (and cannot) reasonably find. Cf. Evoqua Water
    Techs., 940 F.3d at 235. Core Health’s liability on Max Rack’s second theory (that Core Health
    used the Max Rack mark to sell Freedom Racks) could be described as equally “thin.” See
    Visible Sys. Corp., 
    551 F.3d at 82
    . This theory rested on isolated and likely accidental references
    to Max Rack’s mark on Core Health’s website. As we have said, the collective references barely
    established the likelihood of confusion that could support an injunction, and Core Health’s
    marketing director voluntarily removed all references she learned of during her deposition. As
    the district court recognized, moreover, Max Rack suffered no damages from the references,
    further undermining any entitlement to fees.
    Likewise, Core Health did not litigate this case in an “unreasonable manner.” Octane,
    572 U.S. at 554. To be sure, the district court noted that Core Health failed to turn over its costs
    data. But that shortcoming largely harmed Core Health, not Max Rack, and the district court did
    not mention the discovery issue when awarding fees. See Max Rack, 
    2020 WL 2128614
    , at *7.
    Max Rack also did not have unclean litigating hands. Its witnesses peppered the record with
    references to the addendum and complaints about Core Health’s allegedly unfair refusal to pay
    royalties in perpetuity. Yet its counsel admitted that “this lawsuit isn’t about the addendum.”
    Tr., R.102, PageID 2632. In short, if this case warranted attorney’s fees, we fail to see what case
    would not.
    For its part, the district court relied on precedent that it read as creating a bright-line rule
    permitting fees if a defendant’s intentional infringement continued after the plaintiff sued. See
    Max Rack, 
    2020 WL 2128614
    , at *7. This rule applied here, the court reasoned, because the jury
    found Core Health’s conduct to be intentional and because Core Health admitted that its
    infringement continued even through the trial. See 
    id.
     Its reasoning rested on legal error. The
    district court’s rigid rule conflicts with the Supreme Court’s more flexible totality-of-the-
    circumstances approach to this inquiry. See Octane, 572 U.S. at 554. Perhaps unsurprisingly,
    not one of the cases that the court cited for this rule mentioned Octane. See Zeroez Franchising
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.             Page 34
    Sys., Inc. v. Distinctive Cleaning, Inc., 
    103 F. Supp. 3d 1032
    , 1050 (D. Minn. 2015); Ohio State
    Univ. v. Skreened Ltd., 
    16 F. Supp. 3d 905
    , 921 (S.D. Ohio 2014); Ramada Franchise Sys., Inc.
    v. Boychuk, 
    283 F. Supp. 2d 777
    , 787, 792–93 (N.D.N.Y. 2003). In addition, although the jury
    found intentional infringement, that finding reasonably rested on Core Health’s unauthorized
    sales of Max Racks. That conduct ended well before the trial. Lastly, the district court’s
    conclusion partially relied on OC Fitness’s alleged infringement of the Max Rack mark. See
    Max Rack, 
    2020 WL 2128614
    , at *7 n.4. As we have explained, however, that conduct cannot
    be held against Core Health at all, let alone provide the grounds for an attorney’s fees award.
    * * *
    In sum, we affirm in part, reverse in part, and remand. We affirm the district court’s
    denial of Core Health’s motion for judgment as a matter of law or for a new trial. We also affirm
    the $250,000 profits award and the court’s rejection of the $1 million damages award. But we
    reverse the court’s decision to double the profits award and its decision to grant Max Rack
    attorney’s fees. We thus remand for the court to enter a new judgment consistent with this
    opinion.
    Nos. 20-3598/3600       Max Rack, Inc. v. Core Health & Fitness, LLC, et al.                Page 35
    _____________________________________________________
    CONCURRING IN PART AND DISSENTING IN PART
    _____________________________________________________
    COLE, Circuit Judge, concurring in part and dissenting in part. I would affirm the
    district court in full. Therefore, I respectfully dissent from the majority’s treatment of the district
    court’s resolution of the profits award and attorney’s fees.
    I.
