Amicus Curiae Brief — RiseandShine Corporation, dba Rise Brewing, Petitioner v. PepsiCo, Inc.

No. 24-1016

In the

Supreme Court of the United States

RISEANDSHINE CORPORATION,

DBA RISE BREWING,

Petitioner,

v.

PEPSICO, INC.,

Respondent.

On Writ of Certiorari to the United States

Court of A ppeals for the Second Circuit

TRADEMARK LAW SCHOLARS

AS AMICI CURIAE

IN SUPPORT OF NEITHER PARTY

Rebecca Tushnet

Counsel of Record

Frank Stanton Professor

of the First Amendment

Harvard Law School

1563 Massachusetts Avenue,

Hauser 520

Cambridge, MA 02138

(703) 593-6759

rtushnet@law.harvard.edu

Attorney for Amici Curiae

121243

(800) 274-3321 • (800) 359-6859

i

TABLE OF CONTENTS

Page

TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . i

TABLE OF CITED AUTHORITIES . . . . . . . . . . . . . . . ii

INTEREST OF AMICI . . . . . . . . . . . . . . . . . . . . . . . . . . 1

SUMMARY OF ARGUMENT . . . . . . . . . . . . . . . . . . . . 1

ARGUMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

I.

The Development of the Multifactor Test . . . . . . 3

II. Strength Specifically is a Mixed Question . . . . . 9

III. Adjudicative Facts Bear on Likelihood

of Confusion But Do Not Map Directly To

the Likelihood of Confusion Factors . . . . . . . . . 16

IV. The Multifactor Test Creates a Variably

Mixed Question of Law and Fact . . . . . . . . . . . . 24

CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

APPENDIX — LIST OF AMICI CURIAE . . . . . . . . . 1a

ii

TABLE OF CITED AUTHORITIES

Cases

Page

1-800 Contacts, Inc. v. JAND, Inc.,

119 F.4th 234 (2d Cir. 2024) . . . . . . . . . . . . . . . . . . . . 26

1-800 Contacts, Inc. v. Lens.com, Inc.,

722 F.3d 1229 (10th Cir. 2013) . . . . . . . . . . . . . . . . . . 26

A & H Sportswear, Inc. v. Victoria’s Secret Stores, Inc.,

237 F.3d 198 (3d Cir. 2000) . . . . . . . . . . . . . . . . . . . . . 15

Abercrombie & Fitch Co. v. Hunting World, Inc.,

537 F.2d 4 (2d Cir. 1976) . . . . . . . . . 2, 10, 12, 13, 29, 30

Amazing Spaces, Inc. v. Metro Mini Storage,

608 F.3d 225 (5th Cir. 2010) . . . . . . . . . . . . . . . . . . . . 10

AmBrit, Inc. v. Kraft, Inc.,

812 F.2d 1531 (11th Cir. 1986) . . . . . . . . . . . . . . . . . . 21

Apex Bank v. CC Serve Corp.,

156 F.4th 1230 (Fed. Cir. 2025) . . . . . . . . . . . . . . . . . 18

Appliance Liquidation Outlet, L.L.C. v.

Axis Supply Corp.,

105 F.4th 362 (5th Cir. 2024) . . . . . . . . . . . . . . . . 21-22

Blau Plumbing, Inc. v. S.O.S. Fix-It, Inc.,

781 F.2d 604 (7th Cir. 1986) . . . . . . . . . . . . . . . . . . . . 11

iii

Cited Authorities

Page

Booking.com B.V. v. Matal,

278 F. Supp. 3d 891 (E.D. Va. 2017) . . . . . . . . . . . . . . 11

Bose Corp. v. Consumers Union of U.S., Inc.,

466 U.S. 485 (1984) . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Brookfield Commc’ns, Inc. v. W. Coast Ent. Corp.,

174 F.3d 1036 (9th Cir. 1999) . . . . . . . . . . . . . . . . 10, 21

Bufkin v. Collins,

604 U.S. 369 (2025) . . . . . . . . . . . . . . . . . . . . . . . . 25, 28

Campbell v. Acuff-Rose Music, Inc.,

510 U.S. 569 (1994) . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Century 21 Real Est. Corp. v. Lendingtree, Inc.,

425 F.3d 211 (3d Cir. 2005) . . . . . . . . . . . . . . . . . . . . . . 5

Checkpoint Sys., Inc. v.

Check Point Software Techs., Inc.,

269 F.3d 270 (3d Cir. 2001) . . . . . . . . . . . . . . . . . . . . . 21

Cohn v. Petsmart, Inc.,

281 F.3d 837 (9th Cir. 2002) . . . . . . . . . . . . . . . . . . . . 21

Conopco, Inc. v. May Dep’t Stores Co.,

46 F.3d 1556 (Fed. Cir. 1994) . . . . . . . . . . . . . . . . . . . . 4

Craft Smith, LLC v. EC Design, LLC,

969 F.3d 1092 (10th Cir. 2020) . . . . . . . . . . . . . . . . . . 23

iv

Cited Authorities

Page

Dellums v. U.S. Nuclear Regul. Comm’n,

863 F.2d 968 (D.C. Cir. 1988) . . . . . . . . . . . . . . . . . . . . 6

Deltona Transformer Co. v. NOCO Co.,

--- F.4th ----, No. 24-13590, 2026 WL 2236806

(11th Cir. Aug. 4, 2026) . . . . . . . . . . . . . . . . . . . . . . . . 25

Door Sys., Inc. v. Pro-Line Door Sys., Inc.,

83 F.3d 169 (7th Cir. 1996) . . . . . . . . . . . . . . . . . . . . . 16

Dunagin v. City of Oxford,

718 F.2d 738 (5th Cir. 1983) . . . . . . . . . . . . . . . . . . . . . 6

Fisher Stoves, Inc. v. All Nighter Stove Works, Inc.,

626 F.2d 193 (1st Cir. 1980) . . . . . . . . . . . . . . . . . . . . 22

Fla. Int’l Univ. Bd. of Trs. v. Fla. Nat’l Univ., Inc.,

830 F.3d 1242 (11th Cir. 2016) . . . . . . . . . . . . . . . . 18-19

George & Co. LLC v. Imagination Ent. Ltd.,

575 F.3d 383 (4th Cir. 2009) . . . . . . . . . . . . . . . . . . . . 18

Google LLC v. Oracle Am., Inc.,

593 U.S. 1 (2021) . . . . . . . . . . . . . . . . . . . . . . . . . . 26, 29

Hana Financial, Inc. v. Hana Bank,

574 U.S. 418 (2015) . . . . . . . . . . . . . . . . . . 18, 19, 20, 29

Iberia Foods Corp. v. Romeo,

150 F.3d 298 (3d Cir. 1998) . . . . . . . . . . . . . . . . . . . . . . 5

v

Cited Authorities

Page

In re Chippendales USA, Inc.,

622 F.3d 1346 (Fed. Cir. 2010) . . . . . . . . . . . . . . . 11, 12

Indianapolis Colts, Inc. v.

Metro. Balt. Football Club Ltd. P’ship,

34 F.3d 410 (7th Cir. 1994) . . . . . . . . . . . . . . . . . . . . . . 5

Int’l Info. Sys. Sec. Certification Consortium, Inc. v.

Sec. Univ., LLC,

823 F.3d 153 (2d Cir. 2016) . . . . . . . . . . . . . . . . . . . . . . 5

Int’l Kennel Club of Chicago, Inc. v. Mighty Star, Inc.,

846 F.2d 1079 (7th Cir. 1988) . . . . . . . . . . . . . . . . . . . 22

Jack Daniel’s Props., Inc. v. VIP Prods. LLC,

599 U.S. 140 (2023) . . . . . . . . . . . . . . . . . . . . . 4, 6, 8, 28

Jim S. Adler, P.C. v. McNeil Consultants, L.L.C.,

10 F.4th 422 (5th Cir. 2021) . . . . . . . . . . . . . . . . . . . . 26

Kos Pharms., Inc. v. Andrx Corp.,

369 F.3d 700 (3d Cir. 2004) . . . . . . . . . . . . . . . . . . . . . 17

Labrador Software, Inc. v. Lycos, Inc.,

32 F. Supp. 2d 31 (D. Mass. 1999) . . . . . . . . . . . . . . . 15

Lerner & Rowe PC v. Brown Engstrand & Shely LLC,

119 F.4th 711 (9th Cir. 2024),

cert. denied, 145 S. Ct. 2732 (Mem) (2025) . . . . 25-26

vi

Cited Authorities

Page

Lockhart v. McCree,

476 U.S. 162 (1986) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Louis Vuitton Malletier S.A. v.

Haute Diggity Dog, LLC,

507 F.3d 252 (4th Cir. 2007) . . . . . . . . . . . . . . . . . . . . 15

Merriam-Webster, Inc. v. Random House, Inc.,

35 F.3d 65 (2d Cir. 1994) . . . . . . . . . . . . . . . . . . . . . . . 21

Mushroom Makers, Inc. v. R. G. Barry Corp.,

441 F. Supp. 1220 (S.D.N.Y. 1977),

aff’d, 580 F.2d 44 (2d Cir. 1978) . . . . . . . . . . . . . . . . . 21

Network Automation, Inc. v.

