TTAB Affirms Section 2(e)(5) Functionality Refusal of Serrated Knife Blade
Relying on Applicant's own advertising touting the design of its knife blade and its ease of manufacture (yielding lower cost), the Board affirmed Examining Attorney Caroline L. Moran's Section 2(e)(5) functionality refusal of the design of a "serration pattern" on the blade of a knife (shown below). Applicant GB's design patent served as some evidence of non-functionality, but was outweighed by GB's own advertising touting the design's utilitarian benefits. [I suppose one could say that GB stabbed itself in the foot - ed.]. For the purpose of completeness, the Board also considered Applicant GB's claim of acquired distinctiveness but found the evidence insufficient to support the Section 2(f) claim. In re GB II Corporation d/b/a Columbia River Knife & Tool Company, Serial No. 88730809 (February 8, 2022) [not precedential] (Opinion by Judge Jonathan Hudis).
Functionality: In determining the functionality issue, the Board looked to the ever popular Morton-Norwich factors:
(1) the existence of a utility patent that discloses the utilitarian advantages of the subject matter sought to be registered; (2) advertising by the applicant that touts the utilitarian advantages of the subject matter; (3) facts pertaining to the availability of alternative designs; and (4) facts pertaining to whether the subject matter results from a comparatively simple or inexpensive method of manufacture.
GB described the mark (in its brief on appeal) as "the distinct angled shape for the teeth and the angled, semi-circular beveled portions disposed above the cutting edge." Its advertising stated, among other things, that the serrations "make[] for superior cutting of cordage, rope, and fibrous materials," and that they are "easy to sharpen and they work better than traditional serrations."
GB submitted a copy of its design patent covering the same design, arguing that the patent weighs against functionality, since design patents "by definition protect only ornamental and nonfunctional designs." [N.B.: "functionality" in the design patent realm is a very narrow doctrine, much different from "functionality" in the trademark realm. See TTABlog comment below. - ed.]. The Board agreed that the design patent weighs in favor of applicant, but it "can be outweighed by other evidence supporting the functionality determination."
The Board found GB's self-laudatory advertising to be a "strong factor" in favor of the functionality finding. GB, as is typical, claimed that the advertising was "mere puffery," but the Board pooh-poohed that contention.
Although it need not consider alternative designs if the design at issue is found to be functional based on other evidence, the Board considered several third-party knife designs, "for the sake of completeness as 'part of the overall mix'" and because these designs "further support[s] our finding that Applicant's porposed mark is functional." The Board found that the alternative designs "do not work equally well and are not equally cost effective," and so they supported the refusal to register.
Although GB, during prosecution, claimed that its design required a complicated and expensive manufacturing process, its advertising said the opposite, and so the Board found that the subject knife blade design "results from a comparatively simple or inexpensive method of manufacture."
And so, the Board affirmed the refusal under Section 2(e)(5).
Acquired Distinctiveness: Although the Section 2(e)(5) affirmance was dispositive, the Board chose to consider Applicant GB's Section 2(f) claim. Of course, product designs are not registrable as trademarks without proof of acquired distinctiveness. Wal-mart. And an applicant for registration of a product design mark "faces a heavy burden" of proof under Section 2(f).
The Board considered the six factors set forth in Converse, Inc. v. ITC, 909 F.3d 1110, 128 USPQ2d 1538, 1546 (Fed. Cir. 2018):
(1) Association with a particular source by actual purchasers (typically measured by customer surveys); (2) Length, degree, and exclusivity of use; (3) Amount and manner of advertising; (4) Amount of sales and number of customers; (5) Intentional copying; and (6) Unsolicited media coverage of the product embodying the mark.
GB did not provide any evidence as the fourth, fifth, and sixth factors. It did claim five years of continuous and substantially exclusive use in its application, but that failed to satisfy GB's heavy burden of proof. Even the 15 years of use claimed in its brief (without support) was not enough. The only evidence left was GB's advertising – which failed to instruct consumers to "look for" the serrated design and failed to promote the design as a trademark – and a handful of responses on the Reddit website to a question posted by applicant. However, the Reddit evidence came nowhere near qualifying as a reliable survey and so it was not probative.
The Board concluded that applicant's serrated design has not acquired distinctiveness, and so it affirmed the USPTO's alternative refusal under Sections 1, 2, and 45.
Read comments and post your comment here.
TTABlogger comment: In my view, the existence of a design patent should have no bearing on the issue of whether the proposed mark is functional in a trademark sense because the respective standards for functionality are different. See Sarah Burstein, Faux Amis in Design Law, 105 TMR 1455 (2015). [pdf here]
.Text Copyright John L. Welch 2022.
3 Comments:
This is probably a broader discussion, but *why* are the tests for ornamentation/functionality different between trade dress and design patent? Isn't the purpose of both doctrines the exact same -- to prevent people from receiving protection for "utilitarian" designs that should instead be covered by utility patents?
[Typos all corrected, I think!]
The article you linked makes a great point. Putting that aside for the moment, here’s another reason why the existence of a design patent shouldn’t be evidence of non-functionality, no matter how close it is. In registrability cases, the Federal Circuit makes absolutely clear that the PTO’s registration of a similar mark, whether of a third party or of the applicant itself (see, e.g., In re Cordua Rests., Inc., 823 F.3d 594, 600 (Fed. Cir. 2016)), isn’t binding on the agency in examining a later application. Each application must be examined for compliance with the governing statutory provisions, irrespective of whether the agency erroneously registered the prior mark. The decision by a single examining attorney to allow registration of the prior mark, even one having the exact same flaw, is legally irrelevant. Why is a decision by a single design patent examiner to grant a design patent not subject to the same rule?
The Burstein article explains the difference. I will try. Design patent functionality requires that there be only one way to design something: e.g., a key for a certain keyway. Very, very few design patent applications are rejected on functionality grounds. Not so with trademark functionality. A product design can be functional in a trademark sense even though it is not the only design that fits the purpose.
Post a Comment
<< Home