FOR PUBLICATION

  UNITED STATES COURT OF APPEALS
       FOR THE NINTH CIRCUIT


STONE CREEK, INC., an Arizona           No. 15-17418
corporation,
                Plaintiff-Appellant,       D.C. No
                                        2:13-cv-00688-
                 v.                          DLR

OMNIA ITALIAN DESIGN, INC., a
California corporation,
                 Defendant-Appellee.



STONE CREEK, INC., an Arizona           No. 16-15304
corporation,
                           Plaintiff,      D.C. No.
                                        2:13-cv-00688-
                and                          DLR

ADAM SMITH; COREY ESCHWEILER;
JOSHUA LLOYD BENSON; MARK               ORDER AND
DOUGLAS CHESTER,                         AMENDED
                     Appellants,          OPINION

                 v.

OMNIA ITALIAN DESIGN, INC., a
California corporation,
                 Defendant-Appellee.
2          STONE CREEK V. OMNIA ITALIAN DESIGN

         Appeal from the United States District Court
                  for the District of Arizona
         Douglas L. Rayes, District Judge, Presiding

             Argued and Submitted April 6, 2017
                    Pasadena, California

                     Filed July 11, 2017
                   Amended August 30, 2017

      Before: M. Margaret McKeown and Consuelo M.
    Callahan, Circuit Judges, and Gordon J. Quist, * District
                            Judge.

                           Order;
                 Opinion by Judge McKeown




    *
      The Honorable Gordon J. Quist, United States District Judge for
the Western District of Michigan, sitting by designation.
           STONE CREEK V. OMNIA ITALIAN DESIGN                       3

                          SUMMARY **


                            Trademark

    The panel filed an order denying on behalf of the court a
petition for rehearing en banc and amending its opinion
affirming in part and reversing in part the district court’s
judgment, after a bench trial, in favor of the defendant in a
trademark infringement action under the Lanham Act.

    Defendant Omnia Italian Design, Inc., copied and began
selling the same goods branded with the mark of its (now ex)
business partner, retail furniture company Stone Creek, Inc.

    In its opinion, as amended, the panel reversed in part and
held that Omnia’s use of Stone Creek’s mark was likely to
cause confusion. The panel rejected Omnia’s invocation of
the Tea Rose-Rectanus doctrine, an affirmative defense
under which the use of a mark in a remote geographic area
is protected when the use is in good faith. Agreeing with the
Seventh and Eighth Circuits, the panel held that Omnia’s
knowledge of Stone Creek’s prior use defeated any claim of
good faith. Accordingly, Omnia was liable for infringement
of the Stone Creek mark.

    Agreeing with the Federal Circuit, the panel confirmed
that a 1999 amendment to the trademark statutes did not
sweep away precedent requiring that a plaintiff prove
willfulness to justify an award of the defendant’s profits. The
panel remanded for a determination of whether Omnia had
the requisite intent.
    **
       This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
4         STONE CREEK V. OMNIA ITALIAN DESIGN

    The panel affirmed in part and reversed in part the
district court’s imposition of sanctions under 28 U.S.C.
§ 1927.


                         COUNSEL

Joshua L. Benson (argued) and Cory M. Eschweiler, Glen
Lerner Injury Attorneys, Las Vegas, Nevada, for Plaintiff-
Appellants.

Daniel C. DeCarlo (argued), Brittany H. Bartold, Daniel R.
Lewis, and Jeffry A. Miller, Lewis Brisbois Bisgaard &
Smith LLP, Los Angeles, California, for Defendant-
Appellee.


                          ORDER

    The opinion filed on July 11, 2017, and appearing at

, is hereby amended as follows: on page 1140,
“Omnia asserts that its use of Stone Creek’s mark is
protected under the Tea Rose–Rectanus doctrine and argues
that we may affirm the district court’s judgment of no
liability on this alternative basis.” is replaced with “The Tea
Rose–Rectanus doctrine is an affirmative defense separate
and apart from the underlying infringement claim.
5 McCarthy, supra, § 26:4. Omnia asserts that its use of
Stone Creek’s mark is protected under that doctrine and
argues that we may affirm the district court’s judgment of no
liability on this alternative basis.”

   With this amendment, Judges McKeown and Callahan
vote to deny the petition for rehearing en banc, and Judge
Quist so recommends.
          STONE CREEK V. OMNIA ITALIAN DESIGN                5

   The full court has been advised of the petition for
rehearing en banc, and no active judge has requested a vote
on whether to rehear the matter en banc. Fed. R. App. P. 35.

    The petition for rehearing en banc is denied. No further
petitions for panel or en banc rehearing shall be permitted.



                         OPINION

McKEOWN, Circuit Judge:

    This appeal, set in the high-stakes world of furniture
sales, runs the gamut of trademark infringement issues. The
facts are somewhat unusual: the alleged infringer, leather
furniture manufacturer Omnia Italian Design, Inc.
(“Omnia”), admits that it blatantly copied and began selling
the same goods branded with the mark of its (now ex)
business partner, retail furniture company Stone Creek, Inc.
(“Stone Creek”).

    The first question we confront is whether Omnia’s use of
Stone Creek’s mark was likely to cause confusion. Placing
an identical mark on identical goods creates a strong
likelihood of confusion, especially when the mark is
fanciful. Because Stone Creek also sells in overlapping
marketing channels and the law dictates that other factors
heighten the likelihood that consumers will be confused as
to the origin of the furniture, we reverse the district court’s
contrary determination.

    We also reject Omnia’s invocation of a common-law
defense—known as the Tea Rose–Rectanus doctrine—that
protects use of a mark in a remote geographic area when the
use is in good faith. Omnia’s knowledge of Stone Creek’s
6        STONE CREEK V. OMNIA ITALIAN DESIGN

prior use defeats any claim of good faith. Finally, we
confirm that a 1999 amendment to the trademark statutes
does not sweep away our precedent requiring that a plaintiff
prove willfulness to justify an award of the defendant’s
profits. A remand is necessary to determine whether Stone
Creek can make that showing here.

                       Background

    Stone Creek, which manufactures furniture and sells
directly to customers, has five showrooms in the Phoenix,
Arizona area. Around 1990, Stone Creek adopted and began
using the STONE CREEK mark:




    In 1992, Stone Creek obtained state trademark
protection. Twenty years later, in 2012, Stone Creek
federally registered its mark. As described in the federal
registration, the STONE CREEK mark is a red oval circling
the words “Stone Creek” for various types of furniture.