Advanced Systems Concepts, Inc.,

638 F.3d 1137 (9th Cir. 2011) . . . . . . . . . . . . . . . . . . . 26

New Kids on the Block v. News Am. Publ’g, Inc.,

971 F.2d 302 (9th Cir. 1992) . . . . . . . . . . . . . . . . . . . . . 5

Nutri/Sys., Inc. v. Con-Stan Indus.,

809 F.2d 601 (9th Cir. 1987) . . . . . . . . . . . . . . . . . . . . 10

One Indus., LLC v. Jim O’Neal Distrib., Inc.,

578 F.3d 1154 (9th Cir. 2009) . . . . . . . . . . . . . . . . . . . 18

Paddington Corp. v.

Attiki Importers & Distributors, Inc.,

996 F.2d 577 (2d Cir. 1993) . . . . . . . . . . . . . . . . . . . . . 23

vii

Cited Authorities

Page

Prestonettes, Inc., v. Coty,

264 U.S. 359 (1924) . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Samara Bros. v. Wal-Mart Stores, Inc.,

165 F.3d 120 (2d Cir. 1998),

rev’d, 529 U.S. 205 (2000) . . . . . . . . . . . . . . . . . . . . . 7, 8

Savannah Coll. of Art & Design, Inc. v.

Sportswear, Inc.,

983 F.3d 1273 (11th Cir. 2020) . . . . . . . . . . . . . . . . . . 11

Sorensen v. WD-40 Co.,

792 F.3d 712 (7th Cir. 2015) . . . . . . . . . . . . . . . . . 16, 22

Sterling Jewelers, Inc. v. Artistry Ltd.,

896 F.3d 752 (6th Cir. 2018) . . . . . . . . . . . . . . . . . . . . 24

Syntex Lab’ys, Inc. v. Norwich Pharmacal Co.,

437 F.2d 566 (2d Cir. 1971) . . . . . . . . . . . . . . . . . . . . . 17

Therma-Scan, Inc. v. Thermoscan, Inc.,

295 F.3d 623 (6th Cir. 2002) . . . . . . . . . . . . . . . . . . . . 22

TikTok v. Garland,

604 U.S. 56 (2025) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Tommy Hilfiger Licensing, Inc. v. Nature Labs, LLC,

221 F. Supp. 2d 410 (S.D.N.Y. 2002) . . . . . . . . . . . . . 15

Two Pesos, Inc. v. Taco Cabana, Inc.,

505 U.S. 763 (1992) . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

viii

Cited Authorities

Page

U.S. Bank N.A. v. Village at Lakeridge, LLC,

583 U.S. 387 (2018) . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

United States Patent and Trademark Office v.

Booking.com B. V.,

915 F.3d 171 (4th Cir. 2019),

aff’d, 591 U.S. 549 (2020) . . . . . . . . . . . . . . . . . . . . . . 11

United States v. Gould,

536 F.2d 216 (8th Cir. 1976) . . . . . . . . . . . . . . . . . . . . 10

VIP Prods. LLC v. Jack Daniel’s Props. Inc.,

No. CV-14-02057, 2025 WL 275909

(D. Ariz. Jan. 23, 2025) . . . . . . . . . . . . . . . . . . . . . . . . 28

VIP Prods., LLC v. United States,

185 F.4th 919 (9th Cir. 2026) . . . . . . . . . . . . . . . . . . . 28

Wal-Mart Stores, Inc. v. Samara Bros.,

529 U.S. 205 (2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Water Pik, Inc. v. Med-Sys., Inc.,

726 F.3d 1136 (10th Cir. 2013) . . . . . . . . . . . . . . . . . . 16

Wynn Oil Co. v. Am. Way Serv. Corp.,

943 F.2d 595 (6th Cir. 1991) . . . . . . . . . . . . . . . . . . . . 22

Yankee Pub. Inc. v. News Am. Pub. Inc.,

809 F. Supp. 267 (S.D.N.Y. 1992) . . . . . . . . . . . . . . . . 17

ix

Cited Authorities

Page

Statutes and Other Authorities

U.S. Const. amend. I . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Barton Beebe & C. Scott Hemphill, The Scope of

Strong Marks: Should Trademark Law Protect

the Strong More than the Weak?, 92 N.Y.U. L.

Rev. 1339 (2017) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Barton Beebe, An Empir ical Study of the

Multifactor Tests for Trademark Infringement,

94 Calif. L. Rev. 1581 (2006) . . . . . . . . . . . . . . . . . . 4, 23

Barton Beebe, The Semiotic Analysis of Trademark

Law, 51 UCLA L. Rev. 621 (2004) . . . . . . . . . . . . . . . 19

Robert G. Bone, Taking the Confusion Out of

“Likelihood of Confusion”: Toward A More

Sensible Approach to Trademark Infringement,

106 Nw. U. L. Rev. 1307 (2012) . . . . . . . . . . . . . . . . 3, 23

Graeme B. Dinwoodie & Dev S. Gangjee, The Image

of the Consumer in EU Trade Mark Law, in The

Images of the Consumer in EU Law: Legislation,

Free Movement and Competition Law 339

(Dorota Leczykiewicz & Stephen Weatherill eds.,

2016) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-17

Graeme B. Dinwoodie, Trade Mark Law As

A Normative Project, 2023 Sing. J. Legal

Stud. 305 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11, 12, 17

x

Cited Authorities

Page

Timothy B. Dyk, The Role Of Non-Adjudicative

Facts In Judicial Decisionmaking, 76 Stan. L.

Rev. Online 10 (2023) . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Fifth Cir. Civil Jury Instructions, Trademark

Infringement: Specific Preliminary Instructions

To The Jury ch. 14.11 (2024) https://www.lb5.

uscourts.gov/juryinstructions/Fifth/2024Civil_

Ch14_Trademarks.pdf (last visited Sept. 9,

2026) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Michael Grynberg, The Consumer’s Duty of Care

in Trademark Law, in Research Handbook on

Trademark Law Reform (Graeme B. Dinwoodie

& Mark D. Janis eds., 2021) . . . . . . . . . . . . . . . . . . . . 20

Michael Grynberg, Trademark Litigation As

Consumer Conflict, 83 N.Y.U. L. Rev. 60 (2008) . . . 17

Justin Hughes, The Respective Roles of Judges and

Juries in Copyright Fair Use, 58 Hous. L. Rev.

327 (2020) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Ja c o b Ja c o b y & Way n e D. Ho y e r, T h e

Comprehension and Miscomprehension of Print

Communications (1987) . . . . . . . . . . . . . . . . . . . . . . . . 4

Jacob Jacoby et al., Miscomprehension of Televised

Communications (1980) . . . . . . . . . . . . . . . . . . . . . . . . 4

Richard L. Kirkpatrick & Vijay K. Toke, Likelihood

of Confusion in Trademark Law (2d ed. 2026) . . . . 27

xi

Cited Authorities

Page

Chiranjeev S. Kohli & Rajneesh Suri, Brand

Names That Work: A Study of the Effectiveness

of Different Types of Brand Names, 10 Mktg.

Mgmt. J. 112 (2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Thomas R. Lee, Eric D. DeRosia & Glenn L.

Christensen, An Empirical and Consumer

Ps ych ol o g y An a lys i s of Tr a d e m a r k

Distinctiveness, 41 Ariz. St. L. J. 1033 (2009) . . . . . 12

John P. Liefeld, How Surveys Overestimate

the Likelihood of Consumer Confusion, 93

Trademark Rep. 939 (2003) . . . . . . . . . . . . . . . . . . . . . 5

Daryl Lim, Trademark Confusion Revealed: An

Empirical Analysis, 71 Am. U. L. Rev. 1285

(2022) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5

Jake Linford, Are Trademarks Ever Fanciful?, 105

Geo. L.J. 731 (2017) . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Jake Linford, The False Dichotomy Between

Suggestive and Descriptive Trademarks, 76 Ohio

St. L.J. 1367 (2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Glynn S. Lunney, Jr., Trademark’s Judicial DeEvolution: Why Courts Get Trademark Cases

Wrong Repeatedly, 106 Calif. L. Rev. 1195

(2018) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

xii

Cited Authorities

Page

Dustin Marlan, The Arbitrary Myth: Challenging

Trademark Law’s Elevation of Arbitrary Marks,

41 Berkeley Tech. L.J. 555 (2026) . . . . . . . . . . . . . . . 13

J. Thomas McCarthy, McCarthy on Trademarks

and Unfair Competition (5th ed. 2022) . . . . 12, 14, 16,

17, 19, 21, 23, 27

William McGeveran & Mark P. McKenna, Confusion

Isn’t Everything, 89 Notre Dame L. Rev. 253

(2013) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

William McGeveran, The Imaginary Trademark

Parody Crisis (and the Real One), 90 Wash. L.

