Ezaki Gliko Kabushiki Kaisha v. Lotte International America Co

PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
_______________
No. 19-3010
_______________
EZAKI GLICO KABUSHIKI KAISHA,
a Japanese Corporation d/b/a Ezaki Glico;
EZAKI GLICO USA CORP.,
a California Corporation,
Appellants
v.
LOTTE INTERNATIONAL AMERICA CORP.;
LOTTE CONFECTIONARY CO. LTD.
_______________
On Appeal from the United States District Court
for the District of New Jersey
(D.C. No. 2-15-cv-05477)
District Judge: Honorable Madeline C. Arleo
_______________
Argued: July 9, 2020
Before: McKEE, BIBAS, and FUENTES, Circuit Judges
(Filed: January 26, 2021)
_______________
Jessica L. Ellsworth [ARGUED]
Anna K. Shaw
Neal K. Katyal
Benjamin A. Field
Hogan Lovells US
555 Thirteenth Street, N.W.
Columbia Square
Washington, DC 20004
Steven M. Levitan
Hogan Lovells US
4085 Campbell Avenue
Menlo Park, CA 94025
Aaron S. Oakley
Hogan Lovells US
1601 Wewatta Street
Denver, CO 80202
Katherine B. Wellington
Hogan Lovells US
125 High Street, Suite 2010
Boston, MA 02110
Roy H. Wepner
Lerner David Littenberg Krumholz & Mentlik
20 Commerce Drive
Cranford, NJ 07016
Counsel for Appellants
2
John J. Dabney [ARGUED]
Mary D. Hallerman
Snell & Wilmer
2001 K Street, N.W.
Suite 425 North
Washington, DC 20006
Counsel for Appellees
David H. Bernstein
Debevoise & Plimpton
919 Third Avenue
New York, NY 10022
Bruce R. Ewing
Dorsey & Whitney
51 West 52nd Street
New York, NY 10019
Jonathan E. Moskin
Foley & Lardner
90 Park Avenue
New York, NY 10016
Claudia E. Ray
Kirkland & Ellis
601 Lexington Avenue
New York, NY 10022
Counsel for Amicus Curiae in support of Appellants
International Trademark Association, on Petition for
Rehearing
3
Theodore H. Davis, Jr.
Kilpatrick Townsend
1100 Peachtree Street, Suite 2800
Atlanta, GA 30309
Sara K. Stadler
Kilpatrick Townsend
1114 Avenue of the Americas
The Grace Building, 21st Floor
New York, NY 10036
Counsel for Amicus Curiae in support of Appellants
National Confectioners Association, on Petition for
Rehearing
Adam H. Charnes
Kilpatrick Townsend
2001 Ross Avenue, Suite 4400
Dallas, TX 75201
Charles H. Hooker, III
Kilpatrick Townsend
1100 Peachtree Street, Suite 2800
Atlanta, GA 30309
J. David Mayberry
Kilpatrick Townsend
1114 Avenue of the Americas
The Grace Building, 21st Floor
New York, NY 10036
Counsel for Amicus Curiae in support of Appellants
Mondelez Global LLC, on Petition for Rehearing
4
_______________
OPINION OF THE COURT
_______________
BIBAS, Circuit Judge.
This is a tale of more than just desserts. Decades ago, Ezaki
Glico invented Pocky, a chocolate-covered cookie stick. Pocky
was very popular. And its success drew imitators, including
Lotte’s Pepero. Ezaki Glico now sues Lotte for trade-dress in-
fringement.
The District Court granted Lotte summary judgment, find-
ing that because Pocky’s design is functional, Ezaki Glico has
no trade-dress protection. We agree. Trade dress is limited to
designs that identify a product’s source. It does not safeguard
designs that are functional—that is, useful. Patent law protects
useful inventions, but trademark law does not. We will thus
affirm.
I. BACKGROUND
A. A cookie is born: Ezaki Glico’s Pocky
Ezaki Glico is a Japanese confectionery company. For more
than half a century, it has made and sold Pocky: a product line
of thin, stick-shaped cookies (what the British call biscuits).
These cookies are partly coated with chocolate or a flavored
cream; some have crushed almonds too. The end of each is left
partly uncoated to serve as a handle. Ezaki Glico makes Pocky
in both a standard and an “Ultra Slim” size. Appellant’s Br. 9.
