OPINION

BOYCE F. MARTIN, JR., Circuit Judge.

Justice Hugo Black once wrote, "I was brought up to believe that Scotch whisky would need a tax preference to survive in competition with Kentucky bourbon." Dep't of Revenue v. James B. Beam Distilling Co., 377 U.S. 341, 348-49, 84 S.Ct. 1247, 12 L.Ed.2d 362 (1964) (Black, J., dissenting). While there may be some truth to Justice Black's statement that paints Kentucky bourbon as such an economic force that its competitors need government protection or preference to compete with it, it does not mean a Kentucky bourbon distiller may not also avail itself of our laws to protect its assets. This brings us to the question before us today: whether the bourbon producer Maker's Mark Distillery, Inc.'s registered trademark consisting of its signature trade dress element—a red dripping wax seal—is due protection, in the form of an injunction, from a similar trade dress element on Casa Cuervo, S.A. de C.V.'s Reserva de la Familia tequila bottles. We hold that it is. The judgments of the district court in this trademark infringement case are AFFIRMED.

I.

All bourbon is whiskey, but not all whiskey is bourbon.1 Whiskey, like other distilled spirits, begins as a fermentable mash, composed of water and grains or other fermentable ingredients. The mash is heated and then cooled, yeast is introduced to ferment the sugars in the mash, and the yeast turns the sugars into alcohol and carbon dioxide. This now-alcoholic liquid is then distilled to concentrate the alcohol. GARY REGAN & MARDEE HAIDIN REGAN, THE BOURBON COMPANION 32-33 (1998). The composition of the mash, and the aging, treating, and flavoring of the distilled alcohol, determine the flavor, color, and character of the distilled spirit. In the case of bourbon, the corn-based mash and aging in charred new oak barrels impart a distinct mellow flavor and caramel color. Distillers compete intensely on flavor, but also through branding and marketing; the history of bourbon, in particular, illustrates why strong branding and differentiation is important in the distilled spirits market.

The legend of the birth of bourbon is not without controversy: "As many counties of Kentucky claim the first production of Bourbon as Greek cities quarrel over the birthplace of Homer." H.F. WILLKIE, BEVERAGE SPIRITS IN AMERICA—A BRIEF HISTORY 19 (3d ed. 1949). The generally accepted and oft-repeated story is that "the first Bourbon whiskey . . . made from a mash containing at least fifty percent corn, is usually credited to a Baptist minister, The Reverend Elijah Craig, in 1789, at Georgetown, [Kentucky]," just prior to Kentucky's joining the Union as a state in 1792. Id. But it is more likely that Kentucky whiskey was first distilled at Fort Harrod, the first permanent European settlement in what is now Kentucky, in 1774. CHARLES K. COWDERY, BOURBON, STRAIGHT: THE UNCUT AND UNFILTERED STORY OF AMERICAN WHISKEY 3-4 (2004); accord WILLKIE, supra, at 19. Kentucky's settlers distilled whiskey using methods similar to those "used in Scotland and Ireland for hundreds of years," WILLKIE, supra, at 20, except that Kentucky whiskey was made mostly from corn, a crop unknown to Europeans before Columbus ventured to America. COWDERY, BOURBON, STRAIGHT, supra, at 2. Though "most [American] colonial whiskey was made from rye," id. at 3, corn was easy to grow in Kentucky soil, and surplus corn was often used to make whiskey. Id. at 4.

The name "bourbon" itself is easier to trace: one of the original nine counties of Kentucky was Bourbon County, WILLKIE, supra, at 20, named in honor of the French royal family. Charles K. Cowdery, How Bourbon Whiskey Really Got Its Famous Name, BOURBON COUNTRY READER, July 1996. "[Kentucky] whiskey was shipped from Limestone, a riverside port in Bourbon County," down the Ohio river to the Mississippi, bound for New Orleans. REGAN & REGAN, supra, at 14. Whiskey shipped from the port in Bourbon County came to be known as "Old Bourbon," and later, simply "Bourbon," to distinguish it from Pennsylvania Rye or other whiskeys. Cowdery, How Bourbon Whiskey Really Got Its Famous Name, supra. The name "bourbon" at that time meant whiskey made from mostly corn in Kentucky or points west. But it was likely not until "sometime between 1823 and . . . 1845" that Dr. James Crow "perfect[ed] the sour-mash method of whiskey-making"—the dominant process in use today that, when coupled with aging in charred new oak barrels, produces modern bourbon's familiar caramel color and distinctive taste. REGAN & REGAN, supra, at 15.

