This litigation originates from a dispute over the design appearing on fiberglass pipe. Both Smith Fiberglass Products, Inc. and Ameron, Inc. manufacture fiberglass pipes. Both companies manufacture the pipes by wrapping fiberglass filaments coated with resin around a cylindrical mandrel in two opposite directions, thereby creating two helical patterns that repeatedly cross over each other. The resulting cross-over pattern, which is the best method to maintain pipe strength, is visible through the partially transparent resin unless a dye is added during manufacturing to hide the pattern.

In 1959 Smith began using a helical cross-over pattern1 on the surface of its pipes, which Smith subsequently registered as a trademark. Ameron distinguished its product by adding a black dye to its pipes during the manufacturing process. Other pipe manufacturers differentiated their products either by using dyes or by indicating the company name directly on the pipes.

This arrangement created no difficulties until Ameron branched into the conductive pipe market. Fiberglass pipe normally is not electrically conductive. As a result, a substantial charge of static electricity can build up on the pipe, resulting in a sudden discharge of electricity (sparking). Sparking usually creates no difficulties in the use of fiberglass pipes, but when the pipe is used to carry easily ignitable fuel, sparking creates the possibility of a disastrous explosion.

Conductive pipe solves the sparking problem. Ameron manufactures conductive pipe, Bondstrand 7000, by adding conductive graphite ribbons at regular intervals along with the resin carrying fiberglass filaments. As a result, the ribbons form a helical cross-over pattern, which the black dye used in Ameron pipes initially masked. The dye, however, created difficulties in distinguishing Ameron's conductive pipe from its non-conductive pipe, inspecting the pipe for broken or incorrectly spaced ribbons, and making electrical connections to the ribbons. At the request of a major client, Ameron stopped dying its pipes, revealing a helical cross-over pattern similar to Smith pipes.

Smith sued to prevent Ameron from continuing to incorporate the helical cross-over pattern in its conductive fiberglass pipes. Both federal statutory law and state common law protect trademark interests. See, e.g., Dunn v. Gull, 990 F.2d 348, 350 (7th Cir.1993). Smith sought relief based on three theories: first, trademark infringement pursuant to section 32 of the Lanham Act, 15 U.S.C. § 1114; second, false designation of origin pursuant to section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a); and third, trademark infringement, palming off, and unfair competition pursuant to Wisconsin common law.

For all three of Smith's theories, however, the central issue is the likelihood of consumers in the relevant market confusing the infringer's mark with that of the complainant. See 15 U.S.C. § 1114(1)(a) (trademark infringement); 15 U.S.C. § 1125(a) (false designation of origin); Hirsch v. S.C. Johnson & Son, Inc., 90 Wis. 2d 379, 280 N.W.2d 129, 139 (1979) (Wisconsin common law). After a bench trial the district court found that no likelihood of confusion existed, and in part for that reason held against Smith. This circuit repeatedly has held that the conclusion of a trial court regarding likelihood of confusion is a finding of fact, and we reverse such findings only if they are clearly erroneous. AHP Subsidiary Holding Co. v. Stuart Hale Co., 1 F.3d 611, 616 (7th Cir.1993); Forum Corp. of North Am. v. Forum Ltd., 903 F.2d 434, 438 (7th Cir.1990); see also Scandia Down Corp. v. Euroquilt, Inc., 772 F.2d 1423, 1428 (7th Cir.1985) (likelihood of confusion issue is "all fact and no law"). We will consider a finding of fact to be clearly erroneous only if on the entire evidence we are "left with the definite and firm conviction that a mistake has been committed." United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948)).

This circuit considers the following seven factors, which the district court applied in this case, to be important in evaluating the likelihood of confusion in trademark infringement cases:

(1) similarity between the marks in appearance and suggestion;

(2) similarity of the products;

(3) area and manner of concurrent use;

(4) degree of care likely to be exercised by consumers;

(5) strength of complainant's mark;

(6) actual confusion; and,

(7) intent of defendant "to palm-off his product as that of another."

AHP Subsidiary Holding Corp., 1 F.3d at 615; see also Forum Corp., 903 F.2d at 439. We shall examine these factors in turn.

The district court concluded that the marks on the products are similar. We note, however, that none of the seven factors alone are dispositive, and the weight accorded to each factor will vary from case to case. See id. at 616; Schwinn Bicycle Co. v. Ross Bicycles, Inc., 870 F.2d 1176, 1187 (7th Cir.1989); International Kennel Club of Chicago, Inc. v. Mighty Star, Inc., 846 F.2d 1079, 1087 (7th Cir.1988).

The district court concluded that there are limited similarities between the marks, but that dissimilarities exist as well. The court noted from the bench that "a person without any background in piping could readily distinguish the Ameron product from the Red Thread A product...." The district court formed this opinion from a side-by-side comparison, which Smith asserts was error.