5
In 1978, Ezaki Glico started selling Pocky in the United
States through its wholly owned subsidiary here. Since then, it
has tried to fend off competitors by registering U.S. trademarks
and patents. It has two Pocky product configurations registered
as trade dresses.
Ezaki Glico also has a utility patent for a “Stick Shaped
Snack and Method for Producing the Same.” App. 1013–16.
The first thirteen claims in the patent describe methods for
making a stick-shaped snack. The final claim covers “[a] stick-
shaped snack made by the method of claim 1.” App. 1016. The
width of that stick-shaped snack matches that of Pocky Ultra
Slim.
B. A new cookie comes to town: Lotte’s Pepero
Imitation is the sincerest form of flattery, and others have
noted Pocky’s appeal. Starting in 1983, another confectionery
company called Lotte started making Pepero. These snacks are
also stick-shaped cookies (biscuits) partly coated in chocolate
or a flavored cream, and some have crushed almonds too. It
looks remarkably like Pocky. Here are the two products side
by side:
6
See App. 980–83, 1018–19, 1021–24. Lotte and its U.S. sub-
sidiary have been selling Pepero in the United States for more
than three decades.
C. Ezaki Glico’s trade-dress suit
From 1993 to 1995, Ezaki Glico sent letters to Lotte, noti-
fying Lotte of its registered trade dress and asking it to cease
and desist selling Pepero in the United States. Lotte assured
Ezaki Glico that it would stop until they resolved their dispute.
But Lotte resumed selling Pepero. For the next two decades,
Ezaki Glico took no further action.
In 2015, Ezaki Glico sued Lotte in federal court for selling
Pepero. Under federal law, Ezaki Glico alleged trademark
7
infringement and unfair competition, in violation of the Lan-
ham (Trademark) Act §§ 32 and 43(a), 15 U.S.C. §§ 1114,
1125(a)(1)(A). Under New Jersey law, it alleged trademark in-
fringement and unfair competition, in violation of both the
common law and the New Jersey Fair Trade Act, N.J.S.A.
§ 56:4-1 and 2.
After discovery, the District Court granted summary judg-
ment for Lotte, holding that because Pocky’s product configu-
ration is functional, it is not protected as trade dress. Kaisha v.
Lotte Int’l Am. Corp., No. 15-5477, 2019 WL 8405592, at *3
(D.N.J. July 31, 2019).
Ezaki Glico now appeals. The District Court had jurisdic-
tion under 15 U.S.C. §§ 1119 and 1121(a) and 28 U.S.C.
§§ 1331, 1338, and 1367. We have jurisdiction under 28 U.S.C.
§ 1291.
We review the District Court’s grant of summary judgment
de novo. Cranbury Brick Yard, LLC v. United States, 943 F.3d
701, 708 (3d Cir. 2019). We will affirm if no material fact is
genuinely disputed and if, viewing the facts most favorably to
Ezaki Glico, Lotte merits judgment as a matter of law. Fed. R.
Civ. P. 56(a). Both of Ezaki Glico’s Lanham Act claims de-
pend on the validity of its trade dress. New Jersey’s unfair-
competition and trademark laws are not significantly different
from federal law, so our analysis of Ezaki Glico’s Lanham Act
claims applies equally to dispose of its state-law claims. See
Am. Greetings Corp. v. Dan-Dee Imports, Inc., 807 F.2d 1136,
1141 (3d Cir. 1986); 3 J. Thomas McCarthy, McCarthy on
Trademarks and Unfair Competition § 22:1.50 (5th ed. 2020).
Following the parties’ lead, we focus on federal trademark law.
8
II. TRADE-DRESS LAW DOES NOT PROTECT PRODUCT
DESIGNS THAT ARE USEFUL
Under the statute, the key issue is whether Pocky’s trade
dress is functional. Lotte says that it is; Ezaki Glico says no.
Ezaki Glico equates “functional” with “essential.” Appellants’
Br. 18, 25 (emphases omitted). But that test is too narrow. It
misreads the Lanham Act’s text and its relationship with the
Patent Act. Under both the statute and the case law, a feature’s
particular design is functional if it is useful. And there are sev-
eral ways to show functionality.
A. Patent law protects useful designs, while trademark
law does not
Copying is usually legal. It is part of market competition.