While in the early years "[w]hiskey was whiskey, as everybody knew," some bourbon distillers began to brand their bourbons to capitalize on the differences between "[g]ood Kentucky Bourbon" and all the rest. WILLKIE, supra, at 22. Dr. Crow, a Kentuckian by way of Scotland, "insist[ed] upon strict sanitation in his manufacture," and branded his bourbon with his name; other Kentucky families followed suit in an effort to differentiate their products. Id. Crow's branding tactics seem to have worked, as his bourbon accumulated prominent fans. For example, bourbon drinker Ulysses S. Grant preferred Old Crow over other bourbons, Julia Reed, Bourbon's Beauty, NEWSWEEK, Dec. 21, 2008, as did all three of Congress's "Great Triumvirate," Henry Clay, John C. Calhoun, and Daniel Webster. GERALD CARSON & MIKE VEACH, THE SOCIAL HISTORY OF BOURBON 47 (2010).

Success attracts imitators, and in the late nineteenth century "rectifiers" began to crowd the market, selling "a product that they would call `Kentucky Bourbon' using neutral spirits, flavoring agents and artificial coloring with only some aged whiskey in the product." Mike Veach, The Taft Decision, THE FILSON, Winter 2009, at 4. A hotly contested legal and lobbying war between the rectifiers and traditional "straight whiskey" distillers erupted, culminating in President William Taft's official interpretation, in 1909, of the 1906 Pure Food and Drug Act; Taft's interpretation settled the question of what spirits could be labeled as "whiskey." Id. The rectifiers lost and were required to label their product "imitation whiskey." See id.; see also H. Parker Willis, What Whiskey Is, McCLURE'S MAGAZINE, 1909-10, at 687-903. The ruling only increased distillers' incentives to differentiate themselves and their products. "Before the Taft ruling, few brands were nationally known. . . . But, under the new regulations, labels had to tell both the process and materials of manufacture. Whiskey . . . now began to appear under distinctive labels, competing with other brands on its own merits." WILLKIE, supra, at 26. After Prohibition was repealed, the distilled spirits industry consolidated and matured, id. at 27, and bourbon continued to attract notable adherents. Ian Fleming, the writer who created the James Bond character that famously favored martinis, switched from martinis to bourbon as his drink of choice. John Pearson, Rough Rise of a Dream Hero, LIFE, Oct. 14, 1966, at 113, 126. And Harry S. Truman started his day with a walk followed by "a rubdown, a shot of bourbon, and a light breakfast." Univ. of Va. Miller Cntr., Harry S. Truman: Family Life, http://millercenter.org/president/ truman/essays/biography/7.

In recognition of bourbon's unique place in American culture and commerce, and in the spirit of the Taft decision, Congress in 1964 designated bourbon as a "distinctive product[] of the United States," 27 C.F.R. § 5.22(l)(1), and prescribed restrictions on which distilled spirits may bear the label "bourbon." Federal regulations require that bourbon whiskey to, among other things, be aged in charred new oak barrels, contain certain proportions of mash ingredients, and be barreled and bottled at certain proofs. § 5.22(b). Importantly, whiskey made for consumption within the United States cannot be called bourbon unless it is made in the United States. § 5.22(l)(1). While bourbon is strongly associated with Kentucky, and while "[ninety-five] percent of the world's supply of bourbon comes from Kentucky," Jessie Halladay, Kentucky's Libation Vacations, COURIER-J., Feb. 26, 2012, at D1, some notable bourbons are made in other states.2