Rev. 713 (2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Ninth Cir. Jury Instructions Comm., Manual of

Model Civil Jury Instructions for the District

Courts of the Ninth Circuit § 15.19 (2025 ed.,

current through June 2026) (last visited Sept. 9,

2026) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Note, Fanciful Failures: Keeping Nonsense Marks

Off the Trademark Register, 134 Harv. L. Rev.

1804 (2021) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Malla Pollack, Callmann on Unfair Competition,

Trademarks and Monopolies (4th ed. 2026) . . . . . . 27

Jeremy N. Sheff, Marks, Morals, and Markets, 65

Stan. L. Rev. 761 (2013) . . . . . . . . . . . . . . . . . . . . . . . . 23

xiii

Cited Authorities

Page

Jeremy N. Sheff, Reverse Confusion and the

Justification of Trademark Protection, 30 Geo.

Mason L. Rev. 123 (2022) . . . . . . . . . . . . . . . . . . . 15, 23

Jeremy N. Sheff, Veblen Brands, 96 Minn. L. Rev.

769 (2012) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Hannah Walser, Interpretive Facts: Textualism,

Empiricism, and the Law-Fact Divide, 101 Ind.

L.J. 1071 (2026) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Kenji Yoshino, Appellate Deference in the Age of

Facts, 58 Wm. & Mary L. Rev. 251 (2016) . . . . . . 7, 27

1

INTEREST OF AMICI

Amici, listed in the Appendix, are scholars who teach,

write, and practice in the field of trademark law. We have

no direct interest in the outcome of this case.1

SUMMARY OF ARGUMENT

The issue before the Court is whether trademark

strength in a likely confusion analysis is a question of

fact or a question of law, but this question actually bears

on the confusion inquiry as a whole. Sometimes, factual

issues predominate in this consideration, but sometimes

legal issues do.

This conclusion may sound surprising given the

phrase “likelihood of confusion,” which sounds empirical.

The full version of the standard—likely confusion among

a substantial number of reasonable consumers—clarifies

that normative legal judgments (what is substantial,

what is reasonable, and what constitutes confusion) are

crucial to infringement inquiries. Infringement inquiries

therefore are mixed questions of law and fact in which

legal analysis can often predominate.

The hallmark of fact finding that deserves appellate

deference and clear-error review is individualized

adjudication. The multifactor framework to assess

likelihood of confusion that courts have regularly used

in the 80 years since the enactment of the Lanham Act

1. Counsel for the parties did not author this brief in whole or

in part. No persons other than Amici Curiae contributed money

to fund preparation or submission of this brief.

2

involves propositions that sound factual, but that have

never—not in this case, nor in any other—been tested by

factfinding at the district court level.

This includes the question of trademark strength. The

theory that “suggestive marks are conceptually stronger

than descriptive marks with secondary meaning, thus

favoring the plaintiff more in the confusion inquiry,” is

not an “adjudicative fact” found by the factfinder based

on individualized evidence.

Contrary to some suggestions in Petitioner’s brief,

conceptual strength inquiries do not involve direct

investigation of consumer understanding of the extent to

which a word functions to indicate source when applied to a

product. Conceptual strength is partly empirical (involving

an assessment of, inter alia, dictionary definitions and

other marketplace uses) and partly a matter of prediction

about whether people will need to use “imagination and

thought” to connect a word with the goods or services at

issue. No trademark is “suggestive” or “descriptive” as a

state of nature; Judge Friendly accepted that his division

was “rough[]” at best. Abercrombie & Fitch Co. v. Hunting

World, Inc., 537 F.2d 4, 9 (2d Cir. 1976).

Furthermore, the propositions that suggestive marks

are “stronger” than descriptive marks with secondary

meaning, and thus that, all else being equal, a mark’s

status as suggestive makes confusion more likely than if

it were descriptive, are broad claims about reality.

Courts evolved the multifactor confusion test over time,

and they are still adding variations for new situations—

as they should be. As with many torts, the existence of

3

trademark infringement (given a set of adjudicative facts)

is a classic mixed question of law and fact. No resolution

of this case should treat either likelihood of confusion or

trademark strength as a purely factual question.

ARGUMENT

I.

The Development of the Multifactor Test

The Lanham Act does not specify a test for determining

likelihood of confusion. William McGeveran & Mark P.

McKenna, Confusion Isn’t Everything, 89 Notre Dame L.

Rev. 253, 255 (2013). The choice of what factors to consider

in a given class of cases is itself a legal one. While courts

have converged in most cases on a multifactor test, they

have also altered or departed from it when appropriate.

The multifactor test developed for dealing with

noncompeting goods, where sales substitution due to

confusion was not the primary risk. The standard for

infringement by a direct competitor was simpler: were the

goods and the marks identical? The decision to expand the

multifactor test to competing goods cases was undertaken

by the lower courts mostly without discussion. 2

The only element of the multifactor test that is directly

about consumer confusion is the “actual confusion” factor,

which takes account of complaints, reports, surveys, and

other direct evidence of consumers’ beliefs. The remaining

2. See McGeveran & McKenna, Confusion, supra, at 254;

Robert G. Bone, Taking the Confusion Out of “Likelihood of

Confusion”: Toward A More Sensible Approach to Trademark

Infringement, 106 Nw. U. L. Rev. 1307, 1331-34 & n.164 (2012).

4

factors provide what courts believe to be circumstantial

evidence about the likelihood of confusion. They exist

because most litigated cases don’t involve evidence of

actual confusion, which is thought to be hard to surface

in court even when it exists. 3

Additionally, “actual confusion” evidence itself is

not dispositive about whether confusion is likely among

a substantial number of reasonable consumers. Most

people are confused about many things most of the

time; individual confusion reports may therefore reflect

idiosyncratic, unreasonable, or fleeting beliefs that do not

justify liability.4 Consumer surveys conducted for litigation

purposes are expensive and highly manipulable, so they

may not be good quality evidence either. 5 The multifactor

3. See Barton Beebe, An Empirical Study of the Multifactor

Tests for Trademark Infringement, 94 Calif. L. Rev. 1581, 1608,

1640-42 (2006).

4. See Conopco, Inc. v. May Dep’t Stores Co., 46 F.3d

1556, 1563 (Fed. Cir. 1994) (rejecting idiosyncratic confusion);

Jacob Jacoby & Wayne D. Hoyer, The Comprehension and

Miscomprehension of Print Communications 110-13 (1987)

(finding that an average of 19.3% of messages in magazine ads

were affirmatively misunderstood by consumers; no message

was correctly conveyed to all readers); Jacob Jacoby et al.,

Miscomprehension of Televised Communications 64-73 (1980)

(finding that consumers misunderstood an average of 28.3%

of messages in television commercial ads; 81.3% of consumers

misunderstood at least some portion of those ads, and no ad was

completely understood).

5. Jack Daniel’s Props., Inc. v. VIP Prods. LLC, 599 U.S.

140, 163-65 (2023) (Sotomayor, J., concurring) (cautioning courts

regarding the manipulability of survey evidence); Daryl Lim,

Trademark Confusion Revealed: An Empirical Analysis, 71 Am.

5

test therefore prioritizes the circumstantial evidence of

mark strength, mark similarity, and relatedness of the

parties’ goods and services.

Courts have properly recognized that the multifactor

test is ill-suited for a number of situations, including

comparative advertising, other kinds of “nominative” fair

use,6 resale of legitimate goods,7 noncommercial speech,

and the like. Such cases regularly involve situations where

two key factors—strength of the plaintiff’s mark and

the similarity of the term used by the defendant to that

mark—do not make confusion more likely. For example,

“Pepsi is better than Coke” is not likely to confuse even

though Coke is unquestionably a strong mark and the

goods compete directly. Using the multifactor test in

U. L. Rev. 1285, 1313-14 (2022) (“Constructing a robust survey is

dauntingly hard. . . . As a matter of justice between the parties, the

staggering costs of surveys put defendants at a disadvantage.”); cf.

