Although the aches and pains derived from physical activity have always been part of our American heritage, only recently have they become fashionable and profitable.

In the spring of 1982, Thompson Medical Company (“Thompson”) introduced “Sport-screme,” a topical analgesic designed to relieve the muscle soreness associated with sports activities. Two years later, in the spring of 1984, Pfizer Inc. (“Pfizer”) introduced a similar product, “BEN-GAY SportsGel,” which was also targeted specifically at the “weekend athlete.” On August 16, 1984, Thompson filed suit in the United States District Court for the Southern District of New York, pursuant to Section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a) (1982), seeking a preliminary injunction against Pfizer’s further use of the name “SportsGel.” Judge Griesa, finding a likelihood of confusion between “Sport-screme” and “BEN-GAY SportsGel,” granted the preliminary injunction. He based his decision solely on the similarity of the marks and Thompson’s priority of use.

Because the court below did not address the question whether “Sportscreme” was a protectible mark, and failed to examine the strength of “Sportscreme,” the proximity of the marks, the sophistication of topical analgesic consumers, and Pfizer’s intent in selecting its mark, we vacate the order granting the preliminary injunction and remand the case for further proceedings.

Suits for trademark infringement demand a “comprehensive analysis of all the relevant facts and circumstances.” See Vi-tarroz Corp. v. Borden, Inc., 644 F.2d 960, 968-69 (2d Cir.1981). Accordingly, we set forth the facts giving rise to the dispute in some detail.

I.

Thompson’s “Sportscreme”

Thompson, a New York corporation, is this nation’s second-most successful manufacturer of topical analgesic rubs. ^[1] To cap[*210] italize on the fitness craze of the past decade, Thompson, in 1979, conceived of “Sportscreme,” a topical analgesic directed at the growing group of “weekend warriors,” those sedentary individuals who engage in sporadic physical activity and are thus particularly susceptible to muscle pain and soreness.

After selecting the name “Sportscreme,” Thompson conducted two trademark searches, which revealed that the word “sports” was used in connection with a variety of products. From late 1979 until “Sportscreme’s” launch in February 1982, Thompson engaged in concept development and product formulation, and prepared testing, trade and consumer programs. Sales to consumers began in April 1982.

Patently aware of Pfizer’s dominant position in the topical analgesic market, Thompson embarked on an extensive and intensive advertising campaign to promote and publicize “Sportscreme.” With total expenditures in excess of $8 million, Thompson employed electronic and print media to reach the “weekend athlete.” To this end, “Sportscreme” advertisements were aired on ESPN (a cable television sports network) and carried in Runner magazine. Among the messages conveyed by Thompson was that unlike “smelly BEN-GAY,” “Sportscreme,” with its aromatic scent, leaves its users smelling “nice and clean.” “Sportscreme” samples were distributed without charge at swimming, biking, golfing and triathlon competitions.

The success of this advertising campaign was manifested by “Sportscreme’s” steadily increasing sales volume and market share. From April to November of 1982, retail sales of “Sportscreme” totalled $700,-000; in the fiscal year ending November 30, 1983, sales rose to $1.2 million; and for the fiscal year ending November 30, 1984, retail sales revenues were projected at more than $2.5 million. Indeed, within an eighteen month period, “Sportscreme’s” market share had increased from 0.8% to 4.2%.

Pfizer’s “BEN-GAY SportsGel”

Pfizer, a Delaware corporation with its principal place of business in New York, is a major manufacturer of pharmaceutical products, and is the nation’s leading producer of topical analgesic rubs. Since acquiring the assets of Bengue, Inc. in 1964, Pfizer has enjoyed a 40% share of the topical analgesic market.

For more than three-quarters of a century, Pfizer’s BEN-GAY was typically purchased and used by arthritis sufferers and persons afflicted with ordinary muscle aches. But as Americans became more interested in fitness, the universe of consumers to which topical analgesic rubs could be marketed exploded. This potential source of market expansion was not overlooked by Pfizer executives. Beginning in 1981, BEN-GAY’s advertising campaign assumed a physical-fitness orientation, with major sports figures extolling the virtues of BEN-GAY. A by-product of this advertising was to inculcate in millions of American weekend athletes an understanding of the therapeutic effects of topical analgesics.