Maker's Mark occupies a central place in the modern story of bourbon. The Samuels family, founder of the Maker's Mark distillery in Loretto, Kentucky, has produced whiskey in Kentucky nearly continuously from the eighteenth century through today. REGAN & REGAN, supra, at 161-62. Indeed, Robert Samuels (along with Jacob Beam, Basil Hayden, and Daniel Weller, all of whose surnames are familiar to bourbon connoisseurs) was one of Kentucky's early settlers. COWDERY, BOURBON, STRAIGHT, supra, at 4. Bill Samuels, Sr. formulated the recipe for Maker's Mark bourbon in 1953. His wife, Margie, conceived of the red dripping wax seal and used the family deep fryer to perfect the process of applying it. The company has bottled bourbon for commercial sale under the Maker's Mark name, and has used a red dripping wax seal on its Maker's Mark bourbon bottles, since 1958. Maker's Mark, and craft bourbon generally, garnered national attention when the Wall Street Journal published a front-page article about the bourbon, the red dripping wax seal, and the family behind it. David P. Garino, Maker's Mark Goes Against the Grain to Make its Mark, WALL ST. J., Aug. 1, 1980, at 1. In 1985, Maker's Mark registered a trademark for the dripping-wax-seal element of its trade dress, which it described as a "wax-like coating covering the cap of the bottle and trickling down the neck of the bottle in a freeform irregular pattern." The trademark is silent as to color, but Maker's Mark conceded in submissions before the district court that it sought only to enforce it as applied to the red dripping wax seal.

Jose Cuervo produced a premium tequila, "Reserva de la Familia," beginning in 1995. The tequila bottle had a wax seal that was straight-edged and did not initially feature drips. By 2001, Cuervo had begun selling this tequila in the United States in bottles with a red dripping wax seal reminiscent of the Maker's Mark red dripping wax seal. In 2003, Maker's Mark instituted this suit against Casa Cuervo S.A. de C.V., Jose Cuervo International, Inc., Tequila Cuervo La Rojeña S.A. de C.V., and Diageo North America, Inc. claiming state and federal trademark infringement and federal trademark dilution; sometime thereafter, Cuervo discontinued use of the red dripping wax seal and reverted to a red straight-edged wax seal. In its suit, Maker's Mark sought damages, injunctions against dilution and infringement, and costs. Cuervo counterclaimed for cancellation of the Maker's Mark trademark.

After a six-day bench trial, the district court found that Maker's Mark's red dripping wax seal is a valid trademark and that Cuervo had infringed that trademark. Based on those findings, the district court enjoined Cuervo permanently "from using red dripping wax on the cap of a bottle in the sale, offering for sale, distribution or advertising of Cuervo tequila products at any locality within the United States." The district court found that Cuervo had not diluted the mark and denied Maker's Mark's claim for damages; the district court also denied Cuervo's counterclaim for cancellation of the mark. In a separate opinion, the district court awarded Maker's Mark some of its costs.

Cuervo appeals the district court's determination that the red dripping wax seal is not aesthetically functional, some of the district court's factual findings, its balancing of those findings in determining Cuervo had infringed, and its award of some of Maker's Mark's costs. Cuervo does not appeal the scope of the injunction.

II.

A. Aesthetic Functionality

In a trademark case, this Court reviews a district court's factual findings for clear error and legal conclusions de novo. Wynn Oil Co. v. Thomas, 839 F.2d 1183, 1186 (6th Cir.1988). For a trademark to be enforceable, it must be valid; one way to show a mark's validity is through its "incontestability." A trademark registered for five or more years becomes "incontestable" under 15 U.S.C. § 1065. Incontestability is "conclusive evidence of the validity of the registered mark," id. § 1115(b), except as to certain statutorily enumerated challenges, including the functionality of the mark, id. § 1115(b)(8). A registered mark may be found invalid if it is "functional." See id.; id. § 1052(e)(5). A trademark is functional "if it is essential to the use or purpose of the article or if it affects the cost or quality of the article." Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159, 165, 115 S.Ct. 1300, 131 L.Ed.2d 248 (1995) (internal quotation marks omitted). A trademark may be determined to be functional under traditional functionality doctrine, as in Qualitex. Cuervo also argued the seal was functional under an additional functionality doctrine, "aesthetic functionality." The district court concluded that the red dripping wax seal was not functional under either doctrine and that the mark was valid.