As a rule, unless a patent, copyright, or the like protects an
item, competitors are free to copy it. TrafFix Devices, Inc. v.
Mktg. Displays, Inc., 532 U.S. 23, 29 (2001).
The Constitution does authorize Congress to grant exclu-
sive patents and copyrights “[t]o promote the Progress of Sci-
ence and useful Arts,” but only “for limited Times.” U.S.
Const. art. I, § 8, cl. 8. Utility patents promote “Science and
useful Arts” by protecting inventions that are “new and use-
ful.” 35 U.S.C. § 101. Design patents protect “any new, original
and ornamental design.” Id. § 171(a). In keeping with the Con-
stitution’s time limit, utility patents last for twenty years, and
design patents last for only fifteen years. Id. §§ 154(a)(2), 173.
If there is no patent, or once a patent expires, competitors are
free to copy “publicly known design and utilitarian ideas.” Bo-
nito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 152
9
(1989); accord Qualitex Co. v. Jacobson Prod. Co., Inc., 514
U.S. 159, 164 (1995). This way, sellers can compete and build
on one another’s innovations. That competition improves qual-
ity and lowers consumers’ costs.
By contrast, trademark law protects not inventions or de-
signs per se, but branding. A trademark is a “word, name, sym-
bol, or device . . . used by a person[ ] . . . to identify and distin-
guish his or her goods . . . from those manufactured or sold by
others and to indicate the source of the goods.” 15 U.S.C.
§ 1127. Trademark law can protect a product’s “trade dress[,]
[which] is the overall look of a product or business.” Fair Wind
Sailing, Inc. v. Dempster, 764 F.3d 303, 308 (3d Cir. 2014).
That includes not only a product’s packaging but also its de-
sign, such as its size, shape, and color. Id.; Wal-Mart Stores,
Inc. v. Samara Bros., 529 U.S. 205, 209 (2000).
We are careful to keep trademark law in its lane. Trade
dress, like trademark law generally, is limited to protecting the
owner’s goodwill and preventing consumers from being con-
fused about the source of a product. Shire US Inc. v. Barr
Labs., Inc., 329 F.3d 348, 353 (3d Cir. 2003). We must not
overextend it to protect all of a product’s features, because
“product design almost invariably serves purposes other than
source identification.” TrafFix, 532 U.S. at 29 (quoting Wal-
Mart, 529 U.S. at 213). “Trade dress protection . . . is not in-
tended to create patent-like rights in innovative aspects of
product design.” Shire, 329 F.3d at 353. If it did, it could over-
ride restrictions on what is patentable and for how long. Qual-
itex, 514 U.S. at 164–65. After all, trademarks have no time
limit.
10
The functionality doctrine keeps trademarks from usurping
the place of patents. The Patent and Trademark Office cannot
register any mark that “comprises any matter that, as a whole,
is functional.” 15 U.S.C. § 1052(e)(5). Even after a mark is reg-
istered, it is a defense to infringement “[t]hat the mark is func-
tional.” Id. § 1115(b)(8); see also id. § 1125(a)(3) (providing
that the holder of an unregistered mark must prove that the
mark “is not functional”). Thus, even if copying would confuse
consumers about a product’s source, competitors may copy un-
patented functional designs.
B. Functional designs need not be essential,
just useful
The core dispute here is how to define “functional.” Ezaki
Glico reads it narrowly, equating it with “essential.” Appel-
lant’s Br. 18, 25. But that is not what the word means.
Since the Lanham Act does not define functionality, we
start with its ordinary meaning. A feature’s design is functional
if it is “designed or developed chiefly from the point of view
of use: UTILITARIAN.” Functional (def. 2a), Webster’s Third
New International Dictionary (1966). So something is func-
tional as long as it is “practical, utilitarian”—in a word, useful.
Functional (def. 2d), Oxford English Dictionary (2d ed. 1989).
The word requires nothing more.
Reading functionality as usefulness explains how the Lan-
ham Act fits with the Patent Act. Utility patents, not trade-
marks, protect inventions or designs that are “new and useful.”
35 U.S.C. § 101. If the Lanham Act protected designs that were
useful but not essential, as Ezaki Glico claims, it would invade
11
the Patent Act’s domain. Because the Lanham Act excludes
useful designs, the two statutes rule different realms.
