Centaur Communications, Limited v. A/s/m Communications, Inc.

*1219CARDAMONE, Circuit Judge:

This appeal, from a finding of trademark infringement, necessarily implicates the legal concepts of secondary meaning and likelihood of confusion, for which our precedents establish a long list of factors that must be considered before a determination may be reached. Unfortunately, there is no shortcut. To reach a principled conclusion in a trademark case, it is just as essential to recite the right formulas as it was for Ali Baba to say “Open Sesame” in order to open the door to the treasure cave of the Forty Thieves.

In a judgment of the United States District Court for the Southern District of New York (Pollack, J.) entered on January 20,1987, the district court judge found that the plaintiff-appellee Centaur Communications, Ltd. (Centaur or appellee) had established that the phrase, “Marketing Week,” the title of a Centaur publication, was a protectible trademark under both federal and state law. The trial judge also ruled that the defendant-appellant A/S/M Communications, Inc. (A/S/M or appellant) had infringed upon the mark because the title of its publication, “ADWEEK’s Marketing Week,” would be likely to create confusion among consumers as to the source or origin of A/S/M’s magazine. Based on these findings, it granted injunctive relief. We subsequently stayed a portion of the injunction pending appeal. We now affirm.

BACKGROUND

This appeal involves two weekly publications with similar titles, both of which focus on marketing news. For purposes of clarity, we will refer to Centaur’s publication as Marketing Week and A/S/M’s as ADWEEK’s Marketing Week.1 Although both magazines are concerned with marketing, the emphasis of each is different. Marketing Week concentrates on the British market, though it does have a page devoted to marketing news in the United States, and ADWEEK’s Marketing Week is geared towards the American market.

The orbits of these two magazines recently converged. Marketing Week has been distributed in the United States for about ten years, garnering only a slim circulation of 110 current subscribers. In contrast, its subscriber-base in the United Kingdom numbers 36,000. The genesis of ADWEEK’s Marketing Week is somewhat complicated. A/S/M was formed in 1978. At that time it acquired three regional publications dealing with advertising, and changed their names to ADWEEK. Over the next several years, A/S/M acquired other publishing companies in different regions that also published advertising magazines. By 1986 it was publishing six regional editions of ADWEEK, all aimed towards advertising agency executives. In mid-1985, A/S/M began publishing a national marketing edition of ADWEEK. The national publication differed from the regional editions in that it emphasized attracting marketing executives in so-called “client” companies. The national marketing edition of ADWEEK had a subscription base of 10,000. The title of this new publication was ADWEEK National Marketing Edition. Beginning with the September 6, 1986 edition, A/S/M changed the title to ADWEEK’s Marketing Week. This new publication came to Centaur’s attention shortly thereafter, and the present litigation ensued.

This case commenced when Centaur sued A/S/M for declaratory and injunctive relief on December 15, 1986. The complaint alleged, inter alia, infringement of Centaur’s unregistered trademark and a violation of New York’s unfair competition law. After a December 23, 1986 hearing before the district court, the request for preliminary injunctive relief was denied. Centaur Communications, Ltd. v. A/S/M Communications, Inc., 649 F.Supp. 74 (S.D.N.Y.1986).

*1220Waiving damages, Centaur then sought permanent injunctive relief. After a bench trial on the merits, the district court found both a Lanham Act and an unfair competition violation. Centaur Communications, Ltd. v. A/S/M Communications, Inc., 652 F.Supp. 1105 (S.D.N.Y.1987). As a consequence, it restrained A/S/M from displaying on its cover or in any promotional material, the mark “Marketing Week” either alone or with the word “ADWEEK’s” “occupying a significantly and materially lesser percentage of the entire logo.” Id. at 1125. It further enjoined A/S/M from using the title “Marketing Week” without a license from Centaur. Id. at 1124. Appellant was also required to publish a prominent notice in its next three issues explaining why it had stopped using the title Marketing Week” and stating that its publication was not connected to Centaur’s. Id. at 1126. Finally, the district court awarded appellee attorneys’ fees. Id.

On January 27, 1987, A/S/M moved in this Court for an order expediting the appeal and for a stay pending appeal. We expedited the appeal and stayed that portion of the decree which required that “Ad-week” appear in comparable size lettering to and in conjunction with “Marketing Week” on its cover. Subject to certain conditions, A/S/M was permitted to utilize advertising and promotional materials without the word “Adweek” appearing in comparable size to and in connection with “Marketing Week.” We also stayed the requirement that appellant give notice to its subscribers. Finally, the sum of $10,-000 was set as liquidated damages should the finding of infringement be affirmed on appeal.

The main issue raised is the correctness of the district court’s determination that by its use of the term “Marketing Week” A/S/M violated § 43(a) of the Lanham Act.