Cuervo appeals only the district court's ruling as to aesthetic functionality, arguing that the red dripping wax seal is aesthetically functional, and therefore the mark is not enforceable. The Supreme Court has discussed the concept of aesthetic functionality in dicta, noting that "[i]t is proper to inquire into a significant non-reputation-related disadvantage in cases of [aesthetic] functionality." TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 33, 121 S.Ct. 1255, 149 L.Ed.2d 164 (2001) (internal quotation marks omitted). We have interpreted this dicta to propose that, "where an aesthetic feature (like color), serves a significant function . . . courts should examine whether the exclusive use of that feature by one supplier would interfere with legitimate competition." Antioch Co. v. W. Trimming Corp., 347 F.3d 150, 155 (6th Cir.2003). It seems we have not yet plainly stated which test we would apply under aesthetic functionality doctrine, see Abercrombie & Fitch Stores, Inc. v. Am. Eagle Outfitters, Inc., 280 F.3d 619, 641 n. 16 and 642-43 (6th Cir.2002), or that we have even adopted aesthetic functionality doctrine at all, Antioch Co., 347 F.3d at 155-56 (questioning the validity of aesthetic functionality doctrine in the Sixth Circuit). We need not decide these questions today.

Under the competition theory of functionality adopted by the Sixth Circuit, we have considered two different tests, Abercrombie, 280 F.3d at 641 n. 16 and 642 (citations and internal quotation marks omitted), to determine whether a trademark is functional and, thus, not enforceable:

The inquiry in both tests is factual in nature. See id.

Even assuming we were to recognize aesthetic functionality doctrine, regardless of which test we would apply under that doctrine, the outcome is the same. Under either test, Cuervo's appeal on this claim does not succeed. The district court was not convinced "that it would be difficult or costly for competitors to design around" the mark and we do not disagree. There is more than one way to seal a bottle with wax to make it look appealing, and so Cuervo fails the comparable alternatives test. As to the effective competition test, the district court found that "red wax is not the only pleasing color of wax . . . nor does it put competitors at a significant non-reputation related disadvantage to be prevented from using red dripping wax." The district court's findings are not clearly erroneous, and, based on those findings, Cuervo fails either test.

B. Factual Findings under Frisch

We have recognized four kinds of trademark infringement: palming off, confusion of sponsorship (also known as "association"), reverse confusion of sponsorship, and dilution. Ameritech, Inc. v. Am. Info. Techs. Corp., 811 F.2d 960, 964-65 (6th Cir.1987). As the district court correctly noted, Maker's Mark focuses on "confusion of sponsorship." Maker's Mark does not appeal the district court's adverse ruling on its dilution claim. Confusion of sponsorship "occurs where the goods do not directly compete. In this situation, the goods are unrelated enough that no inference arises that they originated from the same source, but the similarity of the trademarks erroneously suggests a connection between the sources." Id. at 964. In any case, a court considering a claim for trademark infringement must determine the likelihood of consumer confusion. The factors the court should consider are: "1. strength of the plaintiff's mark; 2. relatedness of the goods; 3. similarity of the marks; 4. evidence of actual confusion; 5. marketing channels used; 6. likely degree of purchaser care; 7. defendant's intent in selecting the mark; [and] 8. likelihood of expansion of the product lines." Frisch's Rests., Inc. v. Elby's Big Boy, Inc., 670 F.2d 642, 648 (6th Cir.1982) (quoting Toho Co., Ltd. v. Sears, Roebuck & Co., 645 F.2d 788, 790 (9th Cir.1981)).

We review the district court's factual findings under Frisch for clear error. Tumblebus Inc. v. Cranmer, 399 F.3d 754, 764 (6th Cir.2005). We assess each factor with respect to the relevant consumer market; potential buyers of the "junior" product (here, Cuervo's Reserva de la Familia) are the relevant consumers. Leelanau Wine Cellars, Ltd. v. Black & Red, Inc., 502 F.3d 504, 518 (6th Cir.2007). Cuervo appeals the district court's findings on only three of the eight Frisch factors: strength, similarity, and actual confusion.

