When comparing trademarks to determine if they are confusingly similar, the USPTO applies a 13-factor analysis based on the "DuPont factors," a set of factors first enumerated in a 1973 decision by the CCPA (now part of the Federal Court of Appeals) in a case where one of the parties was named DuPont (E. I. du Pont de Nemours & Co.).
While all factors may be relevant, two factors are usually determinative of the outcome: the similarity of the marks and the relatedness of the goods. The weight assigned to these two factors is interdependent of each other. The greater the similarity of the marks, the less related they need to be for the USPTO to find that they're confusingly similar, and vice versa.
Below are some legal principles the USPTO applies when assessing the first DuPont factor, the similarity of the marks.
The Test Is Not a Side-by-Side Comparison
The proper test is whether the marks are sufficiently similar in terms of their commercial impression such that persons who encounter the marks would be likely to assume a connection between the parties. The standard accounts for the fact that memory is imperfect. Consumers do not carry mental photographs of every trademark they have seen. They carry impressions.
Two marks that look distinct when placed next to each other may still be confusingly similar when evaluated the way actual consumers experience them: sequentially and with imperfect recall.
Marks Must Be Considered as a Whole
A mark must be evaluated in its entirety when assessing likelihood of confusion. That said, considering a mark as a whole does not mean every element gets equal weight.
Some Parts of a Mark Can Carry More Weight Than Others
The USPTO regularly gives more or less weight to particular elements of a mark, notwithstanding that the final conclusion is based on the mark considered in its entirety.
A common reason for giving less weight to a portion of a mark is that it is descriptive or generic with respect to the relevant goods or services. If part of a mark simply describes what is being sold, courts will focus on the distinctive parts when comparing it to a competitor's mark. For example, in Cunningham v. Laser Golf Corp. (2000), the Federal Circuit accepted that little weight could be given to SWING in a comparison of LASERSWING and LASER for golf clubs because the term was descriptive.
Another common reason is the position of an element of the mark within the mark. The beginning portion of a mark is often given more weight in a similarity analysis.
Disclaimed Words Still Count
When the USPTO requires a disclaimer of a generic or descriptive word, that word does not disappear from the confusion analysis. It is still part of the mark as far as consumers are concerned, because consumers have no idea what has been disclaimed.
In Shen Manufacturing Co. v. Ritz Hotel, Ltd. (2004), the Federal Circuit affirmed that it was proper for the TTAB to have considered the disclaimed words PARIS and HOTEL in determining that RITZ PARIS RITZ HOTEL and RITZ were sufficiently different to avoid a likelihood of confusion finding. A disclaimed term may be given less weight, but it cannot be ignored entirely.
In Word-and-Design Marks, the Words Usually Dominate
When a mark combines text and a visual design element, the verbal portion is generally considered the more important source identifier. The reasoning is practical: people are more likely to ask for a product by name than to describe its logo.
There is an important qualification. As the Federal Circuit noted in In re Electrolyte Laboratories, Inc. (1991), there is no general rule as to whether letters or design will dominate in composite marks, and dominance of one over the other is not automatically dispositive. The words-dominate principle is a default, not an absolute. If the design element is particularly striking and the words are weak or descriptive, the analysis may go differently.
When the Goods Are Nearly Identical, Less Similarity Is Needed
The similarity of marks does not exist in a vacuum. It interacts with the similarity of the goods or services at issue. The closer the goods, the less similar the marks need to be to support a finding of likelihood of confusion. Selling nearly identical products to the same consumers through the same channels means even a modest resemblance between marks can create a problem.
Famous Marks Get Extra Protection
The more famous a mark, the less similarity a junior mark needs to create a likelihood of confusion. The Federal Circuit framed this as an inverse relationship in Century 21 Real Estate Corp. v. Century Life of America (1992): the Lanham Act's tolerance for similarity between competing marks varies inversely with the fame of the prior mark. The TTAB applied this principle in Harry Winston, Inc. v. Bruce Winston Gem Corp. (2014), where the fame of the Harry Winston brand was a significant factor in the analysis.
So, a name that is even modestly similar to a well-known brand faces a harder road than it would if the prior mark were obscure. Fame expands the zone of protection.
Similarity Is a Spectrum, Not a Binary Decision
Marks are not simply similar or not. Similarity exists on a continuum, and where a mark falls on that continuum affects how the other DuPont factors are weighed. The Federal Circuit confirmed this in In re St. Helena Hospital (2014), and the TTAB applied the same framing in North Face Apparel Corp. v. Sanyang Industry Co. (2015).
A mark that is highly similar to an existing mark may outweigh other factors that would otherwise weigh against confusion. Conversely, marks that are only modestly similar may require stronger evidence of relatedness in other areas before confusion is found.
Foreign Words Get Translated
Under the doctrine of foreign equivalents, foreign words from common languages are translated into English before being compared to English word marks. For example, if a mark is the Spanish word for a common noun, the USPTO will consider whether the English translation of that noun is confusingly similar to existing marks.
The doctrine applies to common modern languages, not obscure or dead ones, and it is not applied mechanically. If consumers in the relevant market are unlikely to recognize or translate the foreign word, the doctrine may not apply. For Spanish, French, Italian, and other widely spoken languages, applicants should assume it applies.
There Is No Correct Pronunciation of a Trademark
When comparing marks for phonetic similarity, the analysis acknowledges that there is no single correct way to pronounce most trademarks. Different consumers will say the same mark differently, and the likelihood of confusion analysis accounts for the full range of reasonable pronunciations.
There is a sensible exception: when a mark is a recognized word, even in a foreign language, the no-correct-pronunciation principle does not apply. The TTAB drew this line in INTS It Is Not The Same, GmbH v. Disidual Clothing, LLC (2017). If consumers know how to say a word, the analysis focuses on that pronunciation.
Pure Design Marks Are Compared Visually
When neither mark contains words and neither can be spoken, the similarity analysis is necessarily visual. Courts compare the overall appearance, the impression created by the design elements, and whether the two marks would be mistaken for each other by a consumer who has seen one and later encounters the other.
Identical Marks on Related Goods Are Usually Dispositive
When two word marks are identical, and neither is suggestive nor descriptive of the relevant goods, the similarity factor weighs heavily in favor of confusion. For an example, see In re i.am.symbolic, llc (2017).
If two companies use the exact same arbitrary word as their brand name in related markets, confusion is almost inevitable. The strength of the marks and the relatedness of the goods may affect how the other factors are weighed, but starting with identical arbitrary marks is a very difficult position to defend.
Sometimes One Factor Is Enough
While the likelihood of confusion analysis involves balancing multiple factors, a single factor can be dispositive when it is strong enough. The most common example is dissimilarity: if the marks are sufficiently different, no amount of relatedness in the goods or other overlapping factors will establish likelihood of confusion. A finding of clear dissimilarity can end the analysis entirely.
Managing Likelihood of Confusion Problems
A professional clearance search before filing is the most reliable way to identify conflicts before they become refusals. If you have already received a likelihood of confusion Office action, the response process involves assessing each of the DuPont factors and building the strongest available argument against the refusal.
If you are dealing with a likelihood of confusion issue or assessing whether a name is available, book a free consultation.
.svg)