Indianapolis Colts, Inc. v. Metro. Balt. Football Club Ltd. P’ship,

34 F.3d 410, 416 (7th Cir. 1994) (criticizing “the survey researcher’s

black arts” and noting that surveys do not involve real consumption

decisions); John P. Liefeld, How Surveys Overestimate the

Likelihood of Consumer Confusion, 93 Trademark Rep. 939

(2003).

6. See, e.g., New Kids on the Block v. News Am. Publ’g, Inc.,

971 F.2d 302, 308 (9th Cir. 1992) (defining nominative fair use);

Century 21 Real Est. Corp. v. Lendingtree, Inc., 425 F.3d 211,

222 (3d Cir. 2005) (adopting different variant of defense); Int’l

Info. Sys. Sec. Certification Consortium, Inc. v. Sec. Univ., LLC,

823 F.3d 153, 168 (2d Cir. 2016) (adding Third Circuit factors to

multifactor test).

7. See, e.g., Iberia Foods Corp. v. Romeo, 150 F.3d 298 (3d

Cir. 1998) (applying a “material difference” test, in accordance

with general practice).

6

comparative advertising cases would distort the confusion

inquiry rather than aid it. Choosing a test is itself a legal

question about the application of the law to the facts of

a specific case. This Court did precisely that in holding

that the multifactor test was appropriate in Jack Daniel’s.

Context matters even in cases where the multifactor

test is more appropriate. That is because the elements

of the multifactor test involve general claims about

reality, not just case-specific claims. In the language

of the Federal Rules of Evidence, many inputs into the

confusion factors are not adjudicative facts.8 They combine

predictions about the world with normative claims. These

general statements about when confusion is more likely

are not subject to factual development at the district

court level. Properly understood, the multifactor test is

thus a mixed question of law and fact,9 and review must

8. The usual contrasting term is the misnomer “legislative

facts.” See Hannah Walser, Interpretive Facts: Textualism,

Empiricism, and the Law-Fact Divide, 101 Ind. L.J. 1071,

1089 (2026) (“[S]uch facts often involve estimates of probability,

inferences about causation, and other exercises of judgment or

evaluation—judgment constrained by empirical evidence, but

judgment nonetheless.”). Many factors and subfactors in the

multifactor confusion test are those kinds of predictions. Cf.

Dunagin v. City of Oxford, 718 F.2d 738, 748 n.8 (5th Cir. 1983)

(plurality opinion) (holding that “the issue of whether there is

a correlation between advertising [of alcohol] and consumption

[of alcohol] is a legislative and not an adjudicative fact question”

because “[i]t is not a question specifically related to this one case

or controversy,” and because it depends upon “social factors and

happenings which may submit to some partial empirical solution

but is likely to remain subject to opinion and reasoning.”).

9. As are many tort-style inquiries. Cf. Dellums v. U.S.

Nuclear Regul. Comm’n, 863 F.2d 968, 972 n.5 (D.C. Cir. 1988)

7

sometimes go beyond clear error in order for there to be

meaningful appellate review of the non-adjudicative facts

that shape its application.10

As Judge Newman explained in a prescient dissent,

searching appellate review is required in many trademark

cases because trademark cases ask far more of factfinders

than to identify specific historical facts. “[T]he broad

range of reasonable fact-finding as to historical facts and

familiar legal issues like negligence is not appropriate for

issues infused with technical legal meaning, issues with

which juries are almost completely unfamiliar.” Samara

Bros. v. Wal-Mart Stores, Inc., 165 F.3d 120, 135 (2d Cir.

1998) (Newman, J., dissenting in relevant part), rev’d, 529

U.S. 205 (2000).

Deciding trademark cases is not like deciding who

caused a car accident. Such who-hit-whom cases involve

much more of a division between facts and law. The issues

in trademark law are intertwined. As Judge Newman

(“The question of causation or redressability . . . usually is an issue

of predictive fact. As such, it is not at all clear to us that it should

or can be treated as an historical fact. . . . Predictive facts seem

technically more like mixed questions of law and fact. Perhaps

more accurately, . . . they involve as much policy as law.”).

10. See Lockhart v. McCree, 476 U.S. 162, 168 n.3 (1986)

(‘We are far from persuaded . . . that the “clearly erroneous”

standard of Rule 52(a) applies to the kind of “legislative” facts at

issue here.’).”); cf. Kenji Yoshino, Appellate Deference in the Age

of Facts, 58 Wm. & Mary L. Rev. 251, 255 (2016) (arguing for an

intermediate standard of review for “legislative facts,” depending

on “the degree of adversarial testing (broadly construed) to which

the legislative facts had been subjected”).

8

explained, “[i]t is not a matter of giving less deference to

a jury’s fact-finding, reached within an allowable scope.

Rather, it is a matter of recognizing that the scope of

allowable fact-finding is narrower. . . . Otherwise, we

are ceding to juries broad authority to determine the

substantive scope of the law on topics such as antitrust,

copyright, trademark, and, in this case, trade dress.” Id.

at 135-36. Wal-Mart is an excellent example of a case in

which legal questions predominated: a jury found that a

product design was protectable because it was inherently

distinctive, but this Court held that product design cannot

be inherently distinctive, and that, to better promote

competition, any doubts about whether something is

product or packaging should be resolved by treating it as

product design. 529 U.S. at 216.

This Court’s more recent ruling in the Jack Daniel’s

case shows the vital role of appellate review to protect

parody and other forms of protected expression. Jack

Daniel’s Props., Inc. v. VIP Prods. LLC, 599 U.S. 140

(2023). On remand, the legal signal sent by the Court’s

focus on confusion about source and skepticism about

survey evidence in parody cases changed the district

court’s reading of the facts even absent any further

evidence. Given that many trademark owners make

overreaching claims, the Court should preserve the

flexibility appellate courts have historically used in

reviewing confusion claims to protect both competition

and speech.11

11. Cf. TikTok v. Garland, 604 U.S. 56, 83 (2025) (Gorsuch, J.,

concurring) (observing that litigation over an analytic framework

may “take on a life of its own and do more to obscure than to clarify

the ultimate” questions).

9

Whatever the resolution of this case, therefore, Amici

urge the Court not to collapse the various considerations

that go into confusion inquiries—including the “strength

of the mark” factor in the multifactor test—into pure

adjudicative fact. Doing so would hamper courts’ ability

to develop the law of trademark as they have done for the

past nearly eighty years under the Lanham Act. It would

also raise the cost of trademark litigation by making it

harder to resolve cases at summary judgment, which

systematically advantages better-resourced parties in

ways that harm free competition and free speech. See

Glynn S. Lunney, Jr., Trademark’s Judicial De-Evolution:

Why Courts Get Trademark Cases Wrong Repeatedly,

106 Calif. L. Rev. 1195, 1202, 1236, 1261-62, 1272 (2018);

William McGeveran, The Imaginary Trademark Parody

Crisis (and the Real One), 90 Wash. L. Rev. 713, 745–53

(2015); cf. Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S.

205, 214 (2000) (“Competition is deterred … not merely

by successful suit but by the plausible threat of successful

suit[.]”).

II. Strength Specifically is a Mixed Question

Like the other factors in the multifactor test, neither

the strength of the plaintiff’s mark nor the division of the

strength factor into marketplace and conceptual strength

are mentioned in the Lanham Act’s prohibition on likely

confusion. Courts developed these heuristics to help in

predicting whether claimed confusion is actionable. They

are not adjudicative facts but rules of general application,

in the sense that they are “established truths, facts or

pronouncements that do not change from case to case

but apply universally, while adjudicative facts are those

10

developed in a particular case.” United States v. Gould,

536 F.2d 216, 220 (8th Cir. 1976).

Likewise, the Abercrombie spectrum for evaluating

the conceptual strength of word marks is not specified

by the statute. It was created by Judge Friendly as a

synthesis of prior cases and has been refined over time.

The thought is that consumers will not immediately

recognize descriptive terms as trademarks, and only

experience over time can lead them to see the terms as

indicators of source. By contrast, suggestive, arbitrary,

and fanciful terms are less necessary for competitors

and more likely to be recognized immediately as source

indicators. The Abercrombie spectrum concededly doesn’t

work well for non-word marks like images or product

packaging, much less product design.12

Courts generally take for granted the propositions (1)

that a mark that is further towards the “fanciful” side of

the Abercrombie spectrum is conceptually stronger, and

one that is highly descriptive is conceptually weaker, and

(2) that this conceptual strength bears monotonically on

the likelihood of confusion.13

12. See Wal-Mart, 529 U.S. at 216; Amazing Spaces, Inc.

v. Metro Mini Storage, 608 F.3d 225, 242-43 (5th Cir. 2010)

(Abercrombie spectrum a poor fit to evaluate distinctiveness of

image).