1. Strength

To evaluate the strength factor under the Frisch analysis, this Court "focuses on the distinctiveness of a mark and its recognition among the public." Therma-Scan, Inc. v. Thermoscan, Inc., 295 F.3d 623, 631 (6th Cir.2002). One leading commentator usefully characterizes this evaluation as encompassing two separate components: (1) "conceptual strength," or "placement of the mark on the spectrum of marks," which encapsulates the question of inherent distinctiveness; and (2) "commercial strength" or "the marketplace recognition value of the mark." 2 J. THOMAS McCARTHY, McCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 11.83 (4th ed.). In other words, "[a] mark is strong if it is highly distinctive, i.e., if the public readily accepts it as the hallmark of a particular source; it can become so because it is unique, because it has been the subject of wide and intensive advertisement, or because of a combination of both." Homeowners Grp. v. Home Mktg. Specialists, Inc., 931 F.2d 1100, 1107 (6th Cir.1991) (internal quotation marks omitted).

Because the strength of a trademark for purposes of the likelihood-of-confusion analysis depends on the interplay between conceptual and commercial strength, the existence of inherent distinctiveness is not the end of the inquiry. See Therma-Scan, Inc., 295 F.3d at 631-32 (noting that a mark can be inherently distinctive but not especially strong if it fails to attain broad public recognition); Homeowners Grp., Inc., 931 F.2d at 1107 ("The District Court's finding that HMS was an arbitrary and inherently distinctive mark is only a first step in determining the strength of a mark in the marketplace."); see also McCARTHY, supra § 11:83 ("[T]he true relative strength of a mark can only fully be determined by weighing [both] aspects of strength."). Thus, although inherent distinctiveness may provide powerful support for the strength of a mark, the full extent of that support nonetheless depends on the scope of commercial recognition.

Here, the district court appropriately evaluated both components of the strength factor. From the physical characteristics of the mark, the district court specifically found the red dripping wax seal to be inherently distinctive based on its uniqueness and its potential to "draw in the customer" in an unusual manner. This finding of conceptual strength is bolstered by the mark's status as "incontestable," which entitles it to a presumption of strength, though the relative import of that presumption within the overall strength analysis still requires an analysis of "whether the mark is distinctive and well-known in the general population." Therma-Scan, Inc., 295 F.3d at 632; see also Wynn Oil, 839 F.2d at 1187. As to commercial recognition, the district court found the seal "acquired secondary meaning through fifty years of use, extensive advertising and consumer recognition."3 The district court also found that Maker's Mark's advertising was intensive, citing the extent of its advertising budget that "focuses almost entirely on branding the red dripping wax," as well as the significant public attention that the wax seal has received through the media. In further support of these findings, the district court also cited studies showing significant amounts of consumer dialogue about the brand, as well as a high level of recognition among both whiskey drinkers and distilled-spirits drinkers more generally.

Cuervo argues that the district court erred in its evaluation of the strength of the mark by (1) disregarding third-party use of red dripping wax seals; (2) failing to give proper weight to the lack of a survey regarding recognition of the red dripping wax seal; (3) relying in its analysis on Maker's Mark's advertisements without apparent evidence of their dates or circulation; and (4) relying on evidence of the strength of the mark in the overbroad group of distilled spirits drinkers instead of prospective Reserva purchasers.

We recognize that "extensive third-party uses of a trademark [may] substantially weaken the strength of a mark." Homeowners Grp., 931 F.2d at 1108; Herman Miller, Inc. v. Palazzetti Imps. & Exps., Inc., 270 F.3d 298, 317 (6th Cir. 2001) (noting the possibility that the strength of a plaintiff's mark may be "`weakened' by widespread use in the market," causing the mark to "lose its significance as an indication of source." (quoting McCARTHY, supra § 17:17)). Contrary to Cuervo's argument, the district court did consider evidence of third-party use of similar seals on distilled spirits, but rejected that evidence as limited and unconvincing because it concerned seals used on all distilled spirits; the court found that the relevant use of the seals is limited to the "relevant market," and not among all distilled spirits. We agree with the district court's finding and reasoning.

Next, while "survey evidence is the most direct and persuasive evidence" of whether a mark has acquired secondary meaning, "consumer surveys . . . are not a prerequisite to establishing secondary meaning." Herman Miller, Inc., 270 F.3d at 312, 315 (citations and internal quotation marks omitted). Nor is such evidence indispensable to the broader question of commercial recognition. In light of the abundance of other evidence demonstrating market recognition, such as Maker's Mark's extensive marketing efforts focusing on the red dripping wax seal and its widespread publicity, it was not clear error for the district court to overlook the lack of survey evidence because that evidence was not determinative of the strength of the mark.