In 1976, Pfizer contemplated introducing a new line of BEN-GAY products aimed specifically at the sports-active consumer. Although the seeds of the so-called “BEN-GAY Sports Stick” were sown in 1976, preliminary testing from 1977 to 1979 delayed its introduction, and manufacturing complications ultimately forced Pfizer to abandon the stick formulation entirely. Buoyed by the sales increase generated from its sports-related advertising campaign, Pfizer remained determined to capitalize fully on the nation’s physical fitness trend.

Consequently, in the fall of 1983, Pfizer decided to reposition the existing BEN-GAY gel formulation as a sports product, stressing its value to persons who engage in sports activities, while simultaneously deemphasizing its ability to relieve minor arthritis pain. The product was named “BEN-GAY SportsGel.” Thomas Laugh-lin, a group marketing director for BEN-GAY products, testified that the name was[*211] selected because it concisely and accurately described the product: a BEN-GAY gel product for use by sports participants. Laughlin explained that the name “BEN-GAY SportsGel” was the natural derivative of its abandoned predecessor, “BEN-GAY Sports Stick.”

Pfizer’s lawyers then commissioned a trademark search that uncovered, in addition to “Sportscreme,” a litany of marks using the word “sports” in connection with closely related products — Sports-Rub, Sports Balm, Sport Cream, Sport Lotion, Sport Oil, etc. There were also pending trademark applications for two topical analgesic rubs that would compete directly with “BEN-GAY SportsGel,” called SportsRub and Sports Ice. After reviewing this data, Pfizer’s in-house trademark lawyer, William Goebelbecker, concluded that Pfizer was “reasonably safe” in adopting the name “SportsGel.”

Armed with a name and a product, Pfizer next devised a comprehensive packaging strategy. “BEN-GAY SportsGel,” like the other BEN-GAY tube products (and like “Sportscreme”), was packaged in standard IV4 and 3 ounce containers. And like all products in the BEN-GAY “family,” the “SportsGel” container prominently displayed the BEN-GAY logo (emblazoned in red) and the traditional diagonal stripes in three shades of the same color. The package’s red, white and blue color scheme was selected for its “bright, bold, patriotic” qualities. The dominant colors on the “Sportscreme” container were yellow and green.

Both Thompson’s “Sportscreme” and Pfizer’s “BEN-GAY SportsGel” utilized a “fifth panel,” a riser that extends from the back of the box and is believed to enhance a product’s saleability, particularly during the period of its introduction. Before determining the precise packaging configuration, Pfizer conducted a cost-benefit analysis to determine the financial efficacy of utilizing a fifth panel. In light of Pfizer’s desire not to advertise “BEN-GAY Sports-Gel” independently from other products in the BEN-GAY line, the marketing benefits to be derived from the more costly configuration were predicted to exceed the added expense. These calculations proved to be correct. “BEN-GAY SportsGel” was introduced in the spring of 1984, and after just two and one-half months, sales had exceeded one million dollars.

The Trial Court Proceedings

On August 16, 1984, three months after “BEN-GAY SportsGel” had been introduced and full national distribution was under way, Thompson filed suit in the United States District Court for the Southern District of New York, pursuant to Section 43(a) of the Lanham Act, seeking a preliminary injunction against Pfizer’s continued use of the name “SportsGel.” A two-day hearing was conducted by Judge Griesa on August 30 and 31, 1984.

On August 31, Judge Griesa delivered an oral opinion from the bench. Describing the case as a “close one,” the district judge found that Thompson’s survey purportedly establishing actual confusion among consumers, ^[2] was so replete with testing flaws as to be “of limited utility,” and that the “BEN-GAY SportsGel” and “Sportscreme” packages are “strikingly different” in their “essential features.” Nevertheless, Judge Griesa granted the preliminary injunction. His conclusion that the name “SportsGel” infringed Thompson’s use cf. “Sport-screme” was based upon the similarity of the marks and Thompson’s priority of use. He failed, however, to address whether Thompson’s mark was protectible. Neither did he evaluate the strength of “Sport-screme,” the proximity of the marks, the[*212] sophistication of topical analgesic consumers, or Pfizer’s intent in selecting its mark. This appeal followed.

II.