13. See, e.g., Nutri/Sys., Inc. v. Con-Stan Indus., 809 F.2d 601,

605 (9th Cir. 1987) (“Arbitrary or fanciful marks (i.e., Kodak) are

called ‘strong’ marks, whereas descriptive or suggestive marks

are ‘weak.’” (citation modified)); Brookfield Commc’ns, Inc. v. W.

Coast Ent. Corp., 174 F.3d 1036, 1058 (9th Cir. 1999) (suggestive

marks are presumptively weak).

11

Courts did not develop these categories based on

empirical research. Consumers on their own have no

particular reason to know or care whether a mark is

arbitrary, fanciful, suggestive or descriptive, or what

those terms mean for trademarks. This is why genericity

surveys have to train consumers on the difference between

a brand name and a common name before asking them

questions about whether a word is generic—and, notably,

neither “brand” nor “common” are the terms used in

trademark law.14

The law’s preference for inherently distinctive

marks15 —and for fanciful (made-up) or arbitrary marks

over suggestive and descriptive marks that say something

about the product or service at issue—reflects both a

prediction that consumers will immediately understand

the mark at issue to identify source16 and also a legal

14. See, e.g., Booking.com B.V. v. Matal, 278 F. Supp. 3d 891,

915 (E.D. Va. 2017), aff’d sub nom. United States Patent and

Trademark Office v. Booking.com B. V., 915 F.3d 171 (4th Cir.

2019), aff’d, 591 U.S. 549 (2020) (explaining the process of training

survey respondents so they can understand the questions).

15. See, e.g., Blau Plumbing, Inc. v. S.O.S. Fix-It, Inc., 781

F.2d 604, 609 (7th Cir. 1986) (inherently distinctive marks are

“favored”).

16. See Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763,

768 (1992); Savannah Coll. of Art & Design, Inc. v. Sportswear,

Inc., 983 F.3d 1273, 1282 (11th Cir. 2020); In re Chippendales USA,

Inc., 622 F.3d 1346, 1352 (Fed. Cir. 2010) (“Inherent distinctiveness

does not depend on a showing that consumers actually identify the

particular mark with the particular business. . . . [I]f the mark is

inherently distinctive, it is presumed that consumers will view it

as a source identifier.”); Graeme B. Dinwoodie, Trade Mark Law

As A Normative Project, 2023 Sing. J. Legal Stud. 305, 314-15

12

incentive to businesses to select marks that pose less risk

of hampering competition.17 Giving inherent strength

weight in the confusion inquiry provides part of that

incentive.

To the extent that there is empirical research, it

does not support Abercrombie’s categories. Instead,

that research strongly suggests that the Abercrombie

spectrum misunderstands how people actually process

information about products and services.18 Nonsense

(“Inherently distinctive marks are protected upon the basis that

they are likely to act as source-identifiers. . . . [This ‘predictive’

inquiry] is an educated guess about what reality will be when the

mark is used.” (footnotes omitted) (citing J. Thomas McCarthy,

McCarthy on Trademarks and Unfair Competition § 8.02[4] (5th

ed. 2022))).

17. See, e.g., Chippendales, 622 F.3d at 1352; Blau, 781 F.2d

at 609; Dinwoodie, Normative Project, supra, at 315.

18. See, e.g., Chiranjeev S. Kohli & Rajneesh Suri, Brand

Names That Work: A Study of the Effectiveness of Different Types

of Brand Names, 10 Mktg. Mgmt. J. 112, 117 (2000) (finding that

arbitrary names consistently performed the worst overall in key

dimensions of likability, recall, and recognition); Thomas R. Lee,

Eric D. DeRosia & Glenn L. Christensen, An Empirical and

Consumer Psychology Analysis of Trademark Distinctiveness, 41

Ariz. St. L. J. 1033, 1038 (2009) (demonstrating that “descriptive

marks are no less source-indicating than suggestive, arbitrary,

and fanciful marks” and “that it is the non-lexical cues of typical

trademark use (e.g., the location and size of the mark) that cause

consumers to find descriptive marks to be source-indicating”);

Jake Linford, Are Trademarks Ever Fanciful?, 105 Geo. L.J. 731,

735 (2017) (arguing that the category “fanciful” has no empirical

reality); Jake Linford, The False Dichotomy Between Suggestive

and Descriptive Trademarks, 76 Ohio St. L.J. 1367 (2015) (using

linguistic evidence to argue that consumers perceive suggestive

13

marks used for products on Amazon, like BSTOEM

or ZGGCD, are fanciful and thus at the top end of the

“inherently distinctive” category, but they are not

memorable or easily distinguishable from other nonsense

marks and thus not useful to consumers in helping them

remember products with which they have had good or bad

experiences—that is, they are not good at performing

actual trademark functions.19

But whether the Abercrombie spectrum’s ranking of

marks is correct is less important here than the recognition

that the placement of a mark on the conceptual spectrum

is not a pure question of adjudicative fact, nor is how

much weight to give it in the confusion analysis. Indeed,

juries are regularly simply told that strength matters

monotonically—stronger always means more likelihood of

confusion. This instruction makes the weight of strength

in the confusion inquiry more like a legal conclusion than

a factual question. 20

and descriptive terms as fundamentally similar); Dustin Marlan,

The Arbitrary Myth: Challenging Trademark Law’s Elevation

of Arbitrary Marks, 41 Berkeley Tech. L.J. 555, 561 (2026)

(“consumer research demonstrat[es] that arbitrary marks

underperform on measures of recall, recognition, and likeability,

trailing not only fanciful marks but also suggestive and even

descriptive ones”).

19. Note, Fanciful Failures: Keeping Nonsense Marks Off

the Trademark Register, 134 Harv. L. Rev. 1804, 1806 (2021).

20. See, e.g., Ninth Cir. Jury Instructions Comm., Manual

of Model Civil Jury Instructions for the District Courts of the

Ninth Circuit § 15.19, at 446 (2025 ed., current through June

2026) (last visited Sept. 9, 2026) (“The strength of a trademark

is determined by its distinctiveness and recognition in the

marketplace. The stronger a mark is, the greater the scope of

14

Moreover, apparent conceptual strength based on

dictionary definitions isn’t the end of the assessment

of the strength factor. McCarthy, a leading trademark

authority, treats the existence of a crowded market as a

separate input into the assessment of “relative strength.”

1 McCarthy, supra, § 11:85 (“A mark that is hemmed in

on all sides by similar marks on similar goods or services

cannot be very ‘distinctive.’ It is merely one of a crowd of

similar marks. In such a crowd, customers will not likely

be confused between any two of the crowd and may have

learned to carefully pick out one from the other.”). 21

McCarthy does not take a position on whether

crowding is (1) evidence about market strength because

consumers are likely to know that a term is common across

different sources, or (2) evidence that, whatever conceptual

distinctiveness might have been theoretically possible, the

term is now a dead metaphor (e.g., “king size”) no longer

requiring thought or imagination to connect to the goods

or services at issue. Both are reasonable explanations for

why crowding matters. That is because both conceptual

and marketplace strength are loose constructs, not natural

categories; like the other factors, they bear on each other.

protection the law provides.”); Fifth Cir. Civil Jury Instructions,

Trademark Infringement: Specific Preliminary Instructions

To The Jury ch. 14.11, at 17 (2024) https://www.lb5.uscourts.gov/

juryinstructions/Fifth/2024Civil_Ch14_Trademarks.pdf (last

visited Sept. 9, 2026) (same).

21. See also 1 McCarthy, supra, § 11:76 (“One way a mark

is ‘weak’ is that it is known only by a few dedicated customers

because it is the subject of few sales and little advertising or

promotion. Another way a mark may be ‘weak’ is that there are

many similar marks used by others, resulting in a ‘crowded’

trademark marketplace.”) (footnote omitted).

15

For example, early in the Internet era, many

businesses used terms or images relating to dogs known

for their ability to “fetch” or “retrieve.” Given the

connection delays then common, these terms, which might

have seemed suggestive or even arbitrary for software

in the abstract, were so useful that they were properly

classified as descriptive. Labrador Software, Inc. v. Lycos,

Inc., 32 F. Supp. 2d 31, 33-34 (D. Mass. 1999). “Rise” could

be subject to a similar analysis. While the population of the

crowd of similar marks in any given case is an adjudicative

fact, the meaning of that density in the overall context of

an infringement inquiry—how market crowding interacts

with the abstract conceptual strength of the mark—is not.

Additionally, in particular cases, the strength of a

mark may actually help the defendant, not the plaintiff.

