Trademarks

When Search Results Aren't Confusing: Multi Time Machine v. Amazon and the Limits of Initial-Interest Confusion

On rehearing, a divided Ninth Circuit held that a clearly labeled list of substitute products is not trademark infringement—narrowing initial-interest confusion for the era of online search.

October 13, 2025
Online shopper viewing a grid of product search results on a laptop
A search for a brand Amazon did not carry returned labeled rivals—prompting a fight over whether that momentary diversion is actionable confusion. Shutterstock
Educational content, not legal advice. This article explains general legal concepts. It does not create an attorney–client relationship. For your specific situation, consult a licensed attorney.

In Multi Time Machine, Inc. v. Amazon.com, Inc., No. 13-55575 (9th Cir. Oct. 21, 2015), the United States Court of Appeals for the Ninth Circuit confronted a question the boat-and-camera cases of an earlier era never imagined: when a shopper searches an online marketplace for a brand the marketplace does not carry, and the site responds with a labeled list of competing products, has the marketplace committed trademark infringement? In a 2-1 superseding opinion authored by Circuit Judge Barry Silverman, the court answered no. Amazon’s search results “clearly label the name and manufacturer of each product offered for sale” and include photographs, so “no reasonably prudent consumer accustomed to shopping online would likely be confused as to the source.” The decision affirmed summary judgment for Amazon—and sharply curtailed the doctrine of initial-interest confusion in the digital marketplace.

At a glance

  • Case: Multi Time Machine, Inc. v. Amazon.com, Inc., No. 13-55575 (9th Cir. 2015) (Silverman, J.; Bea, J., dissenting).
  • Operative opinion: The amended/superseding opinion issued October 21, 2015, which replaced an earlier July 6, 2015 panel decision that had come out the other way.
  • Posture: The district court granted summary judgment to Amazon; the panel first reversed, then on rehearing withdrew that opinion and affirmed.
  • Holding: A clearly labeled online search-results page displaying competitors’ goods does not create a likelihood of confusion, even when the searched-for brand is absent and the page never says so.
  • The flip: Judge Silverman, who had dissented from the July opinion, wrote the October majority; Judge Gordon Quist switched his vote; Judge Carlos Bea, formerly in the majority, dissented at length.
  • Aftermath: The Supreme Court denied certiorari in 2016, leaving the affirmance in place.
  • Why it matters: The case is the Ninth Circuit’s leading modern statement on keyword/search confusion and a significant narrowing of initial-interest confusion.

The facts are simple and characteristic of e-commerce. Multi Time Machine makes premium military-style watches sold under the federally registered mark MTM SPECIAL OPS, retailing for several hundred dollars. MTM does not sell through Amazon, and—by choice—restricts its distribution. A shopper who typed “mtm special ops” into Amazon’s search box would see not a “no results” message but a results page populated with other manufacturers’ military-style watches, each shown with its brand, manufacturer, and photograph. MTM sued, arguing that by returning rivals without disclosing that it did not carry MTM, Amazon created a likelihood of initial-interest confusion.

Initial-interest confusion, and why the majority rejected it here

Initial-interest confusion is the theory that liability can attach when a defendant uses a mark to capture a consumer’s initial attention—drawing the shopper in—even if any confusion is dispelled before purchase. It is the doctrine behind classic cases about diverting customers with a competitor’s mark, and MTM built its claim on it: the labeled rivals, MTM argued, would not have appeared but for the lure of its mark, and the momentary diversion was itself the harm.

The majority reframed the inquiry around the realities of online shopping. The dispositive question, Judge Silverman wrote, is not whether Amazon “used” the mark by returning results, but “what the consumer saw on the screen and reasonably believed, given the context.” On that screen, every product was labeled with its own brand and maker and pictured. A reasonably prudent online shopper—an increasingly savvy figure in the court’s view—would understand the page as a display of what Amazon does sell, not a representation that those goods come from or are sponsored by MTM. Because the labeling defeated any reasonable belief about source, there was no likelihood of confusion to send to a jury.

