ConAgra, Inc. (ConAgra) sued Geo. A. Hormel Company (Hormel), claiming that Hormel infringed ConAgra's HEALTHY CHOICE trademark as applied to shelf-stable food products. ConAgra claimed Hormel infringed on ConAgra's trademark by utilizing Hormel's HEALTH SELECTIONS trademark as applied to shelf-stable food products. Hormel responded by filing counterclaims regarding ConAgra's use of allegedly false and misleading advertisements.

The parties consented to try this case before me in my capacity as a United States Magistrate Judge, and the Honorable William G. Cambridge, United States District Judge for the District of Nebraska, referred this case to me for trial. Pursuant to Federal Rule of Civil Procedure 52, I now set forth the court's findings of fact and conclusions of law in the form of this memorandum opinion. I find in favor of Hormel on ConAgra's claim of trademark infringement, and I find in favor of ConAgra on Hormel's false advertising claims.

Any finding of fact which is more properly considered a conclusion of law shall be so construed. Alternatively, any conclusion of law which is more appropriately considered as a finding of fact shall be so construed.

I. A.

Although phrased in a variety of ways, there are essentially two questions to be resolved by the court. Those questions are:

1. Does Hormel's use of the mark HEALTH SELECTIONS for a line of shelf-stable products infringe ConAgra's proprietary rights in and to the mark HEALTHY CHOICE under 15 U.S.C. § 1114(1) and 1125(a)?

2. Did ConAgra engage in false, misleading and/or deceptive advertising when advertising and promoting its HEALTHY CHOICE shelf-stable products, primarily in violation of 15 U.S.C. § 1125(a)?

There are state law claims for which the court has pendent jurisdiction, but such state law claims are resolved by decision of the federal claims.

B.

A pretrial conference was held in this matter on the 2nd day of August, 1991 (filing 48). The parties have agreed that the following may be accepted as established facts for purposes of this case only:

JURISDICTION AND VENUE

1. This court has jurisdiction under 15 U.S.C. § 1121 and 28 U.S.C. § 1331, 1338(a) and (b) and pendent jurisdiction.

2. Venue lies in this court pursuant to 28 U.S.C. § 1391(b) and (c).

3. On April 29, 1991, the parties in this action, through their counsel, waived the right to proceed before a judge of the United States District Court and consented to have the United States Magistrate Judge conduct any and all further proceedings in this case (including the trial) and order the entry of final judgment.

PARTIES

4. Plaintiff, ConAgra, Inc. ("ConAgra"), is a corporation organized and existing under the laws of the State of Delaware. ConAgra maintains its principal place of business in Omaha, Nebraska.

5. ConAgra is a diversified food company. It operates its business through a number of independent operating companies, including ConAgra Frozen Food Company.

6. Defendant, Geo. A. Hormel Company ("Hormel"), is a corporation organized and existing under the laws of the State of Delaware. Hormel maintains its principal place of business in Austin, Minnesota.

7. Hormel, like ConAgra, is a diversified food company. Hormel operates its business through six groups: The Prepared Foods Group (which includes the Grocery Products Division), the Meat Products Group, the Frozen Foods and International Group, the Foodservice Group, the Operating Group and the Administrative Group.

8. At all relevant times, Hormel has been a major player in the shelf-stable food business, engaged in manufacturing and selling a variety of food products including shelf-stable foods in interstate commerce.

HEALTHY CHOICE

9. ConAgra adopted the mark HEALTHY CHOICE in 1988. The name was generated in an ideation session in the spring of 1988.

10. On or about May 3, 1988, ConAgra conducted a computerized trademark search for the mark HEALTHY CHOICE.

11. On or about June 22, 1988, ConAgra shipped HEALTHY CHOICE chicken and pasta divan frozen dinners to each of the following locations:

No Frills Supermarket 1817 West Broadway Council Bluffs, Iowa 51501

No Frills Supermarket 8005 Blondo Street Omaha, Nebraska 68134

12. On August 1, 1988, ConAgra filed an application to register the mark HEALTHY CHOICE in International Class 29 on the Principal Register of the United States Patent and Trademark Office. The goods description in that application read "frozen prepared dinners comprised of meat and vegetables and fruit; poultry and vegetables and fruit; seafood and vegetables and fruit." Said mark was published for opposition on December 6, 1988. No opposition being made, Certificate of Registration No. 1,528,468 was issued on March 7, 1989, to ConAgra for the mark HEALTHY CHOICE.

13. In October of 1988, ConAgra began shipping HEALTHY CHOICE product to areas geographically representing approximately twenty-five percent (25%) of the United States.