As to the district court's consideration of advertising evidence, the district court discussed the nature of the advertising and found that advertising efforts by Maker's Mark usually focus directly on the red dripping wax seal. As the record and the district court's opinion show, the district court had before it, and considered, an abundance of Maker's Mark advertisements that specifically feature the red dripping wax seal. Moreover, these advertisements were recent, relevant, and strong enough to convince Business Week, in 2002, to declare the dripping wax seal "one of the most recognizable branding symbols in the world," and CBS Sunday Morning, in 2008, to refer to the process by which the seal is applied as the "famous dip in red sealing wax." These findings support the district court's ultimate conclusion regarding the breadth of market recognition of Maker's Mark's trademarked red dripping wax seal.

Finally, as to the district court's discussion of evidence of the mark's strength within the broader group of distilled spirits drinkers, the district court considered, but did not rest its holding on, this evidence. Instead, the district court based its holding primarily on the seal's "unique design and [Maker's Mark's] singular marketing efforts." We therefore find no error here.

In sum, none of Cuervo's arguments undermines the district court's finding "that the Maker's Mark red dripping wax seal is an extremely strong mark due to its unique design and the company's singular marketing efforts." We therefore conclude that the district court did not clearly err in its evaluation of the strength of the red dripping wax seal.

2. Similarity

In assessing similarity, "courts must determine whether a given mark would confuse the public when viewed alone, in order to account for the possibility that sufficiently similar marks may confuse consumers who do not have both marks before them but who may have a general, vague, or even hazy, impression or recollection of the other party's mark." Daddy's Junky Music Stores, Inc. v. Big Daddy's Family Music Cntr., 109 F.3d 275, 283 (6th Cir.1997) (internal quotation marks omitted). The district court found this factor "narrowly favor[s] Maker's Mark," and found that, though "[v]ery few consumers . . . would buy one product believing it was the other," the seals were facially similar. The district court examined the two seals and found that "nothing on the products other than the red dripping wax . . . would suggest an association between the two."

Cuervo focuses its argument on the relevance of the house marks—product labels identifying the name of the manufacturer—on the bottles. We have held that the presence of a house mark can decrease the likelihood of confusion. Therma-Scan, Inc., 295 F.3d at 634 ("[T]he presence of [a house mark on a product] does not eliminate the similarity between the trademarks. Instead, this labeling diminishes the likelihood of confusion created by the comparable marks and reduces the importance of this factor."); AutoZone, Inc. v. Tandy Corp., 373 F.3d 786, 797 (6th Cir.2004) ("The co-appearance of a junior mark and a house mark is not dispositive of dissimilarity, but it is persuasive."). The district court concluded that this consideration is not "as important in an association case, when the two products are related enough that one might associate with or sponsor the other and still use their own house mark."

In AutoZone, we found that the proximity of the Radio Shack house mark to the "POWERZONE" mark would alleviate any confusion between POWERZONE and AUTOZONE marks. AutoZone does not, however, stand for the proposition that the presence of a house mark always has significant weight in the similarity analysis; it merely states that presence of a house mark is a factor to be considered in the evaluation of similarity and, depending on the facts of the case, may be significant to the overall likelihood of confusion. AutoZone, Inc., 373 F.3d at 796-97. Furthermore, the district court's analysis in this case highlights two factors that diminish the significance of the house marks in the present context. First, testimony in the record indicates that many consumers are unaware of the affiliations between brands of distilled spirits, and that some companies produce multiple types of distilled spirits, which supports the district court's assessment here. Second, the presence of a house mark, as the district court correctly noted, is more significant in a palming off case than in an association case—as the district court reasoned, in an association case "when the two products are related enough . . . one might associate with or sponsor the other and still use their own house mark." Accordingly, the district court did not clearly err in its factual findings under this factor, and we adopt its findings.