Precedent also supports defining functional as useful. In
Qualitex, the Supreme Court described the functionality doc-
trine as protecting competition by keeping a producer from per-
petually “control[ling] a useful product feature.” 514 U.S. at
164. In TrafFix, the Court described functionality as depending
on whether “the feature in question is shown as a useful part of
the invention.” 532 U.S. at 34. It contrasted functional features
disclosed in a utility patent with “arbitrary, incidental, or orna-
mental aspects” that “do not serve a purpose within the terms
of the utility patent.” Id. And in Wal-Mart, the Court contrasted
designs that only “identify the source” with those that “render
the product itself more useful or more appealing.” 529 U.S. at
213. “[M]ore useful or more appealing” is a far cry from es-
sential.
Conversely, a design is not functional if all it does is iden-
tify its maker. “Proof of nonfunctionality generally requires a
showing that the element of the product serves no purpose
other than identification.” Keene Corp. v. Paraflex Indus., Inc.,
653 F.2d 822, 826 (3d Cir. 1981) (quoting SK&F, Co. v. Premo
Pharm. Labs., Inc., 625 F.2d 1055, 1063 (3d Cir. 1980)). But
if a design gives a product an edge in usefulness, then it is func-
tional.
Ezaki Glico resists this reading by focusing on one phrase
from Qualitex. The heart of its claim is the first sentence of its
argument: “A product’s configuration is functional for pur-
poses of trade dress protection only ‘if it is essential to the use
or purpose of the article or if it affects the cost or quality of the
12
article.’ ” Appellants’ Br. 22 (quoting Qualitex, 514 U.S. at
165, and adding the emphasis). But the word “only” is nowhere
on the page it cites. Though Ezaki Glico’s forceful brief repeats
“essential” more than four dozen times and structures its case
around that touchstone, the authority does not support its drum-
beat.
On the contrary, the Supreme Court recognizes several
ways to show that a product feature is functional. One way is
indeed to show that a feature “is essential to the use or purpose
of the article.” Qualitex, 514 U.S. at 165 (quoting Inwood
Labs., Inc. v. Ives Labs., Inc., 456 U.S. 844, 850 n.10 (1995)).
Another is if “it affects the cost or quality of the article.” Id.(Ezaki Glico keeps skipping over this part of the test.) At least
in some cases, a feature is functional and unprotected if the
“exclusive use of [the feature] would put competitors at a sig-
nificant non-reputation-related disadvantage.” TrafFix, 532
U.S. at 32 (quoting Qualitex, 514 U.S. at 165). All of these are
different ways of showing usefulness. (Though this last inquiry
is especially apt for proving aesthetic functionality, the Court
has not specifically limited it to that context. See TrafFix, 532
U.S. at 33.) On the other hand, a feature is “not functional” if,
for instance, “it is merely an ornamental, incidental, or arbi-
trary aspect of the device.” Id. at 30.
We analyze functionality not at the level of the entire prod-
uct or type of feature, but at the level of the particular design
chosen for feature(s). Just “because an article is useful for
some purpose,” it does not follow that “all design features of
that article must be ‘functional.’ ” 1 McCarthy § 7:70 (empha-
ses added). The question is not whether the product or feature
13
is useful, but whether “the particular shape and form” chosen
for that feature is. Id.
For instance, though ironing-board pads need “to use some
color . . . to avoid noticeable stains,” there is no functional rea-
son to use green-gold in particular. Qualitex, 514 U.S. at 166.
Though French press coffeemakers need some handle, there is
no functional reason to design the particular handle in the shape
of a “C.” Bodum USA, Inc. v. A Top New Casting Inc., 927 F.3d
486, 492–93 (7th Cir. 2019) (also noting that the design sacri-
ficed ergonomics). And though armchairs need some armrest,
there is no functional reason to design the particular armrest as
a trapezoid. Blumenthal Distrib., Inc. v. Herman Miller, Inc.,
963 F.3d 859, 867–68 (9th Cir. 2020) (also noting that the de-
sign sacrificed comfort). Ironing-board colors, coffee-pot han-
dles, and armrests are all generally useful. But the particular
designs chosen in those cases offered no edge in usefulness.
Also, a combination of functional and non-functional fea-
tures can be protected as trade dress, so long as the non-func-
tional features help make the overall design distinctive and
identify its source. See Am. Greetings, 807 F.2d at 1143.