The touchstone of our analysis is the Lanham Trade-Mark Act. Animated by economic and marketing realities, the Act’s framers believed that promoting the distin-.guishability of goods would safeguard the public, as well as stimulate competition. ^[3] Section 43(a) of the Act, ^[4] relevant portions of which are printed in the margin, protects against false designation of origin and false description or representation, whether or not a registered trademark is involved. ^[5]

This seemingly straightforward proscription has spawned a weighty, indeed un-wieldly, body of case law, obsessed more with minute calibrations than clear legal principles. To a large extent, of course, this result is to be expected. Razor-thin judgment calls are indigenous to the law of trademark protection.

In determining whether an existing unregistered mark should be protected against the introduction of a “similar” mark by a junior user, we have traditionally employed a variety of approaches, including the mark’s classification; its likelihood of confusion; and the “second-comer” doctrine. These standards, however, suffer from a variety of analytical infirmities that limit their ability to resolve disputes satisfactorily.

Trademark Classification:

Determining A Mark’s Eligibility For Protection

One mode of analysis in trademark infringement cases examines a mark’s eligibility for trademark protection. Writing for the court in Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir.1976), Judge Friendly set forth what quickly became the classic formulation of this principle. He noted that there exist four categories of terms, each one conferring a differing degree of eligibility for trademark protection. “Arrayed in an ascending order which roughly reflects their eligibility to trademark status and the degree of protection accorded, these classes are (1) generic, (2) descriptive, (3) suggestive, and (4) arbitrary or fanciful.” Id. at 9.

Generic terms — “the common descriptive name of an article or substance” ^[6] — are ineligible for any trademark protection. Descriptive marks, those “describ[ing] the qualities, ingredients, effects or other features of the product naturally and in ordinary language” ^[7] are protectible only[*213] where secondary meaning can be established. Suggestive terms require “imagination, thought and perception to reach a conclusion as to the nature of goods,” see Stix Products, Inc. v. United Merchants & Manufacturers, Inc., 295 F.Supp. 479, 488 (S.D.N.Y.1968), and are eligible for protection without proof of secondary meaning. Finally, arbitrary or fanciful marks “enjoy all the rights accorded to suggestive terms as marks — without the need of debating whether the term is ‘merely descriptive’ and with ease of establishing infringement.” Abercrombie & Fitch, supra, 537 F.2d at 11.

More than sixty years ago, Judge Learned Hand, reacting to the amorphousness of these standards, stated, “It is quite impossible to get any rule out of the cases beyond this: That the validity of the mark ends where suggestion ends and description begins.” Franklin Knitting Mills, Inc. v. Fashionit Sweater Mills, Inc., 297 F. 247, 248 (2d Cir.1923), aff’d, 4 F.2d 1018 (2d Cir.1925) (per curiam). It cannot be gainsaid that the judiciary is ill-equipped to distinguish between the descriptively suggestive and the suggestively descriptive mark. ^[8] In addition, societal vicissitudes demand that the categories retain fluidity to accommodate a particular mark’s evolving usage over time. See, e.g., King Seeley Thermos Corp., supra, (“thermos,” once suggestive, was held to be generic); Haughton Elevator Co. v. Seeberger, 85 U.S.P.Q. 80 (1950) (“escalator,” originally fanciful, was held to have become generic). Moreover, the determination whether a mark is descriptive or suggestive cannot be made in a vacuum; it is necessary to surmise the mental processes of those in the marketplace at whom the mark is directed.

The classification of a particular term represents merely a threshold determination. After a mark has been duly catalogued and its eligibility for protection has been determined, we must then examine a litany of factors to evaluate whether consumers are likely to be confused as to the source of the product. ^[9]

Likelihood of Confusion

The ultimate inquiry in most actions for false designation of origin, as with actions for trademark infringement, ^[10] is whether there exists a “likelihood that an appreciable number of ordinarily prudent purchasers [will] be misled, or indeed simply confused, as to the source of the goods in question.” Mushroom Makers, Inc. v. R. G. Barry Corp., 580 F.2d 44, 47 (2d Cir.1978), cert. denied, 439 U.S. 1116, 99 S.Ct. 1022, 59 L.Ed.2d 75 (1979). Writing for the court in Polaroid Corp. v. Polarad Elects. Corp., 287 F.2d 492, 495 (2d Cir.), cert. denied, 368 U.S. 820, 82 S.Ct. 36, 7 L.Ed.2d 25 (1961), Judge Friendly noted that such an assessment properly turns on the examination of many factors:

[The] strength of the mark, the degree of similarity between the marks, the proximity of the products, the likelihood that the prior owner will bridge the gap, actual confusion, and the reciprocal of defendant’s good faith in adopting its own mark, the quality of defendant’s product, and the sophistication of the buyers. Even this extensive catalogue does not[*214] exhaust the possibilities — the court may have to take still other variables into account.