DISCUSSION

1 The Lanham Act Claim

Section 43(a) of the Lanham Act proscribes “false designation of origin” in relation to goods or services. 15 U.S.C. § 1125(a) (1982). This section is the only provision in the Lanham Act that protects an unregistered mark like Centaur’s. Its purpose is to prevent consumer confusion regarding a product’s source, see Lois Sportswear, U.S.A., Inc. v. Levi Strauss & Co., 799 F.2d 867, 871 (2d Cir.1986), and to enable those that fashion a product to differentiate it from others on the market. International Order of Job’s Daughters v. Lindeburg & Co., 633 F.2d 912, 918 (9th Cir.1980), cert. denied, 452 U.S. 941, 101 S.Ct. 3086, 69 L.Ed.2d 956 (1981). In that way producers create goodwill with consumers. The device used to protect both groups is a trademark.

Not all trademarks are entitled to the same degree of legal protection. See, e.g., Lois Sportswear, 799 F.2d at 871; 20th Century Wear, Inc. v. Sanmark-Stardust Inc., 747 F.2d 81, 87 (2d Cir.1984), cert. denied, 470 U.S. 1052, 105 S.Ct. 1755, 84 L.Ed.2d 818 (1985) (20th Century I). As Judge Friendly’s opinion in Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir.1976), observes, there are four shifting levels of trademark protection which, beginning with the least protected, are (1) generic, (2) descriptive, (3) suggestive, and (4) fanciful or arbitrary. Id. at 9.

The two categories implicated on this appeal are generic and descriptive marks. A generic mark “refers, or has come to be understood as referring, to the genus of which the particular product is a species.” Id. at 9. For example, “aspirin” is a generic term. A descriptive mark “describes the qualities or characteristics of a good or service.” Park ’N Fly, Inc. v. Dollar Park and Fly, Inc., 469 U.S. 189, 194, 105 S.Ct. 658, 662, 83 L.Ed.2d 582 (1985). A term can be descriptive in two ways: “[i]t can literally describe the product, or it can describe the purpose or utility of the product.” 20th Century I, 747 F.2d at 88. The district court ruled without objection that the mark, “Marketing Week,” was descriptive. 652 F.Supp. at 1108.

Appellant now suggests that “Marketing Week” is a generic mark. Since this argument was not raised in the trial court, we decline to consider it. See Air et Chaleur, S.A. v. Janeway, 757 F.2d 489, 493 (2d *1221Cir.1985); Schmidt v. Polish People’s Republic, 742 F.2d 67, 70 (2d Cir.1984). Moreover, neither party has addressed the issue of whether “marketing” is the name of a trade or industry, which were that the case, would cause the mark to be generic. Cf. Reese Publishing Co. v. Hampton Int’l Communications, 620 F.2d 7, 10 (2d Cir. 1980) (the title, “Consumer Electronics Monthly” was a generic trademark because “consumer electronics” was the name of a trade). We consider therefore what protection, if any, should be accorded appellee’s descriptive mark.

Because the title, “Marketing Week,” is not registered, Centaur may only succeed in this infringement suit if it proves that (1) the mark has acquired secondary meaning and (2) there is a likelihood of confusion as to the source of the publication. See Thompson Medical Co. v. Pfizer Inc., 753 F.2d 208, 213, 216 (2d Cir.1985); 20th Century I, 747 F.2d at 90. This two-step analysis reflects two different questions. The inquiry into the existence of secondary meaning examines whether the proponent of the mark has acquired a proteetible interest in it. See American Television & Communications Corp. v. American Communications & Television, Inc., 810 F.2d 1546, 1548 (11th Cir.1987); Security Center, Ltd. v. First Nat’l Sec. Centers, 750 F.2d 1295, 1298 (5th Cir.1985). An examination into the likelihood of confusion considers whether that interest has been infringed. See Lois Sportswear, 799 F.2d at 871. We discuss each of these questions.

A. Secondary Meaning

A mark acquires secondary meaning when “it [is] shown that the primary significance of the term in the minds of the consuming public is not the product but the producer.” 20th Century Wear, Inc. v. Sanmark-Stardust Inc., 815 F.2d 8, 10 (2d Cir.1987) (20th Century II) (quoting Ralston Purina Co. v. Thomas J. Lipton, Inc., 341 F.Supp. 129, 133 (S.D.N.Y.1972) (emphasis in original)). Thus, the crux of the doctrine of secondary meaning “is that the mark comes to identify not only the goods but the source of those goods,” even though the relevant consuming public might not know the name of the producer. Id. (“[Consumers often buy goods without knowing the personal identity or actual name of the [producer].”). Nonetheless, someone seeking to establish secondary meaning must show that the purchasing public associates goods designated by a particular mark with but a single — although anonymous — source. Id.; cf. 3 R. Callmann, The Law of Unfair Competition Trademarks and Monopolies § 19.25, at 81-82 n. 7 (L. Altman 4th ed. 1983) (Trademarks) (“It suffices if consumers accept the trademark as an indication of origin.”).

The focus of secondary meaning therefore is the consuming public. Hence, so must be the inquiry into whether the mark owner has a proteetible interest. Although the mark owner strives to create a secondary meaning for its product, it is the consuming public which, in effect, determines whether that effort has succeeded. See Co-Rect Prods., Inc. v. Marvy! Advertising Photography, Inc., 780 F.2d 1324, 1330 (8th Cir.1985); Transgo, Inc. v. Ajac Transmission Parts Corp., 768 F.2d 1001, 1015 (9th Cir.1985), cert. denied, 474 U.S. 1059, 106 S.Ct. 802, 88 L.Ed.2d 778 (1986); Scott Paper Co. v. Scott’s Liquid Gold, Inc., 589 F.2d 1225, 1228 (3d Cir.1978). See also Madrigal Audio Laboratories, Inc. v. Cello, Ltd., 799 F.2d 814, 822 (2d Cir.1986) (same but involving trade name).