    In 2003, Stone Creek met representatives of Omnia—a
manufacturer of leather furniture—at a California trade
show. Dazzled by Omnia’s pitch, Stone Creek agreed to buy
Omnia’s furniture. The two companies entered into an
agreement under which Omnia manufactured leather
furniture branded with the STONE CREEK mark. The
business relationship stayed strong through 2012, but in
2013, Stone Creek discovered that Omnia had been using the
STONE CREEK mark on competing furniture.
           STONE CREEK V. OMNIA ITALIAN DESIGN                        7

    Omnia’s unauthorized use began in 2008 when Omnia
was trying to woo a big client. For many years before that,
Omnia had worked with retailer Bon-Ton Stores, Inc. (“Bon-
Ton”), but Bon-Ton became a “significant” customer in
2008. Bon-Ton signed on for Omnia to supply Bon-Ton’s
leather furniture. However, Bon-Ton did not want to sell
under the Omnia name; instead, Bon-Ton preferred a label
that sounded “American.” Although Omnia offered multiple
options, Bon-Ton opted for STONE CREEK. According to
Omnia, part of the allure of selecting the STONE CREEK
mark was that marketing materials and a logo were already
prepared.

    Omnia copied the logo directly from Stone Creek’s
materials. Omnia’s team used old documents with Stone
Creek’s logo to digitally recreate the identical logo because
they could not achieve sharp resolution by scanning. Then
Omnia plastered the mark onto a host of items, including
binders, leather samples, and color boards for display in
Bon-Ton stores. Most relevant here, Omnia designed
warranty cards with the STONE CREEK mark, and, from
2008 to 2013, sold leather furniture to Bon-Ton branded with
the STONE CREEK mark. Near the end of this period,
Stone Creek—still unaware of Omnia’s misdoings—
obtained its federal trademark registration.

     The STONE CREEK–labeled furniture produced by
Omnia was shipped to Bon-Ton and sold to customers at
Bon-Ton’s various furniture galleries in the Midwest. 1
Omnia’s products reached purchasers living within
200 miles of a Bon-Ton gallery, including portions of
Illinois, Indiana, Iowa, Michigan, Ohio, Pennsylvania, and

    1
      We refer to the Midwest for ease of reference. The relevant states
covered are Illinois, Michigan, Ohio, Pennsylvania, and Wisconsin.
8          STONE CREEK V. OMNIA ITALIAN DESIGN

Wisconsin. The claimed infringing sales all occurred within
that area.

    Stone Creek caught a whiff of what Omnia was doing by
2013. Multiple customers contacted Stone Creek to ask
about product options and store locations in the Midwest.
Another customer called about a warranty issue with a
leather sofa purchased at a Bon-Ton store in Chicago. The
warranty card contained the STONE CREEK mark, which
led the customer to Stone Creek’s website. At that point,
Stone Creek’s president discovered that the mark was being
used on furniture sold on Bon-Ton’s website, and Stone
Creek asked Omnia if it was selling products with the
STONE CREEK mark to other companies.

    To its credit, Omnia was candid. In an email from the
Vice President of Sales, Omnia unequivocally admitted to
selling furniture under the STONE CREEK mark. In a move
not recommended when litigation is certainly impending, the
email observed: “In this day of internet shopping and
surfing, it is unfortunate and probably a nuisance for you that
your stores are receiving inquiries regarding these products
due to the similar name.”

    Stone Creek filed suit in the District of Arizona, alleging
federal and common-law trademark infringement and unfair
competition. After a bench trial, the district court held that
Omnia did not infringe Stone Creek’s trademark because
there was no likelihood of confusion. 2 The court also ruled
at summary judgment that disgorgement of Omnia’s profits
required a showing of willful infringement, but that
determination was never made because Omnia was not

    2
      Earlier in the case, the district court rejected Omnia’s Tea Rose–
Rectanus defense based on Omnia’s lack of good faith.
          STONE CREEK V. OMNIA ITALIAN DESIGN                9

found to infringe. During the course of the case, the court
sanctioned Stone Creek’s attorneys on two occasions for
filings related to Omnia’s profits and damages.

                          Analysis

I. Likelihood of Confusion

     Under the Lanham Act, infringement lies for both
registered and unregistered trademarks when the alleged
infringer’s use “is likely to cause confusion, or to cause
mistake, or to deceive.”           15 U.S.C. §§ 1114(1)(a),
1125(a)(1)(A). The touchstone for trademark infringement
is likelihood of confusion, which asks whether a “reasonably
prudent” marketplace consumer is “likely to be confused as
to the origin of the good or service bearing one of the marks.”
Rearden LLC v. Rearden Commerce, Inc., 

,
1214 (9th Cir. 2012) (citation omitted). Although we review
the district court’s findings and determination of no
likelihood of confusion for clear error, we address legal error
de novo. See Gilman v. Brown, 

, 1017 (9th
Cir. 2016) (“Because the district court applied the wrong
standard, it committed legal error, and the resulting factual
findings are clearly erroneous.”), cert. denied sub nom.
Madden v. Brown, 

(2017); Reno Air Racing
Ass’n, Inc. v. McCord, 

, 1135 (9th Cir. 2006);
see also Dorr-Oliver, Inc. v. Fluid-Quip, Inc., 

,
380 (7th Cir. 1996) (noting that courts “review the district
court’s statement of the law de novo for legal error” in
likelihood of confusion cases).

    The district court properly cited the well-established
factors (the Sleekcraft factors) that guide the likelihood of
confusion inquiry. See AMF Inc. v. Sleekcraft Boats,

, 348–49 (9th Cir. 1979), abrogated on other
grounds by Mattel, Inc. v. Walking Mountain Prods.,
10        STONE CREEK V. OMNIA ITALIAN DESIGN

(9th Cir. 2003). In conducting the analysis,
courts do not merely count beans or tally points. See
Dreamwerks Prod. Grp., Inc. v. SKG Studio, 

,
1129 (9th Cir. 1998). Not all factors are created equal, and
their relative weight varies based on the context of a
particular case. See Network Automation, Inc. v. Advanced
Sys. Concepts, Inc., 

, 1145 (9th Cir. 2011).

    Two particularly probative factors are the similarity of
the marks and the proximity of the goods. See Lindy Pen
Co. v. Bic Pen Corp., 

, 256–57 (9th Cir. 1986).
Other potentially relevant factors include the strength of the
protected mark, evidence of actual confusion, the use of a
common marketing channel, the defendant’s intent in
selecting the allegedly infringing mark, the type of goods
and the degree of consumer care, and the likelihood of
product expansion. 

–49. After
examining the district court’s consideration of the factors in
this case, we conclude that the court legally erred in framing
many of the Sleekcraft factors, leading us to reverse the
finding of no likelihood of confusion. The indistinguishable
marks and goods, coupled with a fanciful mark, evidence of
actual confusion, convergent marketing channels, and
blatant copying, tell the real story.