Id. at 495. Accord Lever Bros. Co. v. American Bakeries Co., 693 F.2d 251 (2d Cir.1982); Vitarroz Corp. v. Borden, Inc., 644 F.2d 960 (2d Cir.1981). The Polaroid court’s prescience has been borne out, as the catalogue of original Polaroid factors has been expanded. See, e.g., McGregor-Doniger Inc. v. Drizzle Inc., 599 F.2d 1126, 1140 (2d Cir.1979) (“this Court has frequently supplemented its consideration of the Polaroid factors by balancing the conflicting interests of the parties involved”); Chandon Champagne Corp. v. San Marino Wine Corp., 335 F.2d 531, 536 (2d Cir.1964) (considering, in addition to the original Polaroid factors, the nature of the senior user’s priority, the senior user’s delay in asserting its claim, and the harm to the junior user as compared to the benefit to the senior user that would result from the requested injunction).

Although the Polaroid balancing test was traditionally confined to products that were “non-competing,” see McGregor-Do-niger Inc., supra, 599 F.2d at 1139, “noncompetitive,” see Scarves By Vera, Inc. v. Todo Imports, Ltd. (Inc.), 544 F.2d 1167, 1173 (2d Cir.1976), or “different,” see Mushroom Makers, Inc., supra, 580 F.2d at 47, it now properly extends to competing products. See Plus Products, supra, 722 F.2d at 1004 n. 6; American International Group, Inc. v. London American International Corp., 664 F.2d 348, 351 n. 2 (2d Cir.1982); Vitarroz, supra, 644 F.2d at 965-67; Kiki Undies Corp. v. Promenade Hosiery Mills, Inc., 411 F.2d 1097,1099 (2d Cir.1969).

The wisdom of the likelihood of confusion test lies in its recognition that each trademark infringement case presents its own unique set of facts. Indeed, the complexities attendant to an accurate assessment of likelihood of confusion require that the entire panoply of elements constituting the relevant factual landscape be comprehensively examined. No single Polaroid factor is pre-eminent, nor can the presence or absence of one without analysis of the others, determine the outcome of an infringement suit.

For example, if the determination of a mark’s strength were to stand as the centerpiece of the Polaroid analysis, with the remaining factors relegated to positions of lesser significance, weak marks would automatically be found unlikely to cause confusion among purchasers as to source; strong marks would invariably be deemed to cause confusion. The likelihood of confusion test would then merely become a variation of the “trademark classification” approach.

The Second-Comer Doctrine

The final mode of analysis traditionally employed in trademark infringement cases is the “second-comer” doctrine, an adjunct of trade piracy. As Learned Hand cogently noted:

“Why it should have chosen a mark that had long been employed by [plaintiff] and had become known to the trade instead of adopting some other means to identify [their] goods is hard to see unless there was a deliberate purpose to obtain some advantage from the trade which [plaintiff] had built up.” Indeed, it is generally true that, as soon as we see that a second comer in a mark has, for no reason that he can assign, plagiarized the “make-up” of an earlier comer, we need no more____[W]e feel bound to compel him to exercise his ingenuity in quarters further afield.

American Chicle Co. v. Topps Chewing Gum, Inc., 208 F.2d 560, 562-63 (2d Cir.1953), quoting Miles Shoes Inc. v. R.H. Macy and Co., 199 F.2d 602, 603 (2d Cir.1952).

A senior user in possession of a distinctive mark has a right not to have a second comer intentionally cause a likelihood of confusion between two marks in an attempt to exploit the reputation of the senior user’s mark, since this would deprive the first user of control over its reputation and goodwill. See Spring Mills, [*215] Inc. v. Ultracashmere House, Ltd., 689 F.2d 1127, 1135 (2d Cir.1982).