That contention, however, is contrary to the law of this circuit. Computer Care v. Service Sys. Enters., Inc., 982 F.2d 1063, 1070 (7th Cir.1992) ("Further, the close similarity between the parties' trade dresses in side-by-side comparison is also a 'factor to be considered in evaluating the likelihood of confusion.' ") (quoting Roulo v. Russ Berrie & Co., 886 F.2d 931, 937 (7th Cir.1989). In addition, Smith conveniently ignores the fact that one of its witnesses raised the side-by-side comparison issue by testifying that people doing repairs in the field might be confused by the similarity of the marks when replacing small sections of pipe. The district court found the opposite to be true, that a person replacing the pipe, during which time the two pipe sections would be next to each other, could distinguish the pipes. The district court properly evaluated the similarity of the products "in light of what occurs in the marketplace," International Kennel Club, 846 F.2d at 1088.

Even if the side-by-side comparison were improper, the issue is irrelevant. The district court discounted the product similarity factor as unimportant in this case because purchasers of fiberglass pipe rely not on appearances but on product specifications. The court's evaluation of the importance to accord each factor is a factual determination, and was not clearly erroneous in this case.

The district court specifically rejected the contention that Smith and Ameron were in direct competition as to the products in question. Smith's witness on the point admitted that he had no knowledge of direct competition between Ameron's conductive pipe and Smith's non-conductive pipe. Smith in essence asks this court to re-evaluate the credibility of the expert testimony presented at trial, a task we are ill-equipped to undertake. United States ex rel. Jones v. De Roberts, 766 F.2d 270, 273 (7th Cir.1986). The conclusion of the district court that no direct competition existed was not clearly erroneous.

The district court found persuasive the testimony of an expert witness for Ameron that the consumers in the markets in which Ameron sells conductive pipe make their purchases in a sophisticated manner. The court further found that the consumers purchase fiberglass pipes using specifications and a bidding process, are "sophisticated and discriminating," and would not be confused by the appearances of the products. The court discounted the testimony of the two witnesses Smith called on this point, finding their testimony to be without foundation and unconvincing.

Once again, we will not re-evaluate the credibility of the expert testimony in this case. The district court, after hearing all of the testimony and seeing all of the exhibits, concluded that Ameron's account was more credible. The record more than adequately supports that finding of fact, and the determination therefore was not clearly erroneous.

The district court found the Smith mark to be distinctive and strong, which Ameron does not contest on appeal.

The district court found no evidence of actual confusion. The two expert witnesses who testified for Smith at trial were unaware of any instances of actual confusion, and only testified generally that confusion seemed likely. Smith did proffer the deposition testimony of a Smith employee who stated that at an Air Force trade show "a gentleman came in the booth and picked up my sample and said something like, have you been over to Ameron's booth, you're not making that pipe [for them] are you? It looks just like a Smith product." (emphasis added). The district court, noting the lack of an exact quote or even the identity of the person making the statement, disregarded this testimony on hearsay2 and relevancy grounds.

Smith's only argument on appeal relies on the propositions that very little proof of actual confusion is necessary to prove likelihood of confusion, Forum Corp., 903 F.2d at 443, and that instances of actual confusion are entitled to substantial weight, Helene Curtis Indus., Inc. v. Church and Dwight Co., 560 F.2d 1325, 1330 (7th Cir.1977). Here, however, Smith presented no instances of actual confusion, but rather sought to admit into evidence a vague hearsay account of what may have been actual confusion. Forum Corp. specifically recognized that the district court may discount de minimis evidence of actual confusion, see Forum Corp., 903 F.2d at 443, which it properly did here.

Smith essentially asks this court to interpret Forum Corp. and Helene Curtis Industries as crafting a new hearsay exception to the Federal Rules of Evidence for paraphrases of state of mind declarations by unknown declarants. Without hesitation we refrain from adopting such a recommendation. Although we require little proof of actual confusion to demonstrate likelihood of confusion, at minimum the district court must find the proof to be admissible.

The district court noted that Ameron did not intend to palm-off its product as Smith pipe. Rather, the court found that Ameron ceased using black dye on its pipes to satisfy the wishes of a major conductive pipe customer. Smith does not challenge that finding on appeal.

As this discussion of the seven likelihood of confusion factors makes clear, the district court properly concluded that five of the factors favored Ameron and that two favored Smith. The district court then weighed the importance of the factors in Ameron's favor against those in Smith's favor, and concluded that no likelihood of confusion existed. Because the pipe purchasers are sophisticated, the products are similar, no actual confusion existed, the products do not directly compete, and Ameron acted in good faith, the district court decided that mere similarity of the marks and the strength of Smith's mark do not suggest that confusion is likely.

How the district court balanced the factors is subject to the clearly erroneous standard of review. AHP Subsidiary Holding, at 616. More than adequate support exists in the record to support the finding that no likelihood of confusion existed. The district court analyzed the importance of the likelihood of confusion factors in this particular case and weighed those favoring Smith against those favoring Ameron. We will not substitute our judgment for that of the district court. Because the determination that no likelihood of confusion existed undermines all three of Smith's causes of action, we need not address Ameron's affirmative defenses. The judgment of the district court is AFFIRMED.