Roar Spirits v. Sutter Home Winery — Ninth Circuit Revives EL BANDIDO YANKEE Trademark Applications

Background

Roar Spirits sells tequila under the name EL BANDIDO YANKEE. It applied to register that word mark and a composite mark reading EL BANDIDO YANKEE TEQUILA COMPANY BLANCO. Sutter Home Winery opposed both applications, arguing that consumers were likely to confuse Roar’s marks with Sutter’s registered BANDIT word mark.

The Trademark Trial and Appeal Board agreed with Sutter. Roar then sought de novo review in federal district court, where the parties could introduce additional evidence. The district court granted summary judgment to Sutter. Roar appealed, relying in part on years of simultaneous sales and a consumer survey that reported no respondents identifying Sutter as the source of Roar’s tequila.

The Court’s Holding

The Ninth Circuit reversed and remanded. Likelihood of confusion is ordinarily a factual question, the panel explained, and summary judgment is uncommon when a reasonable factfinder could weigh the evidence differently. Here, a factfinder could conclude that the competing marks are sufficiently dissimilar and that the absence of actual confusion favors Roar.

Similarity was potentially decisive. EL BANDIDO YANKEE contains two more words and four more syllables than BANDIT. The added word YANKEE also changes the meaning: it modifies BANDIDO and combines references to two distinct geographic regions. The panel said those differences in appearance, sound, meaning, and overall commercial impression could permit a finding that consumers would not confuse the marks.

The district court also should not have disregarded evidence bearing on actual confusion. Roar and Sutter sold substantial quantities of their products during overlapping years, yet Roar identified no marketplace confusion. Its Eveready-style survey likewise found that none of 112 respondents named Sutter as the source of Roar’s tequila. Although registration disputes assess the marks as described in the applications, the panel held that a reasonable factfinder could still give the real-world evidence some weight.

Key Takeaways

• A shared word does not resolve similarity when additional words materially change a mark’s sound, appearance, meaning, and overall impression.
• Likelihood of confusion usually remains a fact question, making summary judgment difficult when reasonable inferences favor the nonmoving party.
• Years of overlapping sales without reported confusion, supported by a properly framed consumer survey, can help an applicant resist a registration opposition.
• The same likelihood-of-confusion standard governs registration and infringement, though the relevant marketplace evidence may be weighed differently in each setting.

Why It Matters

The unpublished decision is not binding precedent, but it offers a practical warning against isolating one common term and treating it as the whole mark. Brand owners and applicants should evaluate marks as consumers encounter them, including the effect of added words and the ideas those words convey.

It also underscores the value of building an evidentiary record before seeking summary judgment. Sales history and survey evidence may not decide a registration dispute by themselves, but they can create a genuine factual dispute that requires a trial rather than an early judgment.

Surfaced via Law360 IP.