     A. Similarity of the Marks and Proximity of the Goods

    The parties do not dispute—nor could they—the district
court’s finding that the factors examining similarity of the
marks and the proximity of the goods favor Stone Creek.
But simply acknowledging the identity of the marks and
goods understates the importance of these factors and our
precedent’s conclusion that identical marks paired with
identical goods can be case-dispositive: “In light of the
virtual identity of marks, if they were used with identical
products or services likelihood of confusion would follow as
         STONE CREEK V. OMNIA ITALIAN DESIGN               11

a matter of course.” Brookfield Commc’ns, Inc. v. W. Coast
Entm’t Corp., 

, 1056 (9th Cir. 1999);
Opticians Ass’n of Am. v. Indep. Opticians of Am., 

, 195 (3d Cir. 1990) (“[L]ikelihood of confusion is
inevitable, when, as in this case, the identical mark is used
concurrently by unrelated entities.”). The case for confusion
based on identical marks and identical goods could hardly be
stronger than here.

    Omnia’s mark is an exact replica of Stone Creek’s logo
that Omnia copied from materials given to Omnia by Stone
Creek. And the goods on which the marks appeared are
identical: not only do Stone Creek and Bon-Ton sell leather
furniture, but they both sell leather furniture that is
manufactured by Omnia. As McCarthy, a leading trademark
scholar, has observed, cases involving identical marks on
competitive goods are rare and “hardly ever find their way
into the appellate reports” because liability is “open and
shut.” 4 J. Thomas McCarthy, McCarthy on Trademarks and
Unfair Competition § 23:20 (4th ed. 2017); Wynn Oil Co. v.
Thomas, 

, 1190–91 (6th Cir. 1988). We
examine the remaining factors to determine whether they
displace the powerful case of likelihood of confusion.

   B. Strength of the Protected Mark

    The strength of the mark is a key factor with two
components: the mark’s recognition in the market (i.e., its
commercial strength) and the mark’s inherent
distinctiveness (i.e., its conceptual strength). See Lahoti v.
Vericheck, Inc., 

, 508 (9th Cir. 2011). The
district court analyzed only half of the equation when it said
that “[t]he STONE CREEK mark is strong in Arizona, but it
is not recognized in the [Midwest] for its relationship to
12         STONE CREEK V. OMNIA ITALIAN DESIGN

Stone Creek.” 3 As we explain below, this crabbed view of
commercial strength is misplaced. But the court’s omission
of the inherent distinctiveness of the mark is even more
significant because, on the spectrum of conceptual strength,
Stone Creek’s mark falls at the high end as a fanciful or
arbitrary mark. See Entrepreneur Media, Inc. v. Smith,

, 1141 (9th Cir. 2002). Its mark is not a
suggestive or descriptive mark that is less likely to act as a
unique identifier of the source in a consumer’s mind. 

The district court committed legal error in not
adducing and evaluating this aspect of the strength factor.
See 

(describing the importance of the strength
factor in determining the scope of trademark protection).

     C. Evidence of Actual Confusion

     Given the other factors favoring likelihood of confusion,
it is not surprising that Stone Creek has put forward several
instances of actual confusion. Such evidence is not
necessary for a finding of likelihood of confusion, but it
bears on the inquiry and is particularly potent. See

–53.           The district court
disregarded Stone Creek’s evidence because the court
focused on whether there was “actual confusion by any
consumer in the [Midwest] who purchased Omnia furniture
believing it was manufactured or sold by Stone Creek.” This
approach misapprehends the breadth of likelihood of
confusion, which can exist even when consumers are not



     3
       To reach its conclusion, the court relied on survey evidence
indicating limited awareness of the Stone Creek brand in the Midwest.
Because we ultimately conclude that Stone Creek has established a
likelihood of confusion, we need not reach Stone Creek’s challenges to
the admission of Omnia’s expert and surveys.
         STONE CREEK V. OMNIA ITALIAN DESIGN               13

“confused as to the source of a product they actually
purchase.” Brookfield 

.

    When we widen the lens to embrace the full scope of
qualifying actual confusion evidence, we credit the examples
of customers seeking to purchase furniture or having already
purchased Bon-Ton furniture who misdirected their queries
to Stone Creek. Cf. Nutri/Sys., Inc. v. Con-Stan Indus., Inc.,

, 606 (9th Cir. 1987). These consumers did not
ask whether Stone Creek and Bon-Ton are affiliated; they
were actually confused as to the source. See Cohn v.
Petsmart, Inc., 

, 842 n.7 (9th Cir. 2002) (per
curiam). The customers requested information about
product options and store locations—issues likely to affect
their buying habits and their view of the company associated
with the mark. See Rearden 

(characterizing trademark infringement as concerned with
consumers’ purchasing decisions). They also contacted
Stone Creek about warranty problems with Omnia furniture
from Bon-Ton. These occasions of actual confusion cannot
be dismissed out of hand but must be considered in context
and in light of the other evidence of likelihood of confusion.
See 4 McCarthy, supra, § 23:14; see also Streamline Prod.
Sys., Inc. v. Streamline Mfg., Inc., 

, 457 (5th
Cir. 2017) (“[C]ourts may not ignore competent evidence of
actual confusion.” (citation omitted)); Daddy’s Junky Music
Stores, Inc. v. Big Daddy’s Family Music Ctr., 

,
284 (6th Cir. 1997) (explaining that one instance of actual
confusion “favors plaintiff at least to some extent”).

   D. Convergence of Marketing Channels

    The district court’s consideration of overlapping
marketing channels rested on the faulty legal assumption that
geographic separation automatically means no intersection
in marketing channels. This misconception led the court to
14       STONE CREEK V. OMNIA ITALIAN DESIGN

unduly weigh this factor against Stone Creek. Even when
parties “operate in different geographical areas,” they may
still “provide [goods] in convergent marketing channels.”
Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 

,
1509 (9th Cir. 1986). This principle is highlighted here
because Stone Creek uses its website as a substantial channel
to market its furniture beyond its physical stores in Arizona.

    The parties tacitly agree that the analysis of marketing
channels should look to Bon-Ton’s sales to customers rather
than Omnia’s sales to Bon-Ton. Thus, we focus on Stone
Creek and Bon-Ton. Both are retail furniture stores, and
their products are identical pieces of leather furniture
manufactured by Omnia.          Selling similar, let alone
interchangeable, products suggests that Stone Creek and
Bon-Ton share the same general class of customers. See
Pom Wonderful LLC v. Hubbard, 

, 1130 (9th
Cir. 2014).

    That Stone Creek does most of its business in the
Phoenix area does not foreclose overlapping marketing
channels with Omnia. The district court found that Stone
Creek      placed     its    mark     on      its  website,
stonecreekfurniture.com, as early as 2000 and optimized
search engine searches so that its website appears early in
Internet search results. During the period of alleged
infringement, Stone Creek also advertised in a nationwide
magazine with readership in the Midwest. Since its
inception, Stone Creek has made more than $200,000,000 in
sales; $610,384 of those sales occurred in the Midwest.