Notably, the majority declined to grind mechanically through every Sleekcraft factor. It treated the multifactor test as a flexible tool, not a rigid checklist, and reasoned that where clear labeling resolves the source question, the framework’s factors need not be marched through one by one. That methodological move—deciding confusion on the screen rather than on a factor tally—is part of why the opinion remains influential and contested.

Judge Bea’s dissent

Circuit Judge Carlos Bea dissented at length, and his opinion is essential reading for anyone tracking the doctrine. He charged that the majority had, “sub silentio,” effectively overruled the circuit’s initial-interest-confusion doctrine. In his view, whether a reasonably prudent online shopper would be confused is a fact question for a jury, not something an appellate court should resolve on summary judgment—particularly given that Amazon’s page never told the shopper that MTM watches were unavailable, while competitors’ rival products sat where the searched-for brand should have been.

Bea’s critique frames the central tension: the majority’s “reasonably prudent online consumer” is sophisticated enough to read labels and infer substitution, while the dissent’s consumer might be momentarily diverted by the unexplained appearance of rivals. Which consumer the law assumes effectively decides the case—and the choice is contestable, which is why the panel itself flipped between July and October.

A case that reversed itself

The procedural history is its own lesson. In the July 6, 2015 opinion, a different majority (Judge Bea, joined by Judge Quist) found triable issues and reversed summary judgment for Amazon; Judge Silverman dissented. On rehearing, Judge Quist changed his vote, the panel withdrew the July opinion, and the October 21, 2015 superseding opinion—Silverman writing, joined by Quist, Bea dissenting—affirmed for Amazon. The Supreme Court later denied review.

Self-reversal by the same panel is rare and signals genuine doctrinal difficulty. The question—how much an online platform may rely on labeling to dispel confusion—did not yield to the older factor tests cleanly, and the swing vote reflected that. For litigants, the history is a caution: confusion outcomes in novel digital contexts can be close, panel-dependent, and unstable.

Open questions

  • What counts as “clear” labeling? The majority leaned on conspicuous brand-and-photo labeling. How prominent must labels be, and does an unlabeled or ambiguously labeled results page change the answer?
  • Is initial-interest confusion still viable in the Ninth Circuit? The dissent says the majority gutted it; later courts continue to debate how much of the doctrine survives for search and keyword contexts.
  • Should confusion on novel interfaces be a jury question? The split turned on whether the “reasonably prudent online consumer” inquiry is one for judges on summary judgment or for fact-finders.

Implications

  • Labeling is a powerful defense. Online marketplaces and retailers that clearly identify each product’s brand and source have a strong argument against confusion when displaying substitutes.
  • Initial-interest confusion is narrower online. After Multi Time Machine, the mere appearance of rivals in response to a branded search is not, by itself, actionable in the Ninth Circuit.
  • The “reasonably prudent online consumer” is sophisticated. Courts increasingly assume shoppers read labels—an assumption that favors platforms and disadvantages brand owners relying on momentary diversion.
  • Factor tests bend to context. The court applied Sleekcraft flexibly, deciding the case on what the screen conveyed rather than on a rigid factor tally.
  • Outcomes can be panel-dependent. The self-reversal shows how unsettled digital confusion law remains; comparable facts may divide judges.

Frequently asked questions

What did the Ninth Circuit ultimately decide? In its October 21, 2015 superseding opinion, the court affirmed summary judgment for Amazon, holding that a clearly labeled search-results page showing competitors’ watches did not create a likelihood of confusion, even though Amazon did not carry MTM and did not say so.

What is initial-interest confusion? It is a theory of liability for using a mark to capture a consumer’s initial attention or divert interest, even if the confusion is dispelled before any purchase. MTM relied on it; the majority found it did not apply because clear labeling prevented any reasonable confusion about source.

Why did the panel reverse itself? The same three judges first ruled for MTM in July 2015, then reheard the case. Judge Quist switched his vote, the original dissenter (Judge Silverman) wrote the new majority, and the court affirmed for Amazon in October 2015. The Supreme Court later denied certiorari.

Authorities and sources