But a product’s design, including its shape, is often useful
and thus functional. For example, when Nabisco sued Kellogg
for making its shredded wheat pillow-shaped, just like
Nabisco’s, the Supreme Court rejected the unfair-competition
claim. The pillow shape is functional because using another
shape would increase shredded wheat’s cost and lower its qual-
ity. Kellogg Co. v. Nat’l Biscuit Co., 305 U.S. 111, 122 (1938).
For the same reason, the Court rejected a challenge to copying
the exact shape of a pole lamp. Sears, Roebuck & Co. v. Stiffel
14
Co., 376 U.S. 225, 231–32 (1964). And if an inventor created
a new light-bulb shape that improved illumination, he could
not trademark that shape. Qualitex, 514 U.S. at 165. That
would be true even if consumers associated the bulb shape with
its inventor, because trademarking it would “frustrat[e] com-
petitors’ legitimate efforts to produce an equivalent illumina-
tion-enhancing bulb.” Id. So long as the design improves cost,
quality, or the like, it cannot be protected as trade dress. The
shape need only be useful, not essential. Conversely, a distinc-
tive logo, pattern, or other arbitrary shape or style may be non-
functional and protectable as a trade dress.
As the leading trademark treatise concurs, “functional”
means useful. “To boil it down to a phrase: something is ‘func-
tional’ if it works better in this shape.” 1 McCarthy § 7:63. That
includes features that make a product cheaper or easier to make
or use. Id. Because the functionality bar is supposed to keep
“trade dress from creating ‘back-door patents,’ . . . . the test of
what is ‘functional’ should be very similar to that of patent
law.” Id. § 7:67.
C. Evidence of functionality
There are several ways to prove functionality. First, evi-
dence can directly show that a feature or design makes a prod-
uct work better. See Am. Greetings, 807 F.2d at 1142 (treating
as functional “tummy graphics” on teddy bears because they
signal each bear’s personality). Second, it is “strong evidence”
of functionality that a product’s marketer touts a feature’s use-
fulness. Id. at 1142–43. Third, “[a] utility patent is strong evi-
dence that the features therein claimed are functional.”
TrafFix, 532 U.S. at 29. Fourth, if there are only a few ways to
15
design a product, the design is functional. Keene, 653 F.2d at
827. But the converse is not necessarily true: the existence of
other workable designs is relevant evidence but not inde-
pendently enough to make a design non-functional. Id.; 1
McCarthy § 7:75 (interpreting TrafFix, 532 U.S. at 33–34).
Our list is not exhaustive; there may be other considera-
tions. The Federal Circuit and other sister circuits also use sim-
ilar inquiries. See, e.g., Georgia-Pac. Consumer Prods. LP v.
Kimberly-Clark Corp., 647 F.3d 723, 727–28 (7th Cir. 2011);
Disc Golf Ass’n, Inc. v. Champion Discs, Inc., 158 F.3d 1002,
1006 (9th Cir. 1998); In re Morton-Norwich Prods., Inc., 671
F.2d 1332, 1340–41 (C.C.P.A. 1982) (predecessor to the Fed-
eral Circuit).
With these definitions and inquiries in mind, we can now
apply them to this case.
III. POCKY’S TRADE DRESS IS FUNCTIONAL
To decide whether a trade dress is functional, we look at the
usefulness of the exact feature or set of features claimed by the
trade dress. See Am. Greetings, 807 F.2d at 1141. Ezaki Glico
has two registered Pocky trade dresses, both broad. The first
“comprises an elongated rod comprising biscuit or the like,
partially covered with chocolate.” App 10, 1448. The second
consists of the same sort of snack, along with almonds on top
of the chocolate or cream.
16
In a picture, Ezaki Glico’s trade dresses include all cookies
like these:
App. 292. The trade dresses are presumptively valid because
they are registered and incontestable. See 15 U.S.C. § 1115. So
Lotte bears the burden of proving that they are functional. Id.§§ 1115(a), (b)(8).
Ezaki Glico argues that none of these features is essential
to make the snack easy to eat. But that is the wrong test. Lotte
has shown that Pocky’s design is useful and thus functional.
A. Pocky’s design makes it work better as a snack
Every feature of Pocky’s registration relates to the practical
functions of holding, eating, sharing, or packing the snack.
