The opinion of the court was delivered by: JAMES F. HOLDERMAN
JAMES F. HOLDERMAN, District Judge:
In two previous opinions issued respectively on January 31, 1992 (Storck I) and July 2, 1992 (Storck II) this court entered a preliminary injunction orders in favor of plaintiffs Storck USA, L.P. and August Storck K.G. ("Storck") regarding the trade dress of its Werther's Original candy and against the trade dress of defendant Farley Candy Company, Inc.'s ("Farley") Butter Toffee candy. (PX 2 and PX 275, respectively.) The facts detailed therein will not be repeated here. After the entry of each injunction order, Farley modified the design of its Butter Toffee trade dress in compliance with each respective order.
Shortly after this court's July 2, 1992 injunction order, Farley began to market its Butter Toffee candy using a bag and packaging bearing a third trade dress design (PX 306-310) which complied with both of the court's injunction orders. Storck, in the fall of 1992, conducted a survey as to possible consumer confusion regarding the trade dress of its Werther's Original (PX 1) and the Farley's Butter Toffee third trade dress design (PX 306). In November 1992 Storck filed its third motion for preliminary injunction. After allowing time for further discovery by the parties, the court in early March 1993 conducted a multi-day hearing as to Storck's motion. Based upon the facts adduced, Storck's motion must be denied.
I. Storck's Likelihood of Success on the Merits.
The Seventh Circuit in Computer Care v. Service Systems Enterprises, Inc., 982 F.2d 1063, 1067-68 (7th Cir. 1992) reiterated the essential requirements of a trade dress infringement claim stating:
In order to prove trade dress infringement, the plaintiff must establish that: (1) its trade dress is "inherently distinctive" or has acquired "secondary meaning"; (2) the similarity of the defendant's trade dress to that of the plaintiff creates a "likelihood of confusion" on the part of consumers; and (3) the plaintiff's trade dress is "non-functional"; Roulo, 886 F.2d at 935; see also Abbott Labs., 971 F.2d at 20; Schwinn Bicycle, 870 F.2d at 1182-83. The third element, that of "functionality," is actually an affirmative defense on which the defendant bears the burden of proof. Abbot Labs., 971 F.2d at 20; W.T. Rogers Co. v. Keene, 778 F.2d 334, 338 (7th Cir. 1985).
Therefore to demonstrate a better than negligible likelihood of success on the merits Storck must present proof as to each essential element of its trade dress infringement claim and must present proof which has a better than negligible change of overcoming Farley's evidence supporting Farley's affirmative defense as to the functionality and unprotected nature of the trade dress features for which Storck seeks injunctive relief. To do so Storck had relied primarily on survey evidence. The court will address that evidence.
The results of Storck's survey (PX 369), according to Storck's expert Robert Lavidge's "best estimate" was 21% confusion (26.6% - 5% noise) of those surveyed in the Storck survey. Farley criticizes Storck's survey as not reflecting the marketplace setting and as employing "forced answer" questions. Farley's survey, which Storck criticized as "a reading test," resulted in 7.6% confusion (DX 70) among those surveyed in the Farley survey.
Although the percentage of confusion shown by Storck's most recent survey is substantially less than previous survey results obtained by Storck as to previous Farley's Butter Toffee packaging (PX 275) (13% less than the 34% shown by the survey conducted before Storck II), Storck's survey as to Farley's third trade dress design (hereinafter "Third Bag,") if trustworthy, demonstrates sufficient survey confusion to support a finding of likelihood of confusion. See James Burrough, Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266 (7th Cir. 1976). Farley's survey's 7.6% confusion percentage weighs against a finding of likelihood of confusion. See Henri's Food Products, Co. v. Kraft, Inc., 717 F.2d 352 (7th Cir. 1983).
The court finds, however, that several factors diminish the significance of the results of Storck's survey as to Farley's Third Bag regarding the issue of the likelihood of confusion as to source. The "forced answer" or "same or different" survey question technique, though approved by the Seventh Circuit in other contexts, see McGraw-Edison Co. v. Walt Disney Productions, 787 F.2d 1163, 1172 (7th Cir. 1986), may have enhanced Storck's survey results as to the confusion among the survey participants. The extent of this was not shown. More importantly, however, as discussed in the next section, some of the survey confusion may have resulted from the fact that certain features in the trade dress of Farley's Third Bag which may be similar to elements of the Werther's Original trade dress which have not been proven to be protected under the Lanham Act. These unprotected trade dress elements contributed to a skewing of the survey results in a way which clouds rather than clarifies the issue of likelihood of confusion as to source among consumers.
B. The Trade Dress at Issue.
The total image of Werther's Original trade dress is distinctive but the trade dress features of Farley's Third Bag which Storck seeks now to enjoin relate to generic, functional, and ...