On Confusion: A Restatement of Trademark's Likelihood-of-Confusion Test

Abstract

Trademark law’s most litigated question is decided by a test no one can explain. Every circuit recites its own list of eight or ten or thirteen confusion factors, no court has ever said how those factors combine into an outcome, and scholars have spent two decades cataloging the test’s failures.

Two recent empirical studies have changed the picture. Analyzing more than 4,000 decisions before the Trademark Trial and Appeal Board and more than 10,000 federal court opinions spanning 1970 to 2025, those studies demonstrate that just two factors, similarity of the marks and competitive proximity of the goods or services, predict the bottom-line confusion finding with accuracy exceeding 93%. The remaining factors function as structured exceptions rather than coequal inputs to an open balancing exercise: narrow, asymmetric overrides that flip the two-factor prediction in fewer than 7% of comparisons.

This Article translates those findings into doctrine. Adopting the format of a Restatement, it sets out black-letter rules, explanatory comments, and case-derived illustrations that reduce the confusion test to a gated framework: a two-factor threshold that generates a rebuttable presumption, followed by a structured defeater analysis drawn from the secondary factors, with stated extensions for reverse confusion and for procedural application at summary judgment, preliminary relief, trial, and appeal.

Courts can adopt the framework without overruling a single precedent; the black letter restates what courts have been doing all along, and the procedural applications that go beyond current practice are labeled as such. The framework governs only how factor findings combine; the law governing how each factor is assessed remains untouched. The result is a test that is clearer, more predictable, and faithful to what courts actually do.

Keywords: trademark, likelihood of confusion, trademark infringement, multifactor test, Lanham Act, trademark law, Polaroid factors, DuPont factors, Sleekcraft factors, likelihood of confusion factors, restatement, consumer confusion, empirical legal studies, trademark litigation, intellectual property, unfair competition, mark similarity, similarity of marks, relatedness of goods, trademark strength, famous marks, trademark fame, actual confusion, consumer sophistication, defendant intent, trade channels, reverse confusion, Trademark Trial and Appeal Board, TTAB, judicial decision-making, judicial behavior, content analysis of judicial opinions, large language models, legal doctrine, doctrinal reform

JEL Classification: O34, K41, K11, L15, D18, M31