    The district court’s factual findings confirm that Stone
Creek’s print and online advertising reached the Midwest.
Stone Creek’s marketing traction in the Midwest is
buttressed by the finding that Stone Creek had customers and
made sales in that region at the same time that Bon-Ton was
          STONE CREEK V. OMNIA ITALIAN DESIGN                15

selling the STONE CREEK–labeled furniture sourced from
Omnia. To be sure, those sales represent a small fraction of
Stone Creek’s total sales, but, in light of the particular facts
of this case, the small volume of the overall sales does not
undercut Stone Creek’s distribution of furniture in the
Midwest.

    What appears to have led the district court astray in
analyzing the marketing channels factor (and permeates
much of the court’s discussion of other factors) is a myopic
focus on the considerable distance between Stone Creek’s
physical showrooms in Arizona and Bon-Ton’s in the
Midwest. As the court explained, Stone Creek follows the
retail furniture business model, where sales are generally
limited to customers living within a drivable distance from
the brick-and-mortar stores. But the territorial separation is
not so well-delineated or physically defined: Stone Creek
and Bon-Ton were simultaneously advertising and selling
under the STONE CREEK mark in the Midwest. Thus, the
likelihood of confusion is not so diminished that Stone
Creek’s nationwide right to enforce its mark in the alleged
infringer’s area has not yet ripened into a remedy. See
Mister Donut of Am., Inc. v. Mr. Donut, Inc., 

,
844 (9th Cir. 1969) (finding that likelihood of confusion was
not precluded because the parties were “no longer confined
to separate and distinct market areas”).

   E. Intent in Selecting the Allegedly Infringing Mark

    Omnia’s reason for adopting the STONE CREEK mark
also plays a critical role: when the alleged infringer intended
to deceive customers, we infer that its conscious attempt to
confuse did in fact result in confusion. Playboy Enters., Inc.
v. Netscape Commc’ns Corp., 

, 1028 (9th Cir.
2004). Recognizing the difficulty of collecting evidence of
a party’s motive, we have held that choosing a designation
16         STONE CREEK V. OMNIA ITALIAN DESIGN

with knowledge that it is another’s trademark permits a
presumption of intent to deceive. Hokto Kinoko Co. v.
Concord Farms, Inc., 

, 1096 (9th Cir. 2013).

    The district court found that “Omnia adopted and used
the STONE CREEK mark with full knowledge of Stone
Creek’s senior use,” an explicit finding establishing the
factual predicate to apply the presumption. 4 Yet the court
nowhere assesses or acknowledges the presumption, and that
error colored its analysis and conclusion. Omnia’s bare
assertion that the mark was picked for its “American” sound
does not counteract the intent to deceive, especially when
Omnia had endless options for suitably “American”-
sounding names to offer to Bon-Ton. This rationale is no
more convincing than if a contemporary marketeer decided
to appropriate the long-standing Häagen-Dazs mark
believing that a foreign-sounding name would appeal to
customers. 5 And Omnia’s replication of the mark from
Stone Creek’s materials and unauthorized use on identical
furniture contribute to an intentional appropriation. See Au-


     4
       We see no reason to modify the applicability of the presumption
simply because Stone Creek had not yet federally registered its mark at
the time that Omnia adopted the mark. Omnia’s understanding of the
scope of Stone Creek’s rights is relevant to rebutting the presumption,
but it does not undercut the rationale for applying the presumption in the
first place.
     5
       The irony, of course, is that the creators, who started the business
in New York, invented the fanciful mark because they thought there
would be cache in the foreignness of the mark. As it turned out, they
were prescient about market identification because customers thought
the ice cream came from Scandinavia. See Alison Spiegel, Häagen-Dazs
Doesn’t Come From Where You Think It Comes From, HuffPost (May
13, 2015, 7:00 AM), http://www.huffingtonpost.com/2015/05/13/haage
n-dazs-comes-from_n_7266208.html.
          STONE CREEK V. OMNIA ITALIAN DESIGN                17

Tomotive Gold, Inc. v. Volkswagen of Am., Inc., 

, 1076 (9th Cir. 2006).

    Additionally, conjectural statements by Omnia’s
president about the scope of Stone Creek’s business do not
amount to a “good faith belief that there [would be] no
conflict between the marks as used on [Omnia’s] goods.”
4 McCarthy, supra, § 23:115. Although the president
“understood” that Stone Creek sold in Phoenix, he never
researched where or how the furniture was sold. Nor did
Omnia ask or investigate where Stone Creek’s customers
were located before using the mark. Omnia has not unseated
the presumption, and its deceptive intent is “entitled to great
weight” in the ultimate determination of likelihood of
confusion. Kendall-Jackson Winery, Ltd. v. E. & J. Gallo
Winery, 

, 1048 (9th Cir. 1998) (citation
omitted).

   F. Degree of Consumer Care and Likelihood of Product
      Expansion

    The remaining two factors—the degree of consumer care
based on the type of goods and the likelihood of product
expansion—do not support either party; at best, they weakly
support Omnia. As the district court found, since furniture
is an expensive good, likelihood of confusion is diminished
because we justifiably expect that consumers will make
purchases with more research and closer scrutiny. Multi
Time Mach., Inc. v. Amazon.com, Inc., 

, 937
(9th Cir. 2015), cert. denied, 

(2016). But
this factor is less instructive in cases like this one where the
marks and goods are identical for the simple reason that even
the trained eye will not be able to discern any difference.
McGregor-Doniger Inc. v. Drizzle Inc., 

, 1137
(2d Cir. 1979), superseded on other grounds as recognized
by Cross Commerce Media, Inc. v. Collective, Inc., 

       STONE CREEK V. OMNIA ITALIAN DESIGN

155 (2d Cir. 2016); 4 McCarthy, supra, § 23.96. Similarly,
while the district court did not clearly err in finding that
Stone Creek has not demonstrated non-speculative plans to
expand, that fact has minimal legal significance because
Stone Creek established an overlap in goods and marketing
channels. See Playboy 

.

     G. Conclusion

    We reverse the district court’s finding of no likelihood
of confusion because it is based on faulty legal foundations.
We credit the court’s factual findings, but its circumscribed
view of the legal landscape left the court with an incomplete
picture. The slam-dunk evidence of a conceptually strong
mark together with the use of identical marks on identical
goods is difficult to surmount. Viewing the facts through the
correct legal lens, there is no substantial argument that the
other factors and evidence overcome the robust case that
Omnia’s use of the STONE CREEK mark is likely to cause
confusion.

II. The Tea Rose–Rectanus Doctrine

    Our determination of a likelihood of confusion with
respect to the STONE CREEK mark does not end the
infringement analysis. The Tea Rose–Rectanus doctrine is
an affirmative defense separate and apart from the
underlying infringement claim. 5 McCarthy, supra, § 26:4.
Omnia asserts that its use of Stone Creek’s mark is protected
under that doctrine and argues that we may affirm the district
court’s judgment of no liability on this alternative basis. The
district court rejected this defense, and so do we.