This test has resisted doctrinal classification. Some courts have simply incorporated the second-comer doctrine into the Polaroid analysis. See, e.g., Spring Mills, Inc., supra. But because the two elements necessary to trigger the second-comer doctrine — a strong mark and bad intent — are themselves included in the Polaroid analysis, the doctrine would appear to be a redundancy, unless it retains independent significance.

This seemingly superfluous approach may be an outgrowth of Polaroid’s recent expansion into the realm of competing products. Prior to Vitarroz and its progeny, there did not exist a systematic mode of analysis capable of incorporating a second-comer’s bad intent and the strength of the senior user’s mark in trademark infringement cases involving competing products. As a result, the second-comer doctrine offered an avenue of relief for the senior user. But with Polaroid’s expansive coverage, this is no longer the case. Consequently, courts that purport to apply the second-comer “test” should recognize that they are merely emphasizing two of the Polaroid factors.

In one sense, however, this doctrine remains fully extant. A second-comer may violate Section 43(a) by falsely representing his goods as those of the trademark owner. See Sutton Cosmetics (P.R.) Inc. v. Lander Co., Inc., 455 F.2d 285, 287-88 (2d Cir.1972). A claim of “passing off” may be established even if the goods are not confusingly similar; the wrong lies in the misrepresentation of the source. See Warner Bros. v. American Broadcasting Co., 720 F.2d 231, 247 (2d Cir.1983).

III.

Against this backdrop, and in light of the Lanham Act’s avowed purpose of ensuring consistent and uniform legal rights and remedies in a national economy, ^[11] we shall now attempt to clarify the law in the area of trademark infringement.

The starting point of our examination is determining whether a mark is eligible for protection. In making such an evaluation, we rely upon the classifications clearly set forth in Abercrombie & Fitch, Co., supra. Recognizing that our initial categorization may determine a mark’s eligibility for protection, and aware that the distinctions between generic and descriptive, ^[12] and descriptive and suggestive are often illusory, we believe a new emphasis is necessary.

The link between secondary meaning and likelihood of confusion is critical, yet it has often been overlooked by courts. Section 43(a) is meant to protect the public from confusion regarding the source of goods produced. ^[13] If a mark has secondary meaning, a purchaser will associate it with a certain producer, and will be likely to make that same association when an identical mark (or a confusingly similar[*216] mark), is used on another producer’s product. In the absence of secondary meaning, however, there will be no such association in the minds of the purchasing public. ^[14] Consonance with the avowed purpose of Section 43(a) of the Lanham Act, then, demands an examination of secondary meaning to determine whether an unregistered mark is eligible for protection.

If an unregistered mark is deemed descriptive, proof of secondary meaning is required for the mark to be protectible. See 20th Century Wear, Inc. v. Sanmark-Stardust Inc., 747 F.2d 81, 87, 90 n. 10 (2d Cir.1984). Suggestive marks are eligible for protection without any proof of secondary meaning, since the connection between the mark and the source is presumed.

Where a mark is ineligible for protection, there is no need to examine likelihood of confusion. Such an inquiry would, of course, be redundant, since a determination that a mark is ineligible for protection establishes that consumers do not associate that mark with a particular source. The rationales for denying protection to marks that do not connote a source of origin include: (1) the public’s inherent “right” to call a product by its name, see Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 119, 59 S.Ct. 109, 113, 83 L.Ed. 73 (1938); (2) exclusive appropriation of a generic/descriptive mark is “unfair” to competitors, see King Seeley Thermos Co., supra, 321 F.2d at 581; and (3) such an appropriation creates or entrenches a monopoly, see Canal Co. v. Clark, 80 U.S. (13 Wall.) 311, 324, 20 L.Ed. 581 (1871); Aloe Creme Laboratories, Inc. v. Milsan, Inc., 423 F.2d 845, 849 (5th Cir.), cert. denied, 398 U.S. 928, 90 S.Ct. 1818, 26 L.Ed.2d 90 (1970).

It is against this landscape of trademark law that we turn to the precise issue presented on appeal — whether the district judge properly granted Thompson’s motion for a preliminary injunction.

IV.