Again, it is not always the general public’s understanding but — depending upon the product — often only a segment of consumers that need be examined. See Playboy Enters. v. Chuckleberry Publishing, Inc., 687 F.2d 563, 567 (2d Cir.1982) (the litigants’ magazines appealed to the same consumer group — heterosexual adult males); President & Trustees of Colby College v. Colby College-N.H., 508 F.2d 804, 808 (1st Cir.1975) (college could establish secondary meaning by showing “the congruence of its name and its institution in the minds of an appreciable number of individuals broadly associated with other institutions of higher education in a given geographic area.”); Inc. Publishing Corp. v. Manhattan Magazine, Inc., 616 F.Supp. *1222370, 386 (S.D.N.Y.1985), aff'd mem., 788 F.2d 3 (2d Cir.1986) (relevant consumer group in trademark case involving magazine titles comprised of advertisers, subscribers, and newsstand purchasers). Such is the case at hand.

Moreover, it is only necessary to show that a substantial segment of the relevant group of consumers made the requisite association between the product and the producer. See Levi Strauss & Co. v. Blue Bell, Inc., 778 F.2d 1352, 1358 (9th Cir.1985) (en banc); Union Carbide Corp. v. Ever-Ready Inc., 531 F.2d 366, 381 (7th Cir.), cert. denied, 429 U.S. 830, 97 S.Ct. 91, 50 L.Ed.2d 94 (1976); 3 R. Callmann, Trademarks, supra, § 19.26, at 86. The district court here found that the relevant consuming public (i.e., market) of the magazines was “executives in the international marketing and advertising community in the United States.” 652 F.Supp. at 1110. Because neither party challenges this definition of the group, it is the starting point for analysis.

In Thompson Medical we collected the various precedents that discussed the question of secondary meaning, 753 F.2d at 217, and drew from them the elements utilized in determining whether secondary meaning had been created, noting that “no ‘single factor [was] determinative,’ and [that] every element need not be proved.” Id. (quoting American Footwear Corp. v. General Footwear Co., 609 F.2d 655, 663 (2d Cir.1979)). The elements are (1) advertising expenditures, (2) consumer studies linking the mark to a source, (3) unsolicited media coverage of the product, (4) sales success, (5) attempts to plagiarize the mark, and, (6) length and exclusivity of the mark’s use. Id. After applying these criteria to its definition of the relevant market the district court found that the mark “Marketing Week” had acquired a secondary meaning. Our first task is to analyze these factors in order to decide whether its findings were clearly erroneous. See McGregor-Doniger Inc. v. Drizzle Inc., 599 F.2d 1126, 1133 n. 4 (2d Cir.1979); see also American Television, 810 F.2d at 1549; Blue Bell, Inc., 778 F.2d at 1355; 3 R. Callmann, Trademarks, supra, § 19.27, at 91.

(1) Advertising Expenditures

The characteristics of the relevant market are important in considering advertising expenditures. Compared to other advertising campaigns, Centaur’s efforts were relatively modest. Cf. Aloe Creme Laboratories, Inc. v. Milsan, Inc., 423 F.2d 845, 850 (5th Cir.), cert. denied, 398 U.S. 928, 90 S.Ct. 1818, 26 L.Ed.2d 90 (1970) (expenditure of 3 million dollars for nationwide magazine and newspaper advertisements). It expended $10,000 in a direct mail solicitation campaign targeted at 26,-000 subscribers of Advertising Age.

Nonetheless, the record discloses that Centaur publicized Marketing Week by other means. For the past three years it has mailed brochures to the top 100 American advertising agencies. Its senior director has traveled to the United States twice a year for the last eight years to make sales presentations to advertising agencies and the media. Concededly, these trips were undertaken for the purpose of selling advertising space, but they were in effect advertising expenditures because they served to put Marketing Week before the relevant group of consumers. Similarly, Marketing Week’s appearance in a source guide to European marketing and advertising, entitled Marketing Europe 1985, is also the functional equivalent of advertising because this guide was distributed to people involved in advertising and marketing. Moreover, Centaur produces a video guide to advertising agencies in England, many of which are overseas offices of American agencies. Centaur has also sponsored various conferences featuring speakers from American advertising agencies and companies under the “Marketing Week” title. And, it has generated in recent years up to $250,000 in, revenues from U.S. advertisers buying space in Marketing Week.

Thus, these activities support the conclusion that Centaur’s efforts were effective in causing the relevant group of consumers to associate Marketing Week with it. See, *1223e.g., First Brands Corp. v. Fred Meyer, Inc., 809 F.2d 1378, 1383 (9th Cir.1987) (test of secondary meaning is not the size of the expenditure used to create it but its effectiveness).