   The Tea Rose–Rectanus doctrine has its roots in the
common law: it is named for a pair of Supreme Court cases,
Hanover Star Milling Co. v. Metcalf, 

(1916)
         STONE CREEK V. OMNIA ITALIAN DESIGN               19

(“Tea Rose”), and United Drug Co. v. Theodore Rectanus
Co., 

(1918). The central proposition underlying
the two cases is that common-law trademark rights extend
only to the territory where a mark is known and recognized,
so a later user may sometimes acquire rights in pockets
geographically remote from the first user’s territory. The
question we address is whether Omnia acquired common-
law rights in the Midwest under the Tea Rose–Rectanus
doctrine.

    Omnia’s common-law rights, if they exist, are not wiped
out merely because Stone Creek later filed a federal
registration. Although federal registration presumptively
entitles the senior user to nationwide protection, 15 U.S.C.
§ 1057(b), the Lanham Act preserves legal and equitable
defenses that could have been asserted prior to registration,

Under this rule, already-established common-
law rights are carved out of the registrant’s scope of
protection. 

Johnny Blastoff, Inc. v. L.A.
Rams Football Co., 

, 435 (7th Cir. 1999). In
other words, the geographic scope of a senior user’s rights in
a registered trademark looks like Swiss cheese: it stretches
throughout the United States with holes cut out where others
acquired common-law rights prior to the registration.
5 McCarthy, supra, § 26:31. Because Omnia began using
the mark in 2008, well before Stone Creek’s federal
registration in 2012, the Tea Rose–Rectanus defense is
available to Omnia if it is applicable.

    To take advantage of the Tea Rose–Rectanus doctrine,
the junior user must establish good faith use in a
geographically remote area. See 

;
cf. Grupo Gigante SA De CV v. Dallo & Co., 

,
1096 & n.26 (9th Cir. 2004). Like the district court, we limit
20        STONE CREEK V. OMNIA ITALIAN DESIGN

our discussion to the question of good faith because it is
dispositive.

     The varying descriptions of good faith in the leading
Supreme Court cases have spawned a circuit split, and our
circuit has not yet weighed in. See Grupo 

n.26. On one side, some circuits have held that the
junior user’s knowledge of the senior user’s prior use of the
mark destroys good faith. See, e.g., Nat’l Ass’n for
Healthcare Commc’ns, Inc. v. Cent. Ark. Area Agency on
Aging, Inc., 

, 735 (8th Cir. 2001); Money Store
v. Harriscorp Fin., Inc., 

, 674–75 (7th Cir.
1982). In contrast, other circuits have held that knowledge
is a factor informing good faith, but the “focus is on whether
the [junior] user had the intent to benefit from the reputation
or goodwill of the [senior] user.” GTE Corp. v. Williams,

, 541 (10th Cir. 1990); see C.P. Interests, Inc.
v. Cal. Pools, Inc., 

, 700 (5th Cir. 2001). We
conclude that the better view is that there is no good faith if
the junior user had knowledge of the senior user’s prior use.

    Looking back to the origins of the Tea Rose–Rectanus
doctrine informs why knowledge defeats a claim of good
faith use. In Tea Rose, the senior user began selling “Tea
Rose” flour in approximately 1872; many years later, the
junior user began selling “Tea Rose” flour without any
knowledge of the senior user’s prior 

–
08. At the time that the trademark infringement action was
filed, the senior user had made sales in Massachusetts, Ohio,
and Pennsylvania, while the junior user’s sales had reached
Mississippi, Alabama, Georgia, and Florida. 

Rectanus arose on similar facts: the senior user began selling
“Rex” drugs around 1877 and operated in New England,
while the junior user began selling “Rex” drugs around 1883
and operated in Kentucky, with neither party being aware of
         STONE CREEK V. OMNIA ITALIAN DESIGN               21

the other’s use of the “Rex” mark for more than twenty

–96. In both cases, the Supreme Court
held that the senior user could not enjoin the junior user’s
use of the same mark because the junior user adopted the
mark in good faith and had developed a local reputation in
an area where the mark was not recognized as designating
the senior user. See 

Tea 

–16.

    When describing good faith, the Supreme Court
emphasized that the junior user had no awareness of the
senior user’s use of the mark. The Court in Tea Rose states
that the junior user “adopted and used [the trademark] in
good faith without knowledge or notice that the name ‘Tea
Rose’ had been adopted or used . . . by anybody 

. The Court also refers to the situation as one
where the two parties “independently” employ the same
mark. 

And the Court’s reasoning concentrates on
knowledge:

       Under the circumstances that are here
       presented, to permit the [senior user] to use
       the mark in Alabama, to the exclusion of the
       [junior user], would take the trade and good
       will of the latter company—built up at much
       expense and without notice of the former’s
       rights—and confer it upon the former, to the
       complete perversion of the proper theory of
       trademark rights.

(emphasis added).

    The same focus on notice emerges in Rectanus, which
grants protection for an “innocent” junior user who has “hit
upon” the same mark and avers that the parties acted “in
perfect good faith; neither side having any knowledge or
22       STONE CREEK V. OMNIA ITALIAN DESIGN

notice of what was being done by the 

,
103. The Court also relies on a case that says that the
defendants there acted in good faith because they “believ[ed]
[their] use to be original with them.” Richter v. Anchor
Remedy Co, 

, 455 (C.C.W.D. Pa. 1892), aff’d sub
nom. Richter v. Reynolds, 

(3d Cir. 1893). Seventy
years later, Justice Brennan stressed that application of the
Tea Rose–Rectanus doctrine requires an absence of
knowledge. See K Mart Corp. v. Cartier, Inc., 

,
314 n.8 (1988) (Brennan, J., concurring in part and
dissenting in part) (“[A] firm can develop a trademark that is
identical to a trademark already in use in a geographically
distinct and remote area if the firm is unaware of the
identity.”).

    The Seventh and Eighth Circuits and the Trademark
Trial and Appeal Board (“TTAB”) agree with this reading.
The Seventh Circuit put it explicitly: “A good faith junior
user is one who begins using a mark with no knowledge that
someone else is already using it.” Money 

. The court went on to analyze whether the junior user
in that case had constructive or actual knowledge of the
senior user’s use. 