14. In February of 1989, ConAgra became aware that Edwards Baking Company in Atlanta, Georgia, was seeking to register the mark HEALTHY CHOICE for bakery goods. ConAgra acquired Edwards Baking Company's rights to the mark HEALTHY CHOICE and its application to register the mark in consideration for a Ten Thousand Dollar ($10,000) payment. The application acquired from Edwards Baking Company eventually resulted in the issuance of Certificate of Registration No. 1,552,156 now owned by ConAgra.

15. On or about June 13, 1989, ComSource Independent Foodservice Companies, Inc., and North American Foodservice, Inc. (Collectively "ComSource"), as plaintiffs, sued ConAgra, Inc., in the United States District Court for the Northern District of Georgia (Atlanta, Georgia).

16. ComSource is a national food service organization which provides products and services to independent food service distributors throughout the United States. These distributors sell food products to institutional feeding establishments such as restaurants, hospitals and nursing homes.

17. ComSource took the position in the litigation that ConAgra's adoption and use of its mark HEALTHY CHOICE for frozen dinners sold to retail grocery stores infringed ComSource's trademark rights in the mark HEALTH CHOICE for nutritious food products sold through the food service industry.

18. ConAgra denied that it was infringing ComSource's trademark rights and a trial on ComSource's Motion for Preliminary Injunction was held in Atlanta in November of 1989. ConAgra filed Findings of Fact and Conclusions of Law in connection with that trial.

19. Prior to a decision in the litigation, the parties resolved their differences and entered into a settlement agreement.

20. Pursuant to the Settlement Agreement, ComSource conveyed to ConAgra all of its right, title and interest in and to the mark HEALTHCHOICE.

21. In addition to frozen dinners, ConAgra has marketed the following food products under the mark HEALTHY CHOICE:

Product Description Date

Frozen Entrees 10/30/89 Shelf-Stable Products 10/15/90 Egg Substitute 11/26/90 Ice Cream 02/15/91 Breakfast Items 04/01/91 (approx.)

22. ConAgra has filed the following applications to register HEALTHY CHOICE on certain food products:

Mark Serial No.

HEALTHY CHOICE 74/041,263 HEALTHY CHOICE 74/066,713 HEALTHY CHOICE FOR KIDS 74/076,282 HEALTHY CHOICE 74/132,427 HEALTHY CHOICE DESIGN 74/172,150 HEALTHY CHOICE DESIGN 74/172,323

23. In November of 1990, ConAgra became aware that Campbell Soup Company ("Campbell") had commenced using the mark HEALTHY REQUEST on soup.

24. On the 29th day of November, 1990, ConAgra sent a cease-and-desist letter to Campbell requesting that Campbell immediately cease using the mark HEALTHY REQUEST on soups. In that letter ConAgra contended that HEALTHY REQUEST was "similar in appearance and identical in meaning to [ConAgra's HEALTHY CHOICE trademark]."

25. On or about the 7th day of December, 1990, Campbell instituted a declaratory judgment action in the United States District Court for the District of New Jersey styled Campbell Soup Company v. ConAgra, Inc., Civil Action No. 90-4965 (SSB).

26. On or about the 12th day of March, 1991, the litigation between ConAgra and Campbell was settled by entry of a Consent Decree.

27. By June of 1987, ConAgra, as part of its "project AGNEW," began the process of selecting a trademark for a proposed new line of frozen dinners intended to be low in calories, fat, sodium and cholesterol.

28. ConAgra conducted several trademark searches for proposed trademarks to be used on or in association with its healthy frozen dinners. These searches included a trademark search of HEART'S CHOICE on April 7, 1988, a search for HEALTHY CUISINE conducted on April 7, 1988, searches for HEALTHY CHOICE and HEARTLINE on May 3, 1988, and a search for marks which included "BETTER" in combination with "CHOICE" which was dated Spring of 1988.

29. ConAgra wanted a product name that communicated to the consumer the product concept of being "healthy" and good for the heart. In fact, other marks considered in the trademark selection process included "HEART'S CHOICE," "GOOD FOR YOU" and "HEALTHY HEART."

30. The "HEALTHY HEART" term was dropped from consideration after ConAgra was told by the USDA [United States Department of Agriculture] that the use of that term would be disallowed because it implied the ConAgra dinners had health benefits. ConAgra selected the HEALTHY CHOICE name for its "healthy" frozen dinners in the Spring of 1988 and subsequent to the trademark searches referenced in paragraph 28, supra.

31. Since October of 1988, ConAgra's HEALTHY CHOICE products have included fish or seafood dinners and entrees.

32. Following ConAgra's marketplace introduction of its frozen HEALTHY CHOICE products, ConAgra filed another federal trademark application to register the HEALTHY CHOICE mark as applied to shelf-stable products. That application was filed in March of 1990 and was based on ConAgra's intent-to-use [ sic] the mark on shelf-stable goods.