(2) Consumer Studies Linking Mark to Source

The most vigorously disputed aspect of this case is the evidence of consumer associations linking the mark to appellee. The dispute encompasses the district court’s reliance on “anecdotal” testimony by Centaur’s witnesses and its treatment of A/S/M’s survey. The trial court credited testimony of the head of another trade magazine, Advertising Age, Mr. Ranee Crain, and of the chief executive officer of a large advertising agency, Mr. Edward Ney, that Marketing Week had substantial recognition and acceptance in the relevant market.

Appellant mounts several challenges to the use of this evidence. First, it attacks reliance on anecdotal testimony by consumers instead of use of a consumer survey. This attack is misplaced. We have never rejected use of direct testimony by consumers. Cf. 20th Century II, 815 F.2d at 10 (in passing on the question of secondary meaning, the court noted the absence of testimony from consumers or a consumer survey). And, though surveys have become the usual way of demonstrating secondary meaning, see Mattel, Inc. v. Azrak-Hamway Int’l, Inc., 724 F.2d 357, 361 (2d Cir.1983) (per curiam), they are not the only way. A/S/M also attacks the credibility of Mr. Crain, arguing that he is biased because his magazine competes with AD-WEEK’s Marketing Week. The trial court was in a better position to determine credibility, particularly in light of appellant's thorough investigation of the alleged bias. Hence, the testimony cannot be discounted on this basis. Cf. Playboy Enterprises, 687 F.2d at 569 (appellate court, lacking trial court’s opportunity to observe the witness, credited trial court’s characterization of witness’s testimony).

Yet the testimony of the two witnesses is not entitled to much weight. For example, we have previously considered as probative testimony from a witness that a product had “a small and declining share of the market and that ‘on a consumer level the brand was ... virtually unknown to the vast majority of people.’ ” Abraham Zion Corp. v. Lebow, 761 F.2d 93, 105 (2d Cir.1985). The justification for little weight is because the basis for Mr. Crain’s statement that Marketing Week was well known in “top marketing circles and advertising circles and in top international circles in the country” is not disclosed. Although the witness testified that he had questioned various people about Marketing Week, it is unclear whether they were representative of the relevant market. Again, the witness estimated that there are approximately 5,000 to 10,000 individuals in that market, but does not reveal whether he spoke to 5 or 50 of them. Similarly, the record does not indicate how representative Mr. Ney’s experience at his particular advertising agency was compared to other agencies. Cf. In re Northland Aluminum Prods., Inc., 777 F.2d 1556, 1560 (Fed.Cir.1985) (discounting testimony of individuals concerning secondary meaning of term because the record did not establish that they were representative of the relevant group of consumers). Thus, though the testimony of these witnesses is relevant, its significance is limited.

A/S/M conducted two surveys designed to test consumer association of the mark “Marketing Week” to Centaur. No consumer in either survey made the association. These surveys were admitted into evidence, but their significance was discounted by the district court, which stated that they were “of dubious value in the totality of the evidence.” 652 F.Supp. at 1110. Judge Pollack treated the survey evidence properly. A survey must use the right frame of reference, that is, the group whose associations or attitudes are at issue. See Universal City Studios, Inc. v. Nintendo Co., 746 F.2d 112, 118 (2d Cir.1984); see also Manhattan Magazine, 616 F.Supp. at 392-93. Although the above cases concerned flaws in a survey designed to test consumer confusion, there is no *1224reason not to apply the basic principle in the context of secondary meaning.

The surveys here, according to the testimony of Dr. Hans Zeisel, were designed to test the reactions of “people in executive positions in marketing in American business enterprises and institutions.” These people did not correspond to the relevant group — “executives in the international marketing and advertising community in the United States” — defined by the district court. Thus, the value of the surveys was correctly discounted, leaving direct evidence of consumer association of “Marketing Week” with Centaur as inconclusive.

(3) Sales Success

As previously mentioned, American subscriptions to Marketing Week are limited. In another context, such a small base might be fatal. But the paucity of subscribers in the instant case must be considered in light of two factors. First, some of the subscriptions are held by large advertising agencies and companies where the magazine is read by more than one person. Second, Centaur sold advertising to Business Week, the New York Times, Hilton International, and McGraw Hill. Because these advertising decisions typically involve advertising agencies and client companies, they suggest a recognition among the relevant group of consumers. In sum, the limited success of Centaur in attracting subscribers tends to undercut a finding of secondary meaning, but in light of the discussed considerations, it is not dispositive.

(4) Unsolicited Media Coverage

With respect to unsolicited media coverage, Marketing Week has been cited in such publications as Business Week and the Associated Press. Not much significance may be ascribed to such citations because it is not clear what they indicate about the relevant group of consumers. Cf. Harlequin Enters. v. Gulf & W. Corp., 644 F.2d 946, 950 (2d Cir.1981) (unsolicited media coverage reported on the “enthusiasm and loyalty” of the product’s customers).