The Eighth Circuit follows the
same approach, parroting the language from Tea Rose and
Rectanus. See Nat’l Ass’n for Healthcare 

(“adopted the [mark] in good faith, without
knowledge of [the] prior use”). And the TTAB, the
administrative board charged with deciding certain
trademark disputes and appeals, similarly holds that
“appropriat[ing] a mark with knowledge that it is actually
being used by another” means “that use is not believed to be
a good faith use.” Woman’s World Shops Inc. v. Lane Bryant
Inc., 5 U.S.P.Q.2d 1985, 1988 (T.T.A.B. 1988).
          STONE CREEK V. OMNIA ITALIAN DESIGN                23

    The courts that have ruled the other way have latched on
to one line in the Tea Rose case which reads:

       [W]here two parties independently are
       employing the same mark upon goods of the
       same class, but in separate markets wholly
       remote the one from the other, the question of
       prior appropriation is legally insignificant;
       unless, at least, it appear that the second
       adopter has selected the mark with some
       design inimical to the interests of the [senior]
       user, such as to take the benefit of the
       reputation of his goods, to forestall the
       extension of his trade, or the 

(emphasis added). But this brief reference
to “design inimical” does not override the central focus on
knowledge; it is not without significance that “design
inimical” does not appear anywhere else in the opinion. The
Court in Rectanus repeats the “design inimical” language as
a direct quote of the language from the Tea Rose case and
mentions offhand that the junior user did not have a “sinister

. More salient are the various
points in the leading opinions that draw a close connection
between “good faith” and “knowledge” or “notice.” See,
e.g., 

(“in perfect good faith; neither side having any
knowledge or notice of what was being done by the other”);

(“in good faith, and without notice of any prior use
by others, selected and used the ‘Rex’ mark”); Tea 

(“trademark was adopted and used [by the
junior user] in good faith without knowledge or notice that
the name ‘Tea Rose’ had been adopted or used by the [senior
user]”); 

(“in good faith and without notice of the
[senior user’s] mark”).
24        STONE CREEK V. OMNIA ITALIAN DESIGN

    Tying good faith to knowledge makes sense in light of
the policy underlying the doctrinal framework. As the
Supreme Court explained, the Tea Rose–Rectanus doctrine
operates to protect a junior user who unwittingly adopted the
same mark and invested time and resources into building a
business with that mark. 

; Tea

. A junior user like Omnia who has
affirmative knowledge of the senior user’s mark has not
serendipitously chosen the same mark and independently
built up its own brand. Instead, a user like Omnia knows that
its actions come directly at the expense of the senior user,
potentially blocking the senior user from entering into the
new market. Viewed in this light, the junior user has acted
in bad faith, which “serve[s] as evidence that the [senior]
user’s mark, at least in reputation, has extended to the new
area.” Developments in the Law Trade-Marks and Unfair
Competition, 68 Harv. L. Rev. 814, 859 (1955); 5 McCarthy,
supra, § 26:12.

    The knowledge standard also better comports with the
Lanham Act. The statutory section preserving the Tea Rose–
Rectanus defense for junior users acting pre-registration
requires that the junior user’s mark “was adopted without
knowledge of the registrant’s prior use.” 15 U.S.C.
§ 1115(b)(5) (emphasis added). More broadly, one major
change effected by the Lanham Act is that securing federal
registration affords nationwide rights regardless of where the
registrant has used the mark, a result accomplished by a
provision that puts would-be users on constructive notice.
See 

1072; 5 McCarthy, supra, § 26:32. In
other words, the Lanham Act displaces the Tea Rose–
Rectanus defense by charging later users with knowledge of
a mark listed on the federal register. If constructive notice is
sufficient to defeat good faith, it follows that actual notice
should be enough too.
          STONE CREEK V. OMNIA ITALIAN DESIGN               25

    Once knowledge is accepted as a determinative factor in
deciding good faith, the Tea Rose–Rectanus doctrine has no
applicability here. The district court found that “[Omnia]
was a non-innocent remote user” who “acquired no common
law trademark rights in the [Midwest].” That conclusion
flows from the parties’ agreement that Omnia adopted Stone
Creek’s mark with knowledge of Stone Creek’s previous
use. The Tea Rose–Rectanus doctrine provides no shelter to
Omnia for infringement of Stone Creek’s mark.

III.   Willfulness and Disgorgement of Profits

     Under the remedies provision of the Lanham Act, a court
may award (1) the defendant’s profits, (2) the damages
sustained by the plaintiff, and (3) the costs of the action.
15 U.S.C. § 1117(a). At issue here is the applicable standard
to award disgorgement of profits. In an effort to shape its
trial strategy, Stone Creek filed a motion for summary
judgment asking the district court to rule that willfulness is
not required for such an award. The court denied the motion
and sanctioned Stone Creek for filing it.

    Because we conclude that Omnia infringed Stone
Creek’s mark, we address the standard for awarding
disgorgement of profits as it will be front and center on
remand. Historically, we have imposed a willfulness
requirement with respect to disgorgement of profits. Now
we must decide what effect, if any, a 1999 amendment to the
Lanham Act’s remedies provision has on our precedent
regarding awarding of profits. We agree with the district
court that the 1999 amendment has not changed the state of
the law on disgorgement and that willfulness is still required.

   The evolution of the remedies provision—including the
ever-persisting circuit split—is key to understanding the
impact of the 1999 amendment, so we recount the history
26       STONE CREEK V. OMNIA ITALIAN DESIGN

and the current state of affairs. Before 1999, the trademark
remedies provision—§ 1117(a)—stated:

       When a violation of any right of the registrant
       of a mark registered in the Patent and
       Trademark Office, or a violation under
       section 1125(a) of this title, shall have been
       established in any civil action arising under
       this chapter, the plaintiff shall be entitled,
       subject to the provisions of sections 1111 and
       1114 of this title, and subject to the principles
       of equity, to recover (1) defendant’s profits,
       (2) any damages sustained by the plaintiff,
       and (3) the costs of the action.

15 U.S.C. § 1117(a) (1996). The spirited debate among the
circuits has been reserved for how the phrase “subject to the
principles of equity” applies to an award of the defendant’s
profits.

    Our circuit fell in line with the camp that requires a
showing of willfulness. We held that an award of the
defendant’s profits “is not automatic and must be granted in
light of equitable considerations”; equity dictates that the
plaintiff must show that the defendant’s infringing acts were
accompanied by some form of intent. Lindy Pen Co. v. Bic
Pen Corp., 

, 1405–06 (9th Cir. 1993),
abrogated on other grounds by SunEarth, Inc. v. Sun Earth
Solar Power Co., 

(9th Cir. 2016) (per
curiam). At that time, the Third Circuit was in accord that
“a plaintiff must prove that an infringer acted willfully
before the infringer’s profits are recoverable.” SecuraComm
Consulting Inc. v. Securacom Inc., 

, 190 (3d
Cir. 1999), superseded by statute as stated in Banjo Buddies,
Inc. v. Renosky, 

(3d Cir. 2005). Other circuits
         STONE CREEK V. OMNIA ITALIAN DESIGN               27

agreed. See, e.g., George Basch Co. v. Blue Coral, Inc.,

, 1540 (2d Cir. 1992); ALPO Petfoods, Inc. v.
Ralston Purina Co., 

, 968 (D.C. Cir. 1990). In
the other camp were circuits who viewed willfulness as one
factor in the overall determination of whether an award of
profits is appropriate. See, e.g., Pebble Beach Co. v. Tour
18 I Ltd., 

, 554–55 (5th Cir. 1998), abrogated
on other grounds by TrafFix Devices, Inc. v. Marketing
Displays, Inc., 

(2001); Roulo v. Russ Berrie &
Co., 

, 941 (7th Cir. 1989). These decisions all
predate the 1999 amendment.