    A. Disgorged Profits Award
    
    15 U.S.C. § 1117
    (a) grants district courts the power, based on principles of equity, to
    award disgorged profits and to increase or decrease a jury’s monetary award. The statute “grants
    a district court a great deal of discretion in fashioning an appropriate remedy in cases of
    trademark infringement.” U.S. Structures, Inc. v. J.P. Structures, Inc., 
    130 F.3d 1185
    , 1191 (6th
    Cir. 1997). Here, the district court did not abuse that broad discretion when it increased the
    jury’s profits award.
    Enhancement of profits under § 1117(a) is used to “provide proper redress to an
    otherwise undercompensated plaintiff where imprecise damage calculations fail to do justice,
    particularly where the imprecision results from defendant’s conduct.”           La Quinta Corp. v.
    Heartland Props. LLC, 
    603 F.3d 327
    , 343 (6th Cir. 2010). That’s exactly what the district court
    did here. As a result of Core Health’s mismanagement and failures at the discovery stage, the
    profits calculation made by the jury was imprecise. Specifically, the only evidence that Core
    Health presented to prove costs was a summary spreadsheet of the “standard cost” of goods with
    no information about the actual cost of the products. And the only witness that Core Health had
    testify about the cost spreadsheet could not fill these gaps. The witness had not participated in
    the calculation of those costs. For these reasons, the district court enhanced the profits award to
    ensure that Max Rack was justly compensated, despite the imprecision resulting from Core
    Health’s conduct.
    Nos. 20-3598/3600          Max Rack, Inc. v. Core Health & Fitness, LLC, et al.           Page 36
    Further, Core Health argues on appeal that the district court abused its discretion by
    doubling the jury’s profits award because any doubt about its costs could only decrease the
    award. But it is equally possible that the jury overestimated the costs based on the limited
    summaries Core Health provided. Core Health urged the court to accept its “standard cost”
    evidence. But Max Rack disputed that evidence, and it would not be an abuse of discretion for
    the court to accept Max Rack’s estimation of costs.
    Ultimately, I would affirm the district court’s decision to enhance the profits award.
    Enhancement of profits was necessary to provide proper redress to Max Rack. For much the
    same reasons, I also do not consider it an abuse of discretion that the district court did not vacate
    the jury’s profits award.
    B. Attorney’s Fees
    A district court may award attorney’s fees in “exceptional cases.” 
    15 U.S.C. § 1117
    (a).
    Courts often consider a trademark infringement case exceptional when the infringing actions
    were intentional, and the infringement continues until the date of trial. See, e.g., Ohio State
    Univ. v. Skreened Ltd., 
    16 F. Supp. 3d 905
    , 921 (S.D. Ohio 2014); Ramada Franchise Sys., Inc.
    v. Boychuk, 
    283 F. Supp. 2d 777
    , 787 (N.D.N.Y. 2003); Zeroez Franchising Sys., Inc. v.
    Distinctive Cleaning, Inc., 
    103 F. Supp. 3d 1032
    , 1050 (D. Minn. 2015).
    Here, the jury found that the trademark infringement was intentional. And evidence
    suggested that this infringement continued for four years, up until the time of trial. Kirt Moritz
    navigated to the Core Health webpage during the lunch break of the first day of trial. He found
    that the assembly manual still referenced the Max Rack, over four years after the license expired.
    And on the second day of trial, even after he had testified about the reference the prior afternoon,
    the reference to the Max Rack was still on the Core Health website.
    Core Health continued to leave lingering references to the Max Rack on its website
    despite repeated warnings. Core Health should have been on notice of its infringing behavior
    after Max Rack filed the lawsuit. At the very least, Core Health must have been on notice when
    its marketing director was informed, during her deposition, that the website still referenced the
    Nos. 20-3598/3600        Max Rack, Inc. v. Core Health & Fitness, LLC, et al.          Page 37
    Max Rack. Despite the fact it would have been a simple fix, Core Health refused to remedy the
    problem for four years.
    Because Core Health continued infringing on the Max Rack trademark for years, it was
    not an abuse of discretion for the district court to award attorney’s fees.
    II.
    I would affirm the district court’s decisions in full and therefore respectfully dissent in
    part from the majority opinion.