(5) Attempts to Plagiarize

The most persuasive Thompson Medical factor is the finding that A/S/M had intentionally copied Centaur’s mark. See 20th Century II, 815 F.2d at 10 (finding of intentional copying was persuasive, if not conclusive, evidence of consumer recognition and goodwill). Appellant was aware of appellee’s publication. Its executive vice-president and principal stockholder had copies of Marketing Week in his office. Additionally, the parties previously had discussed a joint video venture, as well as Centaur’s possible acquisition of some shares in A/S/M.

It seems fairly clear that A/S/M deliberately copied Centaur’s mark. The most telling evidence of this is the obvious similarity of the titles and the fact that A/S/M failed to provide a credible explanation for the change that appeared in its September 6,1986 edition that did not involve copying. A/S/M’s previous title for its magazine was ADWEEK’s National Marketing Edition. But, as the A/S/M official stated at trial, the name “Adweek” was the most prominent feature in the title while the phrase “National Marketing Week” appeared “in a bar under the title.” (emphasis supplied). This testimony suggests that ADWEEK’s Marketing Week evolved from ADWEEK alone. In fact, the evolution from “Adweek” to “Adweek’s Marketing Week” is not so evident. A/S/M claimed in an affidavit submitted in opposition to the preliminary injunction that “Marketing” evolved from the phrase “National Marketing Edition.” Yet this claim is belied by its officer’s statement and by the continued presence — as of April 20, 1987 issue — of the phrase “National Marketing Edition” on the cover of appellant’s publication. Moreover, the trial court was better able than we are to observe the demeanor of appellant’s witnesses testifying to the genesis of the title and therefore to assess their credibility. See Schmitz v. St. Regis Paper Co., 811 F.2d 131, 132 (2d Cir.1987) (per curiam); Financial Information Inc. v. Moody’s Investors Serv., Inc., 808 F.2d 204, 208 (2d Cir.1986).

*1225(6) Length and Exclusivity

Another Thompson Medical factor that supports the finding of secondary meaning is the length and exclusivity of Centaur’s use of the mark. From 1978 Centaur had continuously and exclusively used “Marketing Week” as its mark until A/S/M changed the title of ADWEEK’s National Marketing Edition in September 1986. Of course, no absolute time span can be posited as a yardstick in cases involving secondary meaning. Cf. Ralston Purina Co. v. Thomas J. Lipton, Inc., 341 F.Supp. 129, 133-34 (S.D.N.Y.1972) (comparing cases involving different durations in the use of a mark); 3 R. Callmann, Trademarks, supra, § 19.27, at 89. Instead, the length and exclusivity of a mark’s use is evaluated in light of the product and its consumers. See 3 R. Callmann, Trademarks, supra, § 19.27, at 91. Thus, the relatively small size of the defined market, as well as the interconnections developing between American and British advertising and marketing communities resulting from acquisitions of advertising agencies and the establishment of branch offices, indicate that Centaur’s use of the mark is more significant in this context than it otherwise would have been.

To summarize: given the findings of intentional copying of the mark, its duration and exclusive use, and the relative significance of advertising expenditures, considered in the context of the relevant consumer group, we are unable to say that the district court was clearly erroneous in concluding that “Marketing Week” had achieved secondary meaning. Accordingly, we turn to the second part of the Lanham Act analysis to explore whether the use of “Marketing Week" by A/S/M was likely to confuse consumers as to the source of the product.

B. Likelihood of Confusion

Having determined that Centaur has a protectible property interest in its mark, we take up the question of whether A/S/M has infringed upon that right. In order to succeed on this sort of Lanham Act claim, a plaintiff must show a “likelihood that an appreciable number of ordinarily prudent purchasers are likely to 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) (per curiam), cert. denied, 439 U.S. 1116, 99 S.Ct. 1022, 59 L.Ed.2d 75 (1979). In reviewing the finding that A/S/M’s use of the “Marketing Week” mark presented a likelihood of confusion, it is well to remember that this issue is like a tangle of underbrush. Fortunately, a path has been hewn through this thicket.

In Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492 (2d Cir.), cert. denied, 368 U.S. 820, 82 S.Ct. 36, 7 L.Ed.2d 25 (1961), Judge Friendly set out a multifactor balancing test. The factors are: (1) the strength of the mark, (2) the degree of similarity between the two marks, (3) the proximity of the products, (4) the likelihood that the senior user of the mark will bridge the gap, (5) evidence of actual confusion, (6) the junior user’s bad faith vel non in adopting the mark, (7) the quality of the junior user’s product, and, finally, (8) the sophistication of the relevant consumer group. Id. at 495. At the outset, we note that the district court’s finding on each individual Polaroid factor is subject to the clearly erroneous standard, see Lois Sportswear, 799 F.2d at 873. But the ultimate determination of likelihood of confusion is a legal conclusion reviewable de novo. See Plus Prods. v. Plus Discount Foods, Inc., 722 F.2d 999, 1004-05 (2d Cir.1983). We turn then to the various Polaroid findings.