Especially in situations involving parody or comparative

advertising involving well-known marks, consumers

know what a mark should look like, so strength should

decrease the likelihood of confusion in those situations. 22

Likewise, in so-called “reverse confusion” cases, courts

require more conceptual strength and less marketplace

strength in order to favor a plaintiff. 23 This is another

22. See, e.g., Louis Vuitton Malletier S.A. v. Haute Diggity

Dog, LLC, 507 F.3d 252, 261-62 (4th Cir. 2007); Tommy Hilfiger

Licensing, Inc. v. Nature Labs, LLC, 221 F. Supp. 2d 410, 416

(S.D.N.Y. 2002); Barton Beebe & C. Scott Hemphill, The Scope of

Strong Marks: Should Trademark Law Protect the Strong More

than the Weak?, 92 N.Y.U. L. Rev. 1339, 1364-66 (2017).

23. See, e.g., A & H Sportswear, Inc. v. Victoria’s Secret

Stores, Inc., 237 F.3d 198, 231 (3d Cir. 2000); Jeremy N. Sheff,

Reverse Confusion and the Justification of Trademark Protection,

30 Geo. Mason L. Rev. 123, 144-46 (2022).

16

example of how the legal significance of a factual finding

varies situationally.

III. Adjudicative Facts Bea r on Likelihood of

Confusion But Do Not Map Directly To the

Likelihood of Confusion Factors

Although the Court did not grant certiorari on other

aspects of the multifactor confusion test, we discuss them

here in order to place the strength inquiry in its broader

context. The complexity begins with the concept of likely

confusion. Likely confusion isn’t possible confusion. 24 Nor

is likely confusion shown by the existence of confusion

among a few people out of thousands. 25

Instead, courts look for likely confusion among

an appreciable number of reasonable consumers. This

requires making a number of legal as well as factual

judgments26: For example, the threshold for actionable

24. See, e.g., Sorensen v. WD-40 Co., 792 F.3d 712, 726 (7th

Cir. 2015) (“Possible confusion is not enough; rather, confusion

must be ‘probable.’” (citing 4 McCarthy § 23.3)).

25. See, e.g., Water Pik, Inc. v. Med-Sys., Inc., 726 F.3d 1136,

1150 (10th Cir. 2013) (“We have consistently recognized, however,

that isolated, anecdotal instances of actual confusion may be de

minimis and may be disregarded in the confusion analysis.”); Door

Sys., Inc. v. Pro-Line Door Sys., Inc., 83 F.3d 169, 173 (7th Cir.

1996) (Posner, J.) (“[P]laintiff’s evidence that two consumers (out

of how many thousands?) may have been misled cannot by itself

be thought to create a contestable issue of likelihood of confusion

even if the evidence … is admissible and credible.”).

26. See Graeme B. Dinwoodie & Dev S. Gangjee, The Image

of the Consumer in EU Trade Mark Law, in The Images of

the Consumer in EU Law: Legislation, Free Movement and

17

confusion can be lowered if confusing the products at issue

would have potential health consequences. 27 By contrast,

the threshold should be raised when the defendant’s use

is plausibly parodic, engaged in social commentary, or

otherwise affirmatively beneficial to substantial numbers

of consumers. 28 Such decisions are consequential for

categories of subsequent litigants, not tailored to an

individual case.

In a multifactor test, the factors often bear on one

another. Meaningful appellate review therefore may

Competition Law 339, 345 & n.26, 353 (Dorota Leczykiewicz

& Stephen Weatherill eds., 2016) (explaining connection to tort

law’s reasonable person; noting that “almost all conceptions

of the consumer—as circumscribed by the applicable legal

tests—reflect a blend” of normative and empirical judgments);

Dinwoodie, Normative Project, supra, at 308-09 (“[B]ecause

trade mark law does not seek to prevent all forms of confusion,

the consumer protection rationale presents a series of essentially

normative questions regarding the nature and level of confusion

that should be actionable, along with questions about the forms of

consumer understanding that are properly protected against such

confusion. . . . You need to know what you are measuring—and

why—before you devise rules for doing so.”).

27. See, e.g., Kos Pharms., Inc. v. Andrx Corp., 369 F.3d 700,

715-17 (3d Cir. 2004); Syntex Lab’ys, Inc. v. Norwich Pharmacal

Co., 437 F.2d 566, 569 (2d Cir. 1971); 3 McCarthy, supra, § 23:32.

28. See, e.g., Prestonettes, Inc., v. Coty, 264 U.S. 359, 367-68

(1924) (consumers’ interests in repackaged legitimate goods);

Yankee Pub. Inc. v. News Am. Pub. Inc., 809 F. Supp. 267, 273

(S.D.N.Y. 1992) (consumers’ interests in parody and commentary);

Jeremy N. Sheff, Veblen Brands, 96 Minn. L. Rev. 769, 783-86

(2012) (discussing “non-confused consumers”); Michael Grynberg,

Trademark Litigation As Consumer Conflict, 83 N.Y.U. L. Rev.

60, 77-85, 90-91 (2008) (same).

18

require evaluating how district courts (or the Trademark

Trial and Appeal Board) applied and weighed the

different factors. Such review would have been difficult

had the test’s application been treated as a purely factual

matter. In one recent example, the Trademark Trial and

Appeal Board used a narrow definition of relevant goods

and services when assessing whether the market was

“crowded” for purposes of strength, and then a broad

definition when assessing whether the parties’ offerings

were related enough that consumers would think they

were affiliated. This, the Federal Circuit held, was a legal

error. Apex Bank v. CC Serve Corp., 156 F.4th 1230, 123536 (Fed. Cir. 2025).

Factor interactions are common. For example, mark

similarity is an obvious example of a fact-specific inquiry,

similar to trademark tacking in Hana Financial, Inc.

v. Hana Bank, 574 U.S. 418, 420 (2015). The similarity

must be sufficient, in context, to make confusion likely.

A factfinder may need to determine whether and how

descriptive or generic elements, or stylistic differences,

matter. Differences and similarities in “sight, sound, and

meaning” are all relevant. 29

But there is also an interaction between similarity

and a crowded market. Crowding lessens the weight of

similarity. 30 As noted above, no matter how “strong” a

29. See, e.g., George & Co. LLC v. Imagination Ent. Ltd., 575

F.3d 383, 396 (4th Cir. 2009).

30. See, e.g., One Indus., LLC v. Jim O’Neal Distrib., Inc., 578

F.3d 1154, 1164 (9th Cir. 2009) (“When similar marks permeate

the marketplace, the strength of the mark decreases.”); Fla.

Int’l Univ. Bd. of Trs. v. Fla. Nat’l Univ., Inc., 830 F.3d 1242,

19

mark is conceptually in theory, it may still provide only

narrow exclusion rights if many market participants use

the same or similar metaphors or associations in their

own marks. Barton Beebe, The Semiotic Analysis of

Trademark Law, 51 UCLA L. Rev. 621, 672 (2004) (“It

is rather a mark’s distinctiveness from other marks, its

salience, that makes consumers aware of it and that will

affect whether consumers are likely to confuse it with a

junior, similar mark.”). This kind of interaction between

factors is another reason that appellate review often must

go beyond clear error to fully evaluate a party’s legal

arguments, even when examining mark similarity relies

on adjudicative factfinding.

The one-to-one comparison required in Hana Bank

for determining whether a single trademark maintained

a continuous commercial impression is related to the

similarity question in an infringement inquiry. However,

the former scenario is far less complex, because assessing

whether a mark retained a continuous commercial

impression over time doesn’t additionally require a oneto-many comparison. Whether a Labrador retriever logo

redesigned to be more or less realistic is commercially

continuous with the prior version may not require

consideration of the other logos in the marketplace. All

that is required is a determination of whether the newer

mark is sufficiently close to the former mark that we can

call it the same mark. But an infringement dispute over

1258 (11th Cir. 2016) (given a crowded field, the disputed names

“are not particularly distinct to consumers”); 2 McCarthy, supra,

§ 11:88 (4th ed. 2015) (sufficient evidence of third-party use of

similar marks can “show that customers . . . ‘have been educated

to distinguish between different . . . marks on the basis of minute

distinctions’”).

20

two sellers’ dog logos uses similarity only as one input into

the broader question of confusion, so effects of the level of

similarity and kind of similarity are highly dependent on

the other factors. In other words, Hana Bank concerned

only the claimant’s priority: when did it first use “the

mark”? That question is antecedent to likely confusion,

which concerns the scope of the rights that come with

ownership.

Likewise, the construct of the “reasonable” consumer

from whose perspective likely confusion is assessed is

often both empirical and legal. Trademark law requires

consumers to be “reasonable”—they have to be more

careful to distinguish among providers of real estate

services than of candy bars. This is often a normative

inquiry, not an empirical one. 31 Reasonable consumers

must be more careful when purchases are more important,

and the law does not inquire whether most consumers

are reasonable consumers. Reasonability under the

circumstances is a classic mixed question of law and fact,

requiring “the application of a legal standard” to each

case. 32

Other mixed questions include the “actual confusion”

factor. Although this seems on its fact to be a purely

empirical assessment, its boundaries are regularly

adjusted as a legal matter. Courts regularly say that

evidence of actual confusion is usually hard to find,

31. See Michael Grynberg, The Consumer’s Duty of Care

in Trademark Law, in Research Handbook on Trademark Law

Reform 326, 330-31 (Graeme B. Dinwoodie & Mark D. Janis eds.,

2021).