3. Actual Confusion

The district court stated that "neither party produced meaningful evidence related to actual confusion" and concluded that the lack of evidence was "neutral." The district court reasoned that, though evidence of actual confusion might have been obtainable if it existed, Cuervo sold Reserva for a limited time and in limited quantities, and so the district court did not place weight on the fact that Maker's Mark did not furnish "meaningful" evidence of actual confusion. Despite Cuervo's arguments to the contrary, this finding falls squarely within this Circuit's case law. Though "[e]vidence of actual confusion is undoubtedly the best evidence of likelihood of confusion . . . a lack of such evidence is rarely significant." Daddy's, 109 F.3d at 284 (citation and internal quotation marks omitted). Here, the Reserva product was sold for a short time and in limited quantities; under these circumstances, it is reasonable that no meaningful evidence of actual confusion was available. The district court did not clearly err in finding the lack of actual confusion evidence non-determinative, and we adopt its findings.

C. Balancing the Frisch Factors

We "review de novo the legal question of whether [the district court's Frisch factual findings] constitute a likelihood of confusion." Tumblebus Inc., 399 F.3d at 764 (quoting Champions Golf Club, Inc. v. The Champions Golf Club, Inc., 78 F.3d 1111, 1116 (6th Cir.1996)) (internal quotation marks omitted). Because we find above that the district court did not reversibly err in its factual findings on the three disputed factors—strength, similarity, and actual confusion—and because the parties do not dispute the district court's factual findings under the remaining five factors, we adopt all of the district court's factual findings and balance them de novo.

1. Strength. The district court found the evidence of the strength of the mark heavily favored Maker's Mark. We have held that the strength of the mark supplies the weight it should be accorded in balancing. In general, "[t]he stronger the mark, all else equal, the greater the likelihood of confusion." AutoZone, 373 F.3d at 794 (alteration in original) (quoting Homeowners Grp., Inc., 931 F.2d at 1107). Because the district court found the mark at issue here to be "extremely strong," the strength factor is weighed very heavily.

2. Relatedness of the goods. The district court found the goods were somewhat related because they were part of the same broad category of high-end distilled spirits, but not fully related because the Cuervo product was priced at $100 per bottle, while Maker's Mark sold for $24 per bottle. Where the goods are "somewhat related but not competitive, the likelihood of confusion will turn on other factors." Daddy's, 109 F.3d at 282. Here, the district court found that the products are somewhat related. We accord this factor little weight because the products are competitive only within a very broad category and are only somewhat related; it is thus more appropriate to concentrate the weight of our balancing analysis on other factors.

3. Similarity. The district court found the similarity factor "narrowly favors Maker's [Mark]." "The similarity of the senior and junior marks is `a factor of considerable weight.'" AutoZone, 373 F.3d at 795 (quoting Daddy's, 109 F.3d at 283).

4. Actual confusion. As discussed above, "a lack of such evidence is rarely significant, and the factor of actual confusion is weighted heavily only where there is evidence of past confusion, or perhaps, when the particular circumstances indicate such evidence should have been available." Daddy's, 109 F.3d at 284 (internal quotation marks omitted). The district court found that this factor was neutral. As we noted, the Reserva product was sold for a short time and in limited quantities; under these circumstances, we give the lack of evidence of actual confusion little weight.

5. Marketing channels used by the parties. The court found the channels "similar in some ways and dissimilar in others. Perhaps this factor marginally favors Maker's Mark." The weight of this factor will not add much to a finding of infringement because of the equivocal nature of the district court's factual findings. We accord this factor very little weight.

6. Likely degree of purchaser care. The district court found this factor "clearly favors" Cuervo because of the degree of care potential tequila customers would exercise in purchasing a $100 bottle of Reserva; knowledgeable bourbon customers would also exercise similar care and, further, know that Maker's Mark sells only one kind of liquor. This factor, though strongly in favor of Cuervo, is not dispositive. "[C]onfusingly similar marks may lead a purchaser who is extremely careful and knowledgeable . . . to assume nonetheless that the seller is affiliated with or identical to the other party." Id. at 286. For these reasons, we give this factor substantial weight.

7. Intent. The district court found Cuervo did not intend to infringe, but we give no weight to this finding because "[i]ntent is an issue whose resolution may benefit only the cause of the senior user, not of an alleged infringer." Leelanau, 502 F.3d at 520 (internal quotation marks omitted).