(1) Strength of the Mark

Addressing the first Polaroid factor, the district court found that “Marketing Week” had achieved a position of relative strength. 652 F.Supp. at 1112. The strength of a mark is its tendency to identify the goods sold as emanating from a particular source, even when the source is unknown to the consumer. McGregor-Doniger, 599 F.2d at 1131. As discussed above, the degree of the mark’s distinctiveness depends in part upon its type. See Plus Products, 722 F.2d at 1005. Thus, *1226“Marketing Week,” as a descriptive mark is, by definition, somewhat weak. Yet, a mark’s category is not alone controlling. See, e.g., id. (“Plus” a suggestive mark but nonetheless weak); Lever Bros. v. American Bakeries Co., 693 F.2d 251, 256-57 (2d Cir.1982) (“Autumn” an arbitrary mark but nonetheless weak). Rather, its strength should be examined in its commercial context. See McGregor-Doniger, 599 F.2d at 1133. For example, in considering the strength of a magazine trademark, time and size of circulation and consumer identification are examined. See C.L.A.S.S. Promotions, Inc. v. D.S. Magazines, Inc., 753 F.2d 14, 17-18 (2d Cir.1985).

Marketing Week has been in circulation a comparatively short time with modest sales. But, it circulates in a small market that does not require a large circulation base in order to acquire consumer identification. We think that when this is considered together with Centaur’s various promotional activities, consumers in that small market would associate the mark with its source. Thus, the trial court correctly concluded that “Marketing Week” has achieved relative — if not great— strength in its market context.

(2) Degree of Similarity

The second factor is the similarity of the two marks. Unlike the other factors, this finding is reviewed de novo because an appellate court is in as good a position to evaluate this factor as is a trial court. See Spring Mills, Inc. v. Ultracashmere House, Ltd., 689 F.2d 1127, 1130 (2d Cir.1982); McGregor-Doniger, 599 F.2d at 1133. Here, “the general impression conveyed to the purchasing public by the respective marks” is the pertinent inquiry. C.L.A.S.S. Promotions, 753 F.2d at 18. In C.L.A.S.S. Promotions, we affirmed a finding that the “size, layout, design, and logotype of the two titles” reduced the potential for confusion. Id. Judge Pollack implicitly considered these factors in making his determination that the two titles were quite similar. We see no reason to disturb this finding.

A/S/M contends that the presence of “Adweek” in its title makes the two titles dissimilar. The argument proves too little. The titles are not identical; yet given the size and logotype of “Adweek”, it does not occupy a significant place in the title. Thus, appellant has not shown that the titles are distinct. Hence, this Polaroid factor supports the conclusion that the two titles presented a potential for consumer confusion.

(3) Proximity

The third factor is the proximity of the products, perhaps more accurately described as “competitive proximity.” The concern is whether “it is likely that customers mistakenly will assume either that [the junior user’s goods] somehow are associated with [the senior user] or are made by [the senior user].” Lois Sportswear, 799 F.2d at 874. See also McGregor-Doniger, 599 F.2d at 1134 (“[T]he degree of proximity between the two products is relevant here primarily insofar as it bears on the likelihood that customers may be confused as to the source of the products, rather than as to the products themselves.”) (emphasis in original). Competitive proximity should be measured, in part, with reference to the first two Polaroid factors. See Mobil Oil Corp. v. Pegasus Petroleum Corp., 818 F.2d 254, 258 (2d Cir.1987).

A/S/M correctly points out that the different editorial emphasis — one publication is primarily concerned with marketing news in America and the other with marketing news in Britian — militates against a finding of proximity. Cf. C.L.A.S.S. Promotions, 753 F.2d at 18; Playboy Enterprises, 687 F.2d at 567. This does not end the inquiry however. Both magazines are high quality weekly publications concerned with marketing news. Thus, consumers in the market interested in American marketing news well might assume that Centaur had decided to launch a different magazine primarily concerned with that topic. See Lois Sportswear, 799 F.2d at 874 (consumers of non-designer jeans might assume that maker of designer jeans had entered a different market because of similarity of mark on non-designer jeans). Consequent*1227ly, there is competitive proximity because consumers could have mistakenly assumed that AD WEEK’S Marketing Week was another Centaur publication, particularly in light of the publications’ similarity.

(4) Bridging the Gap

The fourth factor looks to whether the senior user of the mark is likely to enter the market in which the junior user is operating, that is, bridge the gap. If the senior user can show such an intention, it “helps to establish a future likelihood of confusion as to source.” Lois Sportswear, 799 F.2d at 874. One interest that the trademark law protects is “the senior user’s interest in being able to enter a related field at some future time.” Scarves by Vera, Inc. v. Todo Imports Ltd., 544 F.2d 1167, 1172 (2d Cir.1976). The district court found that Centaur had shown a “clear intention to publish an edition of the magazine with an American focus.” 652 F.Supp. at 1113.

There was evidence at trial that Centaur and A/S/M had discussed the possibility of appellee acquiring appellant. Additionally, Centaur had internal discussions about publishing an edition of Marketing Week with an expanded section covering American marketing news. The changing nature of the market for British and American marketing news supports this finding. With the increasing number of takeovers of American agencies by British agencies and the establishment of overseas offices, the distance between the two markets has narrowed. As a result, the incentive to bridge that gap would increase as market barriers become easier to surmount.