32. Hana Bank, 574 U.S. at 423.

21

because consumers rarely bother complaining even if they

ultimately figure out that they were confused. For that

reason, courts say that no inference should be made from

its absence. 33 This assumption that deceived consumers

stay silent is not based on actual evidence of consumer

complaint behavior. Courts then modify this rule with

the proposition that, if the parties have coexisted for a

long time, the absence (or minimal amount) of reported

confusion weighs against the plaintiff. 34 If the plaintiff

is wealthy enough, the court can infer that failure to

conduct a confusion survey weighs against the likelihood

of confusion. 35

Courts also differ on whether misdirected calls or

messages, when a consumer is looking for one entity

but contacts the other, constitute evidence of trademark

confusion or just errors in typing.36 Courts differ about the

33. See, e.g., Brookfield, 174 F.3d at 1050; Checkpoint Sys.,

Inc. v. Check Point Software Techs., Inc., 269 F.3d 270, 291 (3d Cir.

2001) (“[I]t is difficult to find evidence of actual confusion because

many instances are unreported.”); AmBrit, Inc. v. Kraft, Inc., 812

F.2d 1531, 1544 (11th Cir. 1986) (“It is likely that many consumers

who were confused never realized they had been confused and that

many of those who did realize they had been confused chose not

to spend the time to register a complaint.”).

34. See, e.g., Cohn v. Petsmart, Inc., 281 F.3d 837, 842-43

(9th Cir. 2002).

35. See, e.g., Mushroom Makers, Inc. v. R. G. Barry Corp.,

441 F. Supp. 1220, 1231 (S.D.N.Y. 1977), aff’d, 580 F.2d 44 (2d Cir.

1978); Merriam-Webster, Inc. v. Random House, Inc., 35 F.3d 65,

72 (2d Cir. 1994); see also 5 McCarthy, supra, § 32:195 (discussing

negative inferences from absence of survey evidence).

36. Compare, e.g., Appliance Liquidation Outlet, L.L.C. v.

22

import of questions from consumers about whether two

entities are connected—some find evidence of confusion,

while others find that the querying consumer understood

that she didn’t have enough information to know for sure

and thus was not confused about source. 37 These are not

the kinds of factual propositions that are tested by an

individual factfinder; they are legal rules about how to

assess facts.

The “intent” factor, too, requires legal choices to

go along with factual inquiries. Some courts say that a

defendant’s knowledge of the plaintiff’s mark is sufficient

to find bad faith where there is some similarity; others

say that knowledge is not enough. 38 Some courts say

Axis Supply Corp., 105 F.4th 362, 382 (5th Cir. 2024) (finding a

“repeated and persistent flood of mistaken inquires” by relevant

consumers to be evidence of actual confusion), with, e.g., ThermaScan, Inc. v. Thermoscan, Inc., 295 F.3d 623, 636 (6th Cir. 2002)

(consumers were merely “inattentive or careless when attempting

to find the e-mail address for [defendant], rather than confused

about the source of the ear thermometers.”).

37. Compare, e.g., Int’l Kennel Club of Chicago, Inc. v. Mighty

Star, Inc., 846 F.2d 1079, 1090 (7th Cir. 1988) (inquiries regarding

the relationship between plaintiff and defendant’s product

“constitute[d] probative evidence of a likelihood of confusion”),

with, e.g., Fisher Stoves, Inc. v. All Nighter Stove Works, Inc., 626

F.2d 193, 195 (1st Cir. 1980) (instances of consumers “ask[ing] if

the two companies were affiliated” were evidence of consumers

“ha[ving] the difference in source clearly in mind”).

38. Compare, e.g., Wynn Oil Co. v. Am. Way Serv. Corp.,

943 F.2d 595, 603 (6th Cir. 1991) (“Understandably, courts have

held that use of a mark with knowledge of another’s prior use

of the mark supports an inference of intentional infringement.”

(collecting cases)), with, e.g., Sorensen, 792 F.3d at 731 (“Mere

23

that intentional copying is bad faith, while others hold

that—especially with respect to features that might have

non-trademark functions—intentional copying merely

shows a desire to compete, not an intent to confuse. 39 And

reverse confusion cases use a still different definition of

bad intent.40 Each of these definitions requires different

factual proof, and so it is important to know which tilt the

confusion inquiry in a plaintiff’s favor and which do not.

Moreover, the proposition that intended acts are more

likely to succeed than unintended acts, which is why

bad intent is said to make confusion more likely, is once

again a general proposition not tested by individualized

factfinding.41

knowledge of someone else’s mark is insufficient to show intent to

pass off.” See also 3 McCarthy, supra, § 23:115 (collecting cases).

39. See, e.g., Paddington Corp. v. Attiki Importers &

Distributors, Inc., 996 F.2d 577, 587 (2d Cir. 1993) (“Where such

prior knowledge is accompanied by similarities so strong that

it seems plain that deliberate copying has occurred, we have

upheld findings of bad faith.”). But see, e.g., Craft Smith, LLC v.

EC Design, LLC, 969 F.3d 1092, 1106 (10th Cir. 2020) ([W]hen a

competitor copies a product’s design, its purpose is not necessarily

to confuse consumers, but to copy the aspects of that product that

make it more functional.”).

40. See Sheff, Reverse Confusion, at 148-151.

41. It is probably more persuasive to see bad faith as a

separate normative consideration about the defendant’s conduct

rather than an actual empirical claim about confusion. See Bone,

supra, at 1347, 1350-52; Jeremy N. Sheff, Marks, Morals, and

Markets, 65 Stan. L. Rev. 761, 810-11 (2013); cf. Beebe, Empirical

Study of the Multifactor Tests, supra, at 1628-31. The very lack

of clarity in the case law about this question suggests the need

for appellate oversight.

24

Because these questions recur in multiple cases, it

is important for courts to be able to provide predictable

answers—precisely the situation in which clear error

review alone can be insufficient to produce coherent law.

IV. The Multifactor Test Creates a Variably Mixed

Question of Law and Fact

The Sixth Circuit explained why the overall multifactor

test is a mixed question in Sterling Jewelers, Inc. v.

Artistry Ltd., 896 F.3d 752, 755 (6th Cir. 2018), noting

that it has “factual components (e.g., what evidence of

confusion has been shown?) and legal components (e.g.,

what counts as cognizable confusion?).” As it further

noted, “[e]ach factor presents a question of law and fact,

not math. As potential proxies for the kind of confusion

that the Lanham Act prohibits, each factor’s relevance

and relative importance will depend on the facts of each

case.” Id. at 756.

This presents a challenge for review. As this Court

explained:

The appropriate standard of review for a mixed

question depends “on whether answering it

entails primarily legal or factual work.” Some

mixed questions require a court to “expound

on the law” by “amplifying or elaborating on a

broad legal standard.” When applying the law

involves developing legal principles for use in

future cases, appellate courts typically review

the decision de novo.

25

Buf kin v. Collins, 604 U.S. 369, 381 (2025) (citation

modified).

In trademark cases, courts are sometimes immersed

in facts. But they are often enough also, or even

predominantly, relying on normative claims about the

scope of trademark law or factual propositions that weren’t

tested before the factfinder. Additionally, predictive

inquiries into the factors that cause likely confusion

are often refined over time in ways that give guidance

to market participants, who are often sophisticated

commercial entities. Cf. id. at 384-85 (use of objective

reasonable person standard favors de novo review); see

also id. at 399-400 (Jackson, J., dissenting) (“As a general

matter, this Court has long treated the application of a

legal standard to a given set of facts as an exercise that

poses a legal question.”).42 For example, the circuits have,

with decades of experience, reached consensus that socalled “keyword advertising”—using trademarks as paidfor search terms to offer competing goods or services—is

not infringing.43 This was only possible because appellate

42. Cf. Timothy B. Dyk, The Role Of Non-Adjudicative Facts

In Judicial Decisionmaking, 76 Stan. L. Rev. Online 10, 26 (2023)

(arguing that the use of “[b]road statutory language” in the Patent

Act delegated to courts the role of developing and using nonadjudicative facts in their tests); Justin Hughes, The Respective

Roles of Judges and Juries in Copyright Fair Use, 58 Hous. L. Rev.

327, 348 (2020) (though copyright fair use determinations are case

by case, “there is also no question that fair-use jurisprudence is a

body of ‘auxiliary legal principles’ developed through precedent,

suggesting de novo review of the overall fair-use conclusion”).