    The amendment became necessary after Congress made
a substantive change to the Lanham Act in 1996. That year,
Congress added § 1125(c), which created a federal cause of
action for trademark dilution. See H.R. Rep. No. 104-374,
at 2 (1995), as reprinted in 1996 U.S.C.C.A.N. 1029, 1029.
The provision expressly allowed holders of famous
trademarks to enjoin uses that diluted the distinctive quality
of their marks. See 15 U.S.C. § 1125(c) (1996). Section
1125(c) also purported to provide monetary relief under the
remedies provision, § 1117(a), when dilution was “willfully
intended.” But Congress failed to make the requisite cross-
reference in § 1117(a) to harmonize that section with the
amendment and soon discovered the missing link between
the two statutory provisions.

    That statutory mismatch spurred the 1999 amendment.
Congress revised the remedies section, § 1117(a), to include
reference to a “willful violation under section 1125(c).” The
amendment thus made clear that a plaintiff with a dilution
claim could recover money damages. The current version of
§ 1117(a) reads:

       When a violation of any right of the registrant
       of a mark registered in the Patent and
28       STONE CREEK V. OMNIA ITALIAN DESIGN

       Trademark Office, a violation under section
       1125(a) or (d) of this title, or a willful
       violation under section 1125(c) of this title,
       shall have been established in any civil action
       arising under this chapter, the plaintiff shall
       be entitled, subject to the provisions of
       sections 1111 and 1114 of this title, and
       subject to the principles of equity, to recover
       (1) defendant’s profits, (2) any damages
       sustained by the plaintiff, and (3) the costs of
       the action.

15 U.S.C. § 1117(a) (emphasis added). To put it in plain
English, if there is

       [1] a violation of the rights in a registered
       mark, an unregistered mark (§ 1125(a)), or a
       mark used as a domain name (§ 1125(d)), or

       [2] a willful violation under the dilution
       statute (§ 1125(c)),

then, “subject to the principles of equity,” the plaintiff is
entitled to relief. Critically, Congress did not modify the
“subject to the principles of equity” language.

    The contrast in language between clause [1], which does
not reference willfulness, and newly inserted clause [2],
which does, has caused ripples through the circuit courts,
which remain divided on the role of willfulness in awarding
profits. The Federal Circuit, interpreting Second Circuit
jurisprudence, held that “nothing in the 1999 amendment . . .
allows us to depart from . . . precedent requiring willfulness
for the recovery of profits in infringement cases.” Romag
Fasteners, Inc. v. Fossil, Inc., 

, 791 (Fed. Cir.
         STONE CREEK V. OMNIA ITALIAN DESIGN               29

2016), cert. granted, judgment vacated, 

(2017), opinion reinstated in relevant part, 

(Fed. Cir. May 3, 2017) (per curiam). Other
circuits have adhered to or adopted the rule that willfulness
is one piece of the puzzle. See Synergistic Int’l, LLC v.
Korman, 

, 175 & n.13 (4th Cir. 2006); Quick
Techs., Inc. v. Sage Grp. PLC, 

, 347–49 (5th
Cir. 2002). The Third Circuit switched sides, concluding
that the 1999 amendment upended its precedent requiring
willfulness. Banjo 

.

    We have not yet addressed this question. See Fifty-Six
Hope Rd. Music, Ltd. v. A.V.E.L.A., Inc., 

,
1073–74 (9th Cir.) (omitting substantive analysis of the
effect of the 1999 amendment but finding sufficient evidence
of willfulness), cert. denied, 

(2015). We now
decide that the 1999 amendment does not change the
foundation of Ninth Circuit precedent—willfulness remains
a prerequisite for awarding a defendant’s profits.

    Our conclusion is consistent with the Federal Circuit’s
analysis. See Romag 

. We agree
with its approach to start with the history of the amendment
and thoroughly examine the context in which the amendment
came to be. See 

Several circuits have ruled
the other way without looking at the backstory of the
remedies provision. See Synergistic 

&
n.13; Banjo 

–75; Quick 

–49. That history is illuminating and reveals
why the 1999 amendment does not upend our prior
interpretation of the remaining language in § 1117(a).

    The history of enactment convincingly shows that the
1999 amendment was intended only to correct a conspicuous
drafting error in the 1996 version of the remedies provision.
The legislative history bolsters the view that Congress’s sole
30         STONE CREEK V. OMNIA ITALIAN DESIGN

purpose was to correct the mistaken omission of willful
violations of the dilution statute, § 1125(c), from the
remedies provision, § 1117(a). H.R. Rep. No. 106-250, at 4
(1999) (“[The amendment] seeks to clarify that . . . Congress
did intend to allow for . . . damages against a defendant found
to have wilfully [sic] intended to engage in commercial
activity that would cause dilution of a famous mark.”). 6

    Equally important is what Congress changed. Congress
created a new predicate—namely, a willful violation of
§ 1025(c)—that permits monetary recovery. But it did not
touch the other language in § 1117(a), which has
consistently provided for an award of defendant’s profits
under the “principles of equity.” Our holding in Lindy
Pen—that a plaintiff can secure the defendant’s profits only
after establishing willfulness—is based entirely on an
interpretation of that unaltered 

–
06.

    Thus, it would be a mistake to draw a negative
implication from the unrelated and later-introduced
language that the amendment somehow negated our circuit’s
well-settled willfulness requirement. See Russello v. United
States, 

, 23 (1983); Antonin Scalia & Bryan A.

     6
       In 1999, Congress also passed legislation directed at preventing
cyberpiracy related to trademarks and domain names. 15 U.S.C.
§ 1125(d). This subsection of the statute—often referred to as the
cybersquatting provision—provides further evidence that Congress did
not intend a wholesale revision of the remedies provision. In particular,
while the remedies provision was amended to list violations of the
cybersquatting provision (§ 1125(d)) and willful violations of the
dilution provision (§ 1125(c)), the former requires “bad faith intent” for
all forms of relief, whereas the latter requires “willfull[] inten[t]” only
for monetary relief. Compare 15 U.S.C. § 1125(d)(1)(A)(i), with 

(c)(5)(B).
         STONE CREEK V. OMNIA ITALIAN DESIGN               31

Garner, Reading Law: The Interpretation of Legal Texts 331
(2012) (“A clear, authoritative judicial holding on the
meaning of a particular provision should not be cast in doubt
and subjected to challenge whenever a related though not
utterly inconsistent provision is adopted in the same statute
. . . .”). As McCarthy has explained, it would be improper to
“leverage[] this statutory change beyond its intended scope
in order to adjust the equities in ordinary infringement
cases.” 5 McCarthy, supra, § 30:62. This conclusion has
added force because we see no indication that the legislature
meant to take sides in the entrenched circuit split on
willfulness. See Jama v. Immigration & Customs Enf’t,

, 349 (2005) (rejecting congressional
ratification where there was no “judicial consensus so broad
and unquestioned that we must presume Congress knew of
and endorsed it”).