(5) Actual Confusion

The fifth factor looks to whether any consumers had actually been confused by the products bearing the allegedly confusing marks. The district court found actual confusion in this instance based upon misattributions in The Wall Street Journal and New York Times as well as certain responses given by interviewees in the survey conducted at appellant’s instigation. A/S/M’s surveyor found that the interviewers referred to ADWEEK’s Marketing Week as Marketing Week.

The direct evidence we have previously considered on the issue of actual confusion has generally been anecdotal in nature or a market research survey. Inc. Publishing Corp. v. Manhattan Magazine, Inc., 616 F.Supp. 370, 386 (S.D.N.Y.1985), aff'd. mem., 788 F.2d 3 (2d Cir.1986); see also Grotrain Helfferich, Schulz, Th. Steinweg Nachf. v. Steinway & Sons, 523 F.2d 1331, 1340-41 (2d Cir.1975). The newspaper misattributions found in the present case are isolated incidents that are not probative on the issue of actual confusion. See C.L.A.S.S. Promotions, 753 F.2d at 18. Nor do the responses of those surveyed demonstrate instances of actual confusion. Because the surveyors used an improper frame of reference, it is impossible to determine whether the interviewees were aware of Centaur’s publication; therefore, their answers cannot be interpreted as mistakenly referring to one publication when they meant the other.

Evidence of actual confusion is not required to prove the likelihood of confusion between the two marks. Ideal Industries, Inc. v. Gardner Bender, Inc., 612 F.2d 1018, 1024 (7th Cir.1979). The absence of such proof is not especially significant in the present case, particularly given the short time before trial — four months— in which the marks were “competing.” Cf. Plus Products, 722 F.2d at 1006 (absence of instances of actual confusion over three-year period indicated little potential for consumer confusion). Hence, the district court’s error as to this one Polaroid factor does not undermine its ultimate conclusion on the existence of the likelihood of confusion.

(6) Junior User’s Good Faith

The sixth Polaroid factor examines the good faith of the junior user in selecting the mark. Our precedents on this factor appear at least superficially to be at odds. Compare Pegasus Petroleum, 818 F.2d at 258 (intentional copying gives rise to a presumption of a likelihood of confusion) and Harlequin Enterprises, 644 F.2d at 949 *1228(same and collecting cases) with Lois Sportswear, 799 F.2d at 875 (“[Ijntent is largely irrelevant in determining if consumers likely will be confused as to source.”). Lois Sportswear did not signal a departure from the longstanding rule expressed in Pegasus Petroleum, but suggested instead that the absence of good faith cannot control. As we have previously noted: “We have recognized that evidence of intentional copying raises a presumption that a second comer intended to create a confusing similarity of appearance and succeeded____ But if comparison of the [marks] reveals no fair jury issue concerning likelihood of confusion, then intent to copy, even if found from the proffered evidence, would not establish a Lanham Act violation.” Warner Bros. v. American Broadcasting Co., 720 F.2d 231, 246-47 (2d Cir.1983).

Here, appellant was well aware of Centaur’s use of “Marketing Week” as its mark. This awareness can give rise to an inference of bad faith. Cf. Pegasus Petroleum, 818 F.2d at 259 (actual knowledge of registered mark may signal bad faith). This inference was bolstered by the further finding that A/S/M proffered no credible innocent explanation for the change. Cf. Playboy Enterprises, 687 F.2d at 569.

(7) Quality of Junior User’s Product

The seventh factor looks to the quality of the junior user’s product. The district court’s finding that ADWEEK’s Marketing Week is a product of high quality is not disputed and it supports a finding of a likelihood of confusion. The lack of marked difference in quality between goods supports the inference that they emanate from the same source. See Lois Sportswear, 799 F.2d at 875; Plus Products, 722 F.2d at 1006-07.

(8) Sophistication of Consumers

The final Polaroid factor examines the sophistication of the consumers in the relevant market. The parties concede that these consumers are sophisticated. Sophistication of consumers usually militates against a finding of a likelihood of confusion, see, e.g., Plus Products, 722 F.2d at 1007, though it might on occasion increase the likelihood of confusion, depending upon the circumstances of the market and the products. See Grotrian, Helfferich, Schulz, Th. Steinweg Nachf. v. Steinway & Sons, 523 F.2d 1331, 1341-42 (2d Cir.1975). Here it decreased the likelihood of confusion since neither the products nor the marks are identical. Cf. McGregor-Doniger, 599 F.2d at 1137 (suggesting that sophistication of buyers as a factor undercut where marks and related products were identical).

Upon review of the district court’s findings on the various Polaroid factors, we agree with its legal conclusion that there existed a likelihood of confusion as to source as a result of A/S/M’s use of “Marketing Week” in the title of its magazine. Although there have been no instances of actual confusion of the products among the sophisticated consumers, the findings concerning the relative strength of “Marketing Week”, the degree of similarity of the marks, the competitive proximity of the products as well as their quality, Centaur’s intention to bridge the gap, and A/S/M’s intentional copying of Centaur’s mark in bad faith all present a potential for confusion.2

*1229In summary, given the correctness of the findings of secondary meaning and a likelihood of confusion, we affirm the district court’s conclusion that A/S/M infringed Centaur’s trademark and, hence, need not address Centaur’s state law claims.