43. See, e.g., Deltona Transformer Co. v. NOCO Co., --- F.4th

----, No. 24-13590, 2026 WL 2236806, at *8 (11th Cir. Aug. 4, 2026);

Lerner & Rowe PC v. Brown Engstrand & Shely LLC, 119 F.4th

26

courts were able to consider how to weigh the various

factors in a repeating situation, using general principles

of law.44

A judge-made, multifactor test with substantial policy

components may often be predominantly a question of law,

despite key factual components. Google LLC v. Oracle

Am., Inc., 593 U.S. 1 (2021), explained that the “judgemade origins” of copyright fair use, “as well as modern

courts’ use of the doctrine, makes clear that the concept is

flexible, that courts must apply it in light of the sometimes

conflicting aims of copyright law, and that its application

may well vary depending upon context,” 593 U.S. at 20.

The inquiry is necessarily case by case because uses differ,

and thus a mixed question of law and fact. Id. at 23-24.

Where, as in Google, appellate review is concerned

with “legal work” that “provide[s] general guidance for

future cases,” id. at 24, de novo review is appropriate.

Given the multiple legal judgments embedded in balancing

the factors and subfactors when they point in substantially

different directions, that is often the case in trademark

law as well. Justice Kennedy’s statement that “[t]he

common-law method instated by the fair use provision

711, 719, 721-26 (9th Cir. 2024), cert. denied, 145 S. Ct. 2732 (Mem)

(2025); 1-800 Contacts, Inc. v. JAND, Inc., 119 F.4th 234, 249-55

(2d Cir. 2024); Jim S. Adler, P.C. v. McNeil Consultants, L.L.C.,

10 F.4th 422, 428-29 (5th Cir. 2021); 1-800 Contacts, Inc. v. Lens.

com, Inc., 722 F.3d 1229, 1242-43 (10th Cir. 2013).

44. See Network Automation, Inc. v. Advanced Systems

Concepts, Inc., 638 F.3d 1137 (9th Cir. 2011) (addressing “internet

troika” of factors in light of non-case-specific facts about consumer

familiarity with Internet).

27

of the copyright statute presumes that rules will emerge

from the course of decisions”45 also describes the body of

trademark litigation.

As part of this rule-development process, courts

discussing the application of the confusion factors cite

previous court cases as authorities, not external factual

sources as evidence. The very existence of multiple

treatises working through variations of the multifactor

test, such as McCarthy, supra, and Callman,46 and one

treatise just on likely confusion, Richard L. Kirkpatrick

& Vijay K. Toke, Likelihood of Confusion in Trademark

Law (2d ed. 2026), further indicates the systemic need to

“clarify legal principles” and “provide guidance to other

courts resolving other disputes,” U.S. Bank N.A. v. Village

at Lakeridge, LLC, 583 U.S. 387, 398 (2018).

Another reason that appellate review may go beyond

clear error is that trademark law has constitutional

dimensions: it can be used to suppress free speech. If

the likely confusion inquiry provides sufficient First

Amendment protection when the defendant’s accused use

is a trademark use, then it follows that review of a trial

court’s decision on likely confusion has constitutional

significance. Searching appellate review, Bose Corp.

v. Consumers Union of U.S., Inc., 466 U.S. 485, 510-11

(1984),47 is required to make sure the likely confusion

45. Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 596

(1994) (Kennedy, J., concurring) (citation omitted).

46. Malla Pollack, Callmann on Unfair Competition,

Trademarks and Monopolies (4th ed. 2026).

47. Yoshino, supra, at 278 (“ While the constitutional

facts doctrine has largely been applied to adjudicative facts,

28

test has done its job. A district court’s error can allow

a trademark owner to use the courts to suppress a

competitor’s or critic’s speech, justifying greater appellate

oversight. Cf. Bufkin, 604 U.S. at 384 (noting constitutional

significance of probable cause inquiries). Notably, upon

remand in Jack Daniel’s, the district court took this

Court’s well-justified skepticism about the role of survey

evidence in parody cases to heart and reversed its initial

finding.48

CONCLUSION

Confusion factors are court-developed. They involve

normative judgments and include general, predictive

propositions as sub-rules. These are not subject to factfinding at the trial stage. Instead, they are the framework

by which the finder of fact interprets the historical facts

specific to the parties’ dispute. Historical facts are

necessary but not sufficient inputs to the resolution of each

factor (including mark strength), and thus to the overall

likelihood of confusion inquiry.

its logic holds—perhaps even more strongly—with regard to

legislative facts. As the Seventh Circuit has observed, ‘an issue

of “constitutional fact” is reviewed without deference in order to

prevent the idiosyncrasies of a single judge or jury from having

far-reaching legal effects.’ By their nature, case-spanning facts

(that is, legislative facts) will be more likely to have such farreaching effects than case-specific ones (that is, adjudicative

facts).”) (citation modified).

48. VIP Prods. LLC v. Jack Daniel’s Props. Inc., No. CV-1402057, 2025 WL 275909, at *24 (D. Ariz. Jan. 23, 2025), vacated

and remanded sub nom. VIP Prods., LLC v. United States, 185

F.4th 919 (9th Cir. 2026).

29

Meaningful appellate review therefore requires

something more than clear error review as to the

application of these general rules, even when review

of individually relevant factual findings (such as how

many similarly named products exist in the market) is

appropriately only for clear error. Otherwise, the Court

will freeze in place a random snapshot of the common-law

evolution of the confusion inquiry over time, preventing

courts from reaching consensus on new recurring issues.

This Court has long recognized that fact and law

sometimes cannot be cleanly separated. Whether one

version of a mark offers the same commercial impression

as a later version is heavily factual because it is a single

comparison from an ordinary consumer’s perspective.

Hana Bank, 574 U.S. at 420; see also id. at 424-25.

Whether a challenged copyright infringement is a fair use

is often a legal question because the multifactor fair use

test has a number of legal inputs as well as some factual

questions (e.g., how much of the work was taken, what was

the market effect?). Google, 593 U.S. at 24. Likelihood

of confusion is closer to the latter than the former, not

just because it is multifactor, but because many of its

subfactors, like the fair use subfactors, require legal

judgments.

The Court was asked only to rule on one subfactor

of a multifactor test. Whatever its conclusions about the

status of the Abercrombie spectrum, it should recognize

that the overall multifactor inquiry is a mixed question of

law and fact in which adjudicative facts only sometimes

predominate and in which appellate review regularly

needs to go beyond clear error. Additionally, it should

not decide here that the non-statutory multifactor test

30

is always to be used in confusion cases, and it should

not preclude evidence-based challenges to the use or

application of the Abercrombie spectrum.

Respectfully submitted,

Rebecca Tushnet

Counsel of Record

Frank Stanton Professor

of the First Amendment

Harvard Law School

1563 Massachusetts Avenue,

Hauser 520

Cambridge, MA 02138

(703) 593-6759

rtushnet@law.harvard.edu

Attorney for Amici Curiae

APPENDIX

i

TABLE OF APPENDICES

Page

APPENDIX — LIST OF AMICI CURIAE . . . . . . . . . 1a

1a

Appendix

APPENDIX — LIST

OF AMICI CURIAE *

Barton Beebe, John M. Desmarais Professor of Intellectual

Property Law, NYU Law

Graeme B. Dinwoodie, University Disting uished

Professor, Global Professor of Intellectual Property Law,

Chicago-Kent College of Law

Shahrokh Falati, Associate Professor of Law, New York

Law School

Michael Grynberg, Raymond P. Niro Professor of

Intellectual Property Law, DePaul University College

of Law

Laura A. Heymann, James G. Cutler Professor of Law,

William & Mary Law School

Mark A Lemley, William H. Neukom Professor of Law,

Stanford Law School

Yvette Joy Liebesman, Professor of Law, Saint Louis

University School of Law

Jessica Litman, John F. Nickoll Professor of Law,

Michigan Law

Mark P. McKenna, Professor of Law, UCLA School of Law

* Institutional affiliations are provided for informational

purposes only.

2a

Appendix

Joseph Scott Miller, Ernest P. Rogers Chair of Intellectual

Property and Unfair Competition Law, University of

Georgia School of Law

Prof. Tyler T. Ochoa, Datta Center for High Tech Law,

Santa Clara University School of Law

Lisa P. Ramsey, Professor of Law, University of San Diego

School of Law

Zahr K. Said, Santa Clara University School of Law

Jeremy Sheff, Professor of Law, St. John’s University

School of Law

Paul J. Szynol, Lecturer, Michigan Law

Rebecca Tushnet, Frank Stanton Professor of the First

Amendment, Harvard Law School