33. At the time suit was filed, ConAgra's HEALTHY CHOICE shelf-stable products were being test-marketed in four markets — Denver, Des Moines, Omaha and Miami.

HEALTH SELECTIONS

34. In March of 1989, Hormel sought to register the following three trademarks:

HEALTH MENU NUTRITIOUS CUISINE HEALTH SELECTIONS

35. In March of 1989, Hormel was marketing a line of shelf-stable products in a tray container under its trademark TOP SHELF.

36. In connection with Hormel's effort to register the three trademarks described in Paragraph 34, supra, above, Hormel had black-and-white labels depicting each of the three marks printed and applied to shelf-stable products, i.e., Lemon Fillet of Cod and Linguini with Clam Sauce, that were at that time being marketed under the TOP SHELF trademark.

37. In May and June of 1989, Hormel shipped in interstate commerce cases of Lemon Fillet of Cod and cases of Linguini with Clam Sauce under each of the three marks set forth in Paragraph 34, supra.

38. On the 27th day of June, 1989, Hormel filed applications to register each of the three marks described in Paragraph 34, supra, for "main dish entrees with seafood or fish as the primary ingredient."

39. On March 6, 1990, the U.S. Patent and Trademark Office issued Certificates of Registration for the following marks:

Mark Registration No.

HEALTH MENU 1,585,934 HEALTH SELECTIONS 1,585,935

40. Since prior to ConAgra's introduction of its HEALTHY CHOICE frozen products (Fall, 1988), Hormel has manufactured and sold microwaveable shelf-stable products. Other shelf-stable microwaveable products on the market include Chef Boyardee, Lunch Bucket and Light Balance (Doubletree/Dial), Kraft, Healthy Recipes and, most recently, Healthy Choice (ConAgra).

41. During the PTO [Patent and Trademark Office] prosecution of the HEALTH SELECTIONS trademark application, the examining attorney did not cite ConAgra's registered HEALTHY CHOICE mark. Accordingly, on December 12, 1989, the HEALTH SELECTIONS mark was published for opposition by anyone believing it would be damaged by the registration of the mark on the Principal Register.

42. A representative(s) of ConAgra requested a search report for the mark HEALTHY START and a Thomson Thomson full search report on that mark was received and reviewed by representative(s) of ConAgra, Inc., on or shortly after March 7, 1990, the date of the search report. One of the entries in that search report was the pending federal trademark application owned by Geo. A. Hormel Company for the mark HEALTH SELECTIONS for main dish entrees with seafood or fish as the primary ingredient dated June 27, 1989, stating a date of first use in interstate commerce of May 3, 1989, and which showed that the application was published for opposition on December 12, 1989.

43. Since on or before December 12, 1989, and continuing to the present date, the law firm of McGrath, North, Mullin Kratz subscribed to and received the Official Gazette of the United States Patent and Trademark Office. This publication lists federal trademark applications that are published for opposition and federal trademark registrations that issue, among others. During the same time period, McGrath, North, Mullin Kratz represented ConAgra, Inc., with respect to certain trademark matters including but not limited to the trademark HEALTHY CHOICE.

44. No opposition to the initial registration of the HEALTH SELECTIONS mark was lodged by ConAgra, or anyone else, and on March 3, 1990, United States Trademark Registration No. 1,585,935 on the Principal Register issued to Hormel for that mark.

45. Product development of a line of products ultimately marketed under the HEALTH SELECTIONS brand name in a microwaveable cup container began in earnest in June of 1990. Some development work on that product line began earlier.

46. Label development work on the HEALTH SELECTIONS cup products began in September of 1990.

47. Hormel's HEALTH SELECTIONS product has been sold in test markets since February, 1991, which test markets are Des Moines, Omaha, Denver, Phoenix and Dallas.

II.

I turn first to the issue of whether or not Hormel has infringed ConAgra's HEALTHY CHOICE mark. The "ultimate issue on the infringement claim [is] whether the recent mark [HEALTH SELECTIONS] so resembled the registered mark [HEALTHY CHOICE] that its use was `likely to cause confusion, or to cause mistake, or to deceive.'" SquirtCo v. Seven-Up Co., 628 F.2d 1086, 1090-91 (8th Cir. 1980) (citing 15 U.S.C. § 1114(1) and 1127). Both parties agree that when assessing "likelihood of confusion," a district court should consider the following six factors:

a. The strength of the trademark;

b. The similarity between the trademark and the defendant's mark;

c. The competitive proximity of the product on which the respective marks are placed;

d. The intent of the alleged infringer to pass off its goods as those of the trademark holder;

e. The incidence of actual confusion; and

f. The degree of care likely to be exercised by potential customers of the trademark holder. Id. at 1091.