Our inquiry commences with a determination whether Thompson’s “Sportscreme” is eligible for protection. Although the district judge failed to address this issue, an “appellate court is in as good a position as the trial judge” to make such an evaluation. See McGregor-Doniger, Inc., supra, 599 F.2d at 1333. We hold that “Sport-screme” is a descriptive mark and, as such, is eligible only if Thompson can establish that it has acquired secondary meaning.

Thompson asserts that an imaginative link is necessary to connect the concept of a topical analgesic used in connection with sports activities to the name “Sportscreme.” We disagree. No exercise of the imagination is necessary for the public to understand that the product is a cream useful in connection with sports. ^[15] Marks that describe the use to which a product is put are descriptive. See 3 R. Callmann, The Law of Unfair Competition, Trademarks and Monopolies § 18.05, at 27 (L.Altman 4th ed. 1983).

The foregoing analysis comports with the rationale for distinguishing between suggestive and descriptive terms, and the purpose underlying the protection of trademarks. In protecting only those descriptive terms that have acquired secondary meaning, we recognize that potential market entrants should not be foreclosed from using descriptive terms to label their products. This concern, however, is inapposite where suggestive terms are involved. As Judge Lumbard persuasively noted:

The English language has a wealth of synonyms and related words with which[*217] to describe the qualities which manufacturers may wish to claim for their products and the ingenuity of the public relations profession supplies new words and slogans as they are needed.

Aluminum Fabricating Co. of Pittsburgh v. Season-All Window Corp., 259 F.2d 314, 317 (2d Cir.1958).

Thompson points to the availability of alternative names to describe a topical analgesic designed to relieve the aches and pains of sports participants, and argues that this militates in favor of protection. But this assertion misapprehends the relevant concern. The trademark law should not grant, in effect, a monopoly to the first mark that effectively and concisely describes a product’s use or function. Were this exclusive appropriation to occur, future entrants would be required to adopt a “less-descriptive” term, and engage in increased advertising to recoup the lost consumer appeal. Entry barriers would be created, discouraging entry and competition, particularly from small firms. This result is expressly at odds with the purposes of the trademark laws.

Secondary Meaning

Having concluded that “Sportscreme” is a descriptive mark, we must now determine whether it has acquired secondary meaning. As we have noted, “Sportscreme” deserves protection only if it can be established that “consumers have come to associate [it] with a particular manufacturer or source.” Perfect Fit Industries, Inc., supra, 618 F.2d at 953. If, however, secondary meaning cannot be established, “Sport-screme” will be ineligible for protection, and our analysis will conclude without having to address the likelihood of confusion. Absent secondary meaning, purchasers will not associate “Sportscreme” with a particular source of origin. By definition, there could not be likelihood of confusion as to source.

“[P]roof of secondary meaning entails vigorous evidentiary requirements,” Ralston Purina Co. v. Thomas J. Lipton, Inc., 341 F.Supp. 129, 134 (S.D.N.Y.1972), with Thompson bearing the burden of establishing that “Sportscreme” had acquired secondary meaning by the spring of 1984. See Saratoga Vichy Spring Co., Inc. v. Lehman, 625 F.2d 1037, 1042 (2d Cir.1980). In determining whether a mark has acquired secondary meaning, we have examined: advertising expenditures, see Harlequin Enterprises Ltd. v. Gulf & Western Corp., 644 F.2d 946, 949 (2d Cir.1981); McGregor-Doniger Inc., supra, 599 F.2d at 1132, 1133 n. 4; consumer studies linking the name to a source, see Harlequin Enterprises Ltd., supra, 644 F.2d at 950 n. 2; American Footwear Corp. v. General Footwear Co., 609 F.2d 655, 663 (2d Cir.1979); Grotrian, Helfferich, Shulz, Th. Steinweg Nachf. v. Steinway & Sons, 523 F.2d 1331, 1340-41 (2d Cir.1975); sales success, see Harlequin Enterprises Ltd., supra, 644 F.2d at 949 n. 1; unsolicited media coverage of the product, see id. at 950; Scarves By Vera, Inc., supra, 544 F.2d at 1174 (2d Cir.1976); attempts to plagiarize the mark, see Harlequin Enterprises Ltd., supra, 644 F.2d at 950; Perfect Fit Industries, Inc., supra, 618 F.2d at 954-55; and length and exclusivity of the mark’s use, see Saratoga Vichy Spring Co. v. Lehman, 491 F.Supp. 141, 150 (N.D.N.Y.1979), aff’d, 625 F.2d 1037 (2d Cir.1980); American Footwear Corp., supra, 609 F.2d at 663. In assessing the existence of secondary meaning, no “single factor is determinative,” id., and every element need not be proved. Each case, therefore, must be resolved by reference to the relevant factual calculus.