    For these reasons, the district court properly ruled that
Stone Creek must show intentional or willful infringement
before disgorgement of Omnia’s profits could be awarded.
Because the court denied summary judgment based on a
triable issue of fact about whether Omnia infringed, the court
had no need to reach the question of willfulness. We note
that many of the factual findings that the court has already
made—including those on Omnia’s intent in selecting and
using the STONE CREEK mark—may be relevant to
willfulness. See 4 McCarthy, supra, § 23:112. However, we
decline Stone Creek’s invitation to rule that Omnia’s
infringement was willful as a matter of law and instead
remand for the district court to make this determination.

IV.    Sanctions Orders

   The district court, relying on 28 U.S.C. § 1927, twice
sanctioned Stone Creek’s attorneys. Section 1927 permits
sanctions against an attorney who “multiplies the
32        STONE CREEK V. OMNIA ITALIAN DESIGN

proceedings in any case unreasonably and vexatiously” and
tailors the amount awarded to the costs and fees “reasonably
incurred because of such conduct.” Without more, reckless,
but nonfrivolous, filings may not be sanctioned. B.K.B. v.
Maui Police Dep’t, 

, 1107 (9th Cir. 2002).
The district court’s first sanctions order runs afoul of that
rule, but the second order falls well within the court’s
discretion. See Pac. Harbor Capital, Inc. v. Carnival Air
Lines, Inc., 

, 1117 (9th Cir. 2000) (reviewing
for abuse of discretion).

    The court sanctioned Stone Creek’s attorneys for filing a
summary judgment motion on willfulness, reasoning that
summary judgment was not an appropriate vehicle for Stone
Creek to request a legal ruling that willfulness is not required
for a disgorgement of profits. The court stated that it could
not rule on willfulness until it had ruled on infringement and
held that Stone Creek’s arguments were frivolous on the
merits because Fifty-Six Hope forecloses Stone Creek’s
argument that, after the 1999 amendment, a showing of
willfulness is not necessary for disgorgement.

    On the latter point, the district court was incorrect as a
matter of law. Fifty-Six Hope does not address the effect of
the 1999 amendment on the continuing vitality of the
willfulness 

–74. Although the
1999 amendment is referenced in a citation, there is no
analysis or determination about the import of the
amendment. 

Instead, the court in Fifty-Six Hope
took willfulness as a given and did not need to go further
because willfulness was adequately established as a factual
matter. 

Importantly, at the time that Stone Creek
filed its motion, the interplay between the amendment and
the prior version of the statute remained an open question.
This point is underscored in the Ninth Circuit’s Model Jury
         STONE CREEK V. OMNIA ITALIAN DESIGN               33

Instructions: “The Ninth Circuit has not addressed . . .
whether willfulness remained a prerequisite to disgorgement
of a defendant’s profits as a result of the Trademark
Amendments Act of 1999.” Manual of Model Civil Jury
Instructions for the Ninth Circuit 15.29 cmt. (2017).

    Although we ultimately disagree with Stone Creek on the
merits of this issue, its arguments were not frivolous. The
unsettled nature of the question in our circuit provided Stone
Creek with a legitimate basis to ask the district court for a
legal ruling—namely, to determine whether to present
evidence of willfulness and Omnia’s profits, in addition to
Stone Creek’s damages, at trial. The unresolved legal issue,
combined with the fact that another circuit had accepted the
argument that the 1999 amendment did away with the
willfulness requirement, see Banjo 

,
legitimizes Stone Creek’s arguments. See W. Sys., Inc. v.
Ulloa, 

, 873–74 (9th Cir. 1992). We reverse
the sanctions order related to the willfulness issue.

    On the other hand, the court’s second sanctions order
reflects a discretionary judgment adequately grounded in the
law and record. The court sanctioned Stone Creek for not
earlier dropping its actual damages claim when it intended
to pursue only Omnia’s profits. As the court described,
Stone Creek had no evidence to support an actual damages
claim and, with the knowledge that “its actual damages claim
was meritless,” Stone Creek failed to withdraw the claim and
opposed Omnia’s motion to strike the claim.

    It is true that actual damages and defendant’s profits are
two distinct and well-recognized remedies available to the
plaintiff. 15 U.S.C. § 1117(a)(1)–(2); 5 McCarthy, supra,
§ 30:57. The district court’s order does not offend that
principle. The court did not question Stone Creek’s right to
pursue actual damages as an appropriate avenue of recovery;
34        STONE CREEK V. OMNIA ITALIAN DESIGN

instead, the court determined that Stone Creek had not
actually done so.

    As the district court explained, while Stone Creek’s
expert finally determined that proving actual damages would
be “too difficult,” Stone Creek never provided any analysis
from its expert on actual damages or identified what
information it was seeking during discovery to shed light on
the issue. Stone Creek’s expert acknowledged in his
deposition that he had not been asked to perform a damages
analysis or calculation. With the actual damages claim still
on the table, Omnia was forced to expend time and resources
defending against the claim by, for example, taking
depositions and having its expert prepare a report. When
confronted with Omnia’s motion to strike the actual damages
claim before trial, Stone Creek opposed that effort even
though it could point to no evidence to support the claim.
The district court acted within its discretion in concluding
that Stone Creek’s attorneys “unreasonably and vexatiously”
“multiplie[d] the proceedings” and awarding Omnia the
resulting attorneys’ fees.

                         Conclusion

    We reverse the district court’s finding that there is no
likelihood of confusion and reject the application of the Tea
Rose–Rectanus defense. Thus, we hold that Omnia is liable
for infringement of the STONE CREEK mark. We affirm
the district court’s conclusion that willfulness remains a
necessary condition for a disgorgement of profits but remand
for a determination on whether Omnia had the requisite
intent. Finally, we reverse the district court’s sanctions order
with respect to Stone Creek’s summary judgment on
willfulness but uphold the sanctions order with respect to
        STONE CREEK V. OMNIA ITALIAN DESIGN            35

Stone Creek’s continued assertion of the actual damages
claim.

  AFFIRMED IN PART, REVERSED IN PART, AND
REMANDED.

   Costs on appeal shall be awarded to Stone Creek, Inc.