concurring in part and dissenting in part.
I would affirm to the extent the district court held defendants hable for infringement of Conopco’s trade dress and trademark rights. I see no clear error in that court’s finding that there is a likelihood of confusion between defendants’ packaging and the trade dress of the relaunched VICL product. I also see no error in the finding that defendants’ use of Conopco’s registered trademarks was likely to cause confusion among the relevant class of consumers.
This court’s contrary conclusion on likelihood of confusion is reached by impermissi-bly reweighing the evidence that was before the trial court. This is not only beyond its competence but intrudes on the fact finding role of the district court and encourages litigants to ignore the outcome of the trial and retry the case on appeal. Sometimes, like here, it works.
The trial court properly applied the factors set out in Squirtco v. Seven-Up Co., 628 F.2d 1086, 1091, 207 USPQ 897, 900 (8th Cir.1980), to determine whether a likelihood of consumer confusion exists.1 Likelihood of confusion is a matter of fact and such factual findings are entitled to considerable deference and cannot be reversed unless clearly erroneous. Id. The findings are comprehensive, amply supported by the evidence in the record, are not clearly erroneous, and provide a sufficient basis for the determination that a likelihood of confusion exists.
The district court explained that defendants’ employee, Bowers, testified it was “a major directive [of the company] to obtain the new VICL product” and that he was instructed “to make the revised skin care packaging as close as possible to the VICL packaging.” Conopco, 784 F.Supp. at 659 (district court’s findings of fact (FF) 54-55). The court further found as evidence of defendants’ intentional conduct that, without adequate explanation, they failed to place the well known ‘Wenture” name on the front of the package but instead placed only the VASELINE and INTENSIVE CARE marks on the front of the package, including their trademark registration symbols. Id. at 660-61, 666 (FF 64-68, 134-137).2
Defendants did argue that placing a disclaimer on the package alleviated potential confusion. But, supported by the testimony of an expert on consumer psychology, the district court specifically found that consumers rarely, if ever, pay attention to disclaimers. This is especially true if it is obscurely placed, as was defendants’, in small print on the bottom rear label of the package. Conopco, at 667 (FF 140). The court further found based on this evidence that the disclaimer could actually foster confusion because the consumer would likely focus on the most well-known words in it, VASELINE and INTENSIVE CARE, and miscompre-hend the disclaimer entirely. Id. at 667 (FF *1573141). The court also focused on the expert’s testimony that “the average consumer purchasing inexpensive consumer goods, such as skin care lotions, usually reviews the product for approximately three seconds” and “only retain[s] overall impressions of product appearance, including bottle shapes and colors.” Id. at 665 (FF 115).
In addition, the court found that not only were the trade marks VASELINE and INTENSIVE CARE “strong marks” but survey evidence established that a very high percentage (83%) of respondents throughout the country correctly recognized the VICL bottle even with all textual material removed except for the ingredient list. Conopco, at 665 (FF 114). That the court’s detailed comparison of the products and finding that they were “substantially identical” is not clearly erroneous is self-evident from the packages themselves, many examples of which were available at trial. Id. at 665 (FF 116).3
These extensive findings are supported by the evidence in the record including expert testimony that the trial court found credible and persuasive and the testimony of defendants’ own employees. The court’s detailed comparison of the products properly included an analysis of the features asserted by defendants as reducing confusion. The court’s findings on the Squirtco factors — the strength of plaintiffs trademarks and trade dress; the similarity of the products sold in direct competition; the extreme similarity between plaintiffs trade dress and defendants’ packaging; and defendants’ intent to mimic the VICL trade dress — all amply support its conclusions. Defendants have not established that any of these findings taken alone is clearly erroneous or that, as a whole, they are so defective as to cast doubt on the district court’s conclusion about a likelihood of confusion. That the court failed to discuss every possible piece of evidence in its opinion does not render its decision defective or permit the inference that it ignored relevant evidence. See Medtronic, Inc. v. Daig Corp., 789 F.2d 903, 906, 229 USPQ 664, 667 (Fed.Cir.1986) (“We presume that a fact finder reviews all the evidence presented unless he explicitly expresses otherwise.”).
. The following factors must be considered in order to determine whether a likelihood of confusion exists: (1) the strength of the plaintiff's mark; (2) the similarity of the marks; (3) the similarity of the products; (4) the competitive proximity of the products; (5) the defendants' intent; and (6) instances of actual confusion. Squirtco, 628 F.2d at 1091, 207 USPQ at 900.
. The trial court correctly explained that proof of intent to copy another's trade dress creates an inference that there is a likelihood of confusion. See Squirtco, 628 F.2d at 1091, 207 USPQ at 900. I recognize that in some instances the addition of the "Compare to ...” language used with a competitor's marks may alleviate potential confusion. See Calvin Klein Cosmetics v. Parfums de Coeur, Ltd., 824 F.2d 665, 668-69, 3 USPQ2d 1498, 1500-01 (8th Cir.1987) (limited use of trademarks in comparative advertising is permitted when the statements are truthful and non-confusing). However, the evidence of record before the trial court, as explained more fully below, supports its conclusion that, on the facts found here, May’s use of the VASELINE and INTENSIVE CARE marks was likely to cause confusion despite the addition of the comparative statement.
. It is true that evidence of idiosyncratic and mistaken belief by a consumer, such as Mrs. Sickles here, cannot support a finding of actual confusion. See Litton Systems, Inc. v. Whirlpool Corp., 728 F.2d 1423, 1447, 221 USPQ 97, 112 (Fed.Cir.1984). Nevertheless, after discounting entirely the evidence of actual confusion, the finding of likelihood of confusion is amply supported by the other evidence.