Consider each stick’s uncoated handle. Ezaki Glico’s internal
documents show that it wanted to make a snack that people
could eat without getting chocolate on their hands. Pocky was
born when Ezaki Glico found that it could coat just part of a
17
cookie stick, leaving people an uncoated place to hold it. So it
designed Pocky’s handle to be useful.
The same is true of Pocky’s stick shape. As Ezaki Glico
recognizes, the stick shape makes it “easy to hold, so it c[an]
be shared with others to enjoy as a snack.” App. 595. It also
lets people eat the cookie without having to open their mouths
wide. And the thin, compact shape lets Ezaki Glico pack many
sticks in each box, enough to share with friends.
Viewed as a whole, Pocky’s trade dress is functional. The
claimed features are not arbitrary or ornamental flourishes that
serve only to identify Ezaki Glico as the source. The design
makes Pocky more useful as a snack, and its advantages make
Pocky more appealing to consumers for reasons well beyond
reputation. See Kellogg, 305 U.S. at 120. As Ezaki Glico’s own
documents acknowledge, “Pocky provides a functional value
[Enjoy chocolate lightly].” App. 636 (bracketed material in
original).
B. Ezaki Glico promotes Pocky’s utilitarian
advantages
There is plenty of evidence that Ezaki Glico promotes
Pocky’s “convenient design.” App. 646. Its ads tout all the use-
ful features described above. It advertises “the no mess handle
of the Pocky stick,” which “mak[es] it easier for multi-tasking
without getting chocolate on your hands.” App. 648, 651. It
also describes Pocky as “[p]ortable,” since “one compact,
easy-to-carry package holds plentiful amounts of Pocky.” App.
648. “With plenty of sticks in each package, Pocky lends itself
18
to sharing anytime, anywhere, and with anyone.” App. 655.
These promotions confirm that Pocky’s design is functional.
C. There are alternative designs, but that does not
make Pocky’s design non-functional
Lotte could have shaped its Pepero differently. Ezaki Glico
offers nine examples of partly-chocolate-coated snacks that do
not look like Pocky. That is hardly dispositive. As we noted in
Keene, even when there are alternatives, the evidence can still
show that a product design is functional. 653 F.3d at 827. That
is true here. Every aspect of Pocky is useful. The nine other
designs do not make it less so.
D. Ezaki Glico’s utility patent for a manufacturing
method is irrelevant
Finally, Lotte argues that Ezaki Glico’s utility patent for a
“Stick Shaped Snack and Method for Producing the Same”
proves functionality. It does not.
As TrafFix explained, “[a] utility patent is strong evidence
that the features therein claimed are functional.” 532 U.S. at
29. This is because patented items must be “useful.” 35 U.S.C.
§ 101. If a patentee relied on a product’s feature to show that
the product was patentable, that reliance is good evidence that
the feature is useful. As TrafFix put it, the question is whether
the “central advance” of the utility patent is also “the essential
feature of the trade dress” that the owners want to protect. 532
U.S. at 30. So Ezaki Glico’s utility patent would be strong ev-
idence of functionality if the features it claimed overlapped
with its trade dress. But they do not.
19
The trade dress that Ezaki Glico defends is a stick-shaped
snack that is partly coated with chocolate or cream. Yet those
features are not the “central advance” of its utility patent. In-
stead, the patent’s innovation is a better method for making the
snack’s stick shape. The method is useful for making the shape
whether or not the shape itself is useful for anything. Thus, the
patent’s mention of the shape says nothing about whether the
shape is functional.
The District Court erroneously considered the utility patent.
But that error was immaterial. Even setting that aside, many
other factors show that Pocky’s trade dress is functional and so
not protectable. Thus, the District Court properly granted sum-
mary judgment for Lotte. We need not reach other possible
grounds for affirmance.
* * * * *
Though Ezaki Glico created Pocky, it cannot use trade
dress protection to keep competitors from copying it. The Lan-
ham Act protects features that serve only to identify their
source. It does not cover functional (that is, useful) features.
That is the domain of patents, not trademarks. Ezaki Glico has
not borne its burden of showing nonfunctionality. TrafFix, 532
U.S. at 32. There is no real dispute that Pocky’s design is use-
ful, so the trade dress is not protectable. We will thus affirm.
That’s the way the cookie crumbles.
20