Although Thompson offered proof of secondary meaning, at trial — including its advertising expenditures, as well as increases in its sales revenues and market share, Judge Griesa made no finding on this issue. Accordingly, we remand to the district court so it may fully consider all evidence relevant to the issue of secondary meaning. If the district court rules that “Sportscreme” has not acquired secondary meaning, the mark cannot be protected and the inquiry properly concludes. If, on the other hand, the court below finds that[*218] “Sportscreme” has acquired secondary meaning, it may proceed to consider whether Thompson has proved likelihood of confusion as to source and, if so, trademark infringement.

Likelihood of Confusion

Although the court below found a likelihood of confusion as to source between “Sportscreme” and “BEN-GAY Sports-Gel,” it did so based solely on the similarity of the marks and Thompson’s priority of use. Such a conclusion represents a radical departure from this Court’s unwavering adherence to the principles set forth in Polaroid and expanded by its progeny. As we cautioned in Lever Bros. Co., supra, 693 F.2d at 253, “[n]o single ... factor is preeminent, nor can the presence or absence of one determine, without analysis of the others, the outcome of the infringement suit.”

Because Judge Griesa did not address the strength of Thompson’s mark, the proximity of the marks, the sophistication of topical analgesic consumers, Pfizer’s intent in selecting “SportsGel” as its mark, the chance that Thompson would bridge the gap into gel products and Thompson’s delay in filing this infringement action, we believe his finding of likelihood of confusion was made precipitously.

Furthermore, we feel impelled to comment upon the two findings that were made by the district judge — the similarity of the marks and Thompson’s priority of use. In finding the marks to be confusingly similar, the district court contrasted “Sportscreme” with “SportsGel,” emphasizing their bisyllabic nature, as well as their identity of prefix. In so doing, Judge Griesa examined only a “fragment” of Pfizer’s mark, instead of considering that “SportsGel” is linked prominently with the BEN-GAY line of products. See Universal City Studios, Inc. v. Nintendo Co., Ltd., 746 F.2d 112, 117 (2d Cir.1984); Vi-tarroz, supra, 644 F.2d at 968-69; McGre-gor-Doniger, supra, 599 F.2d at 1133-34; Beech-Nut, Inc. v. Warner-Lambert, Co., 480 F.2d 801, 804 (2d Cir.1973). We believe a comparison of each mark in its entirety minimizes the likelihood of consumer confusion as to source.

Moreover, the trial judge held that Pfizer’s duty as a second-comer to the sports rub market required it to select a mark that entirely avoided any possibility of confusion with Thompson’s “Sportscreme.” Pfizer’s failure to do so, according to the district judge, buttressed the finding of trademark infringement. But, in so ruling, the court below misapplied the second-comer doctrine, which we have invoked only where there exists a highly distinctive mark of the senior user and proof of bad faith by the junior user. ^[16] See Spring Mills, Inc., supra, 698 F.2d at 1135-36. Neither requisite element was found by the district judge. Indeed, his conclusion that “the second-comer who knows about the name and package design and so forth of the first-comer has some obligation to name his product and design his packaging so as to avoid confusion,” would, if applied, confer exclusive trademark protection to the first user of a mark, regardless of the mark’s distinctiveness. Such a result would be inimical to the purposes of trademark law.

V.

In light of our views set forth in Section III, we find “Sportscreme” to be a descriptive mark that is protected only if it has acquired secondary meaning. That issue must be determined by the trial court.. If the court finds that “Sportscreme” has not acquired secondary meaning, the mark is not protected and no further inquiry is necessary. Only if the district court rules that “Sportscreme” has acquired secondary meaning must it comprehensively examine the Polaroid factors to determine whether[*219] there exists a likelihood of confusion as to source.

Accordingly, we vacate the order granting the preliminary injunction, and remand to the district court for further proceedings in accordance with this opinion.