II Attorneys’ Fees

Relying on § 35 of the Lanham Act, 15 U.S.C. § 1117 (1982), Judge Pollack awarded Centaur attorneys’ fees. See 652 F.Supp. at 1114-15, 1126. Section 35 authorizes the award of reasonable attorneys’ fees in “exceptional” cases of infringement of registered trademarks. In reviewing the award, we must first consider whether § 35 permits the award of attorneys’ fees in an infringement action involving an unregistered trademark and, second, whether this is an exceptional case justifying such an award.

A number of other circuits have decided that § 35 extends to § 43(a) actions, despite the apparent limitations in its language. See WSM, Inc. v. Wheeler Media Servs., Inc., 810 F.2d 113, 116 (6th Cir.1987); Transgo, Inc., 768 F.2d at 1025-26; Rickard v. Auto Publisher, Inc., 735 F.2d 450, 453-58 (11th Cir.1984); Metric & Multistandard Components Corp. v. Metric’s, Inc., 635 F.2d 710, 715-16 (8th Cir.1980); cf. Standard Terry Mills, Inc. v. Shen Mfg. Co., 803 F.2d 778, 781-82 (3d Cir.1986) (refraining from deciding that § 35 applies to common law trademark actions brought pursuant to § 43(a)). District courts within this Circuit have split over the issue. Compare Yeshiva Univ. v. New England Educ. Inst., Inc., 631 F.Supp. 146, 147 (S.D.N.Y.1986) (attorneys’ fees available) with A.S. & W. Prods., Inc. v. Atlantic Steel Co., 207 U.S.P.Q. 251 (W.D.N.Y.1979) (attorneys’ fees not available).

We think that the Eleventh Circuit’s extended discussion of this question in Rickard is compelling. After a thorough exploration of the legislative history and the general purpose of the Lanham Act, Rickard concluded that Congress could not have intended to limit the relief afforded by § 35 only to cases involving registered trademarks: “Thus, it is clear that, without distinguishing between registered and unregistered marks, Congress intended (1) to simplify trademark practice, (2) to establish uniform regulation of trademarks thereby eliminating the possibility that remedies would vary from state to state, and (3) to provide the greatest protection possible for all trademarks used in interstate commerce.” Rickard, 735 F.2d at 457. Consequently, we hold, as have the majority of the circuits that have considered the question, that § 35 applies to § 43(a).

We turn to the question of whether this is an “exceptional” case. Of course, deliberate and willful infringement can render a case “exceptional” and thus support an award of attorneys’ fees. See Quaker State Oil Ref. Corp. v. Kooltone, Inc., 649 F.2d 94, 95 (2d Cir.1981) (per curiam); see also Springs Mills, Inc. v. Ultracashmere House Ltd., 724 F.2d 352, 357 (2d Cir.1983) (district court abused its discretion in failing to consider award of attorneys’ fees in case involving willful infringement). The award of attorneys’ fees in the instant case was predicated upon such a finding.

Moreover, A/S/M has not shown that the award was an abuse of the district court’s discretion. Although appellant claims that it reasonably doubted the validity of Centaur’s trademark, it failed to point to any investigation it made before it changed the title of its publication. It also claims that a case is unexceptional when the infringement involves no loss of sales. It is true that the absence of lost sales has been a factor in some determinations concerning attorneys’ fees. See, e.g., VIP Foods, Inc. v. Vulcan Pet, Inc., 675 F.2d 1106, 1107 (10th Cir.1982). Yet, such a showing is not required to obtain attorney’s fees. See Ultracashmere House, 724 F.2d at 356-57 (district court should have considered availability of award of attorneys’ fees even though plaintiff had not shown lost sales). Therefore, the award of attorneys’ fees is affirmed. We decline to award Centaur attorneys’ fees for this appeal, as it requested.

*1230CONCLUSION

Accordingly, the judgment is affirmed. The case is remanded to the district court for the purposes of fixing the award of attorneys’ fees and for it to consider and decide Centaur’s request that A/S/M be held in contempt for violation of this Court’s order of January 27, 1987.

. For purposes of this appeal, we will ignore the manner in which the titles are actually displayed on the covers of the magazines except where pertinent, since the manner of the presentation of the titles, or their "trade dress,” is not — save for a few exceptions — directly implicated.

. As we have recently noted, the list of Polaroid factors is nonexclusive. See Pegasus Petroleum Corp., 818 F.2d at 256. Thus, in Chandon Champagne Corp. v. San Marino Wine Corp., 335 F.2d 531 (2d Cir.1964), we also considered the nature of the senior user’s priority, its delay in asserting its claim, and the balance of harm and benefit that would result from granting an injunction against the junior user’s use of a mark. See id. at 536. Chandon made findings on these issues even though the district court had not considered them. The district court here did not make any of these determinations, and the record only permits an examination of the first two factors. First, it should be noted that Centaur’s priority in the use of the mark is somewhat sterile due to the limited sales of Marketing Week in this country, but that conclusion must be viewed in light of appellee’s efforts and the small size of the relevant market. Second, Centaur did not delay in asserting its rights. Thus, these two factors also support the result reached.