Accordingly, it is appropriate to analyze the evidence. Based on the analytical factors outlined in SquirtCo, I shall proceed to do so.

"Likelihood of confusion is a finding of fact." SquirtCo, 628 F.2d at 1091.

A. THE STRENGTH OF THE MARK

As Professor McCarthy has stated: "All trademarks are not equal." 1 J. McCarthy, Trademarks and Unfair Competition § 11:24, at 503. Marks which are "strong" are given protection over a wide variety of related products and services and variations on visual and aural format, and relatively "weak" marks are given more narrow protection, both as to products and format variations. Id.

The "rationale is that the more distinctive, unique and well known the mark, the deeper is the impression it creates upon the public's consciousness and the greater the scope of protection to which it is entitled." Id. at 504 (footnote omitted). As Professor McCarthy suggests, it is "absurd" to separate marks into two neat categories of either strong or weak. Id. Rather, the question is whether the senior user's mark is sufficiently distinctive that a junior user's mark will create customer confusion when used in a different format and appearance, or when used on a different line of goods, or when used in a different territorial area. Id. Thus, the "more customers know and recognize the senior user's mark, the more likely that they will be confused by a junior use." Id.

But even a "weak" mark is entitled to protection against subsequent registration or use by another for a closely similar format on closely competitive goods or services. Id. at 505 (footnote omitted). Thus, whether a mark is "weak" or not is of little importance if the junior mark is identical and the goods are closely related. Id. at 505-06 (footnote omitted).

In this connection, it is important to note that registration of a trademark is prima facie evidence of the validity of the registration, the registrant's ownership of the mark, and of the registrant's exclusive right to use the mark in commerce in connection with the goods and services specified in the certificate of registration. General Mills, Inc., v. Kellogg Co., 824 F.2d 622, 626 (8th Cir. 1987) (citing 15 U.S.C. § 1057(b)). But "a mark's registered status is only an evidentiary tool, and the fact of registration does not affect the plaintiff's ultimate burden of proof in an infringement action." Id.

When assessing the strength of a mark, two factors are normally weighed: (a) the placement of the mark on the spectrum of marks, and (b) the marketplace recognition value of the mark. Id. at 509. This test distinguishes, on a time scale, two separate dimensions of strength. The first inquiry focuses on the inherent potential of the term at the time of its first use. The second evaluates the actual customer recognition value of the mark, in this case at the time the mark is asserted in litigation to prevent another's use.

(1)

The courts have historically looked at the strength of trademarks by endeavoring to place them on a spectrum. Thus, in ascending order, which roughly reflects their eligibility for trademark status, and the degree of protection afforded the marks, the categories are: (a) generic terms which are never entitled to trademark protection; (b) descriptive trademarks, which are not inherently distinctive, but because they have acquired secondary meaning may be entitled to trademark protection; (c) suggestive trademarks, where no secondary meaning is required, and which are considered inherently distinctive, and (d) arbitrary or fanciful trademarks which are considered inherently distinctive and for which no secondary meaning is required. General Mills, 824 F.2d at 625; 1 J. McCarthy, Trademarks and Unfair Competition § 11:1, at 433.

ConAgra contends that its mark is "suggestive," but not "arbitrary" or "fanciful." Plaintiff's post-trial brief at 9 ("the HEALTHY CHOICE mark is suggestive and is entitled to broad protection."). On the other hand, Hormel argues that the mark HEALTHY CHOICE is merely descriptive.

The significance of this distinction is that proof of secondary meaning is required for descriptive marks rather than suggestive marks. General Mills, 824 F.2d at 625. Secondary meaning is: "evidence that the consumers accept and recognize a mark as denoting only one seller or source." Worthington Foods, Inc., v. Kellogg Co., 732 F. Supp. 1417, 1433 (S.D.Ohio 1990).

The terms "descriptive" and "suggestive" are not mutually exclusive. 1 J. McCarthy, Trademarks and Unfair Competition § 11:20, at 490. The borderline is not clear, its location in any given situation "is hazy and only subjectively definable," and the descriptive category "imperceptibly shades over at its fringe into the suggestive domain." Id. § 11:21, at 491 (footnotes omitted).

It is therefore helpful to understand why descriptive marks do not warrant trademark protection in the absence of proof of secondary meaning. It is generally agreed that the descriptive term informs the buyer of the purported quality of the product, and since many other products may have similar qualities, the use of the term will not help the consumer distinguish between products of different sellers. Id. § 11:5, at 444 (footnote omitted). Moreover, descriptive terms are thought to belong to the public and