8. Likelihood of expansion of product lines. The district court found this factor was neutral where neither party put forth evidence of significant expansion plans. Because a "strong possibility that either party will expand his business to compete with the other . . . will weigh in favor of finding that the present use is infringing," Daddy's, 109 F.3d at 287 (internal quotation marks omitted), a finding of little evidence of expansion plans is accorded little to no weight, but does not weigh against Maker's Mark, who, by this test, would benefit by any significant evidentiary showing under this factor, no matter which of the parties intended to expand. For these reasons, we give this factor no weight.

The balance of the factors compels a finding of infringement. Excluding the neutral factors, the majority of the factors—strength, relatedness of the goods, similarity, and marketing channels—favor Maker's Mark. The district court found that Maker's Mark's trademark is "extremely strong," and we have adopted that finding. Further, we have said that the "most important Frisch factors" are similarity and strength of the mark, Gray v. Meijer, Inc., 295 F.3d 641, 646 (6th Cir. 2002); both of these factors favor Maker's Mark. The "likely degree of purchaser care" factor "clearly" favors Cuervo. Though this factor is given substantial weight, this factor alone cannot override the "extreme" strength of the mark that, when coupled with similarity (which itself is given "considerable weight"), and combined with the two other factors weighing in favor of Maker's Mark, together favor a finding of infringement. Buttressing this determination is that, in its briefing, Cuervo complains of errors in the district court's factual determinations, but does not argue that, even given the factual findings made by the district court, a de novo balancing under Frisch should come out in Cuervo's favor. While Cuervo disputes the factual findings themselves and the related outcome of the balancing, it does not argue that the weight given the factors should have been different.

We conclude that there is a likelihood of confusion between the products and that Cuervo has infringed.

III.

In a separate memorandum opinion and order, the district court awarded Maker's Mark $66,749.21 of the $72,670.44 in costs it requested. Cuervo argues that Maker's Mark is not a "prevailing party" under Federal Rule of Civil Procedure 54(d) because the district court only awarded Maker's Mark a permanent injunction, but not damages, on its infringement claim, and denied its request for a permanent injunction on its dilution claim. Cuervo argues that Maker's Mark should have either been awarded no costs or that the costs should have been apportioned.

This Court reviews a district court's decision to award costs under an abuse of discretion standard. Singleton v. Smith, 241 F.3d 534, 539-40 (6th Cir.2001). "Generally, [finding an abuse of discretion] would require the lower court ignoring the criteria set by [the] Sixth Circuit or otherwise a certainty on [this Court's] part that a clear error in judgment was committed." Id. (citation omitted).

Rule 54(d) provides that "costs—other than attorney's fees—should be allowed to the prevailing party." Under Buckhannon Board and Care Home v. West Virginia Department of Health and Human Resources, 532 U.S. 598, 121 S.Ct. 1835, 149 L.Ed.2d 855 (2001), a party is the prevailing party where (1) it receives "at least some relief on the merits of [its] claim," and (2) there is a "judicially sanctioned change in the legal relationship of the parties." Id. at 603, 605, 121 S.Ct. 1835. Here, the district court did not abuse its discretion in awarding costs to Maker's Mark where Maker's Mark secured an injunction against Cuervo. Maker's Mark won on the merits of its infringement claim and the permanent injunction is a judicially sanctioned change in the relationship between the parties. Andretti v. Borla Performance Indus., Inc., 426 F.3d 824, 836 (6th Cir.2005) (holding that even voluntary agreement to injunction "can support a determination that a party prevailed"). Maker's Mark did not need to win every claim to be considered the prevailing party. See Lewis v. Pennington, 400 F.2d 806, 820 (6th Cir.1968). Cuervo does not dispute the particulars of the district court's decision about which costs to allow and which costs to disallow, but merely seeks wholesale denial or apportionment of costs. The district court did not abuse its discretion in awarding some costs to the prevailing party.

IV.

The district court did not reversibly err in its decisions regarding aesthetic functionality and in its factual findings under Frisch. We conclude that the Frisch factors weigh in Maker's Mark's favor. Further, we hold the district court did not abuse its discretion in awarding some of Marker's Mark's costs. The judgments of the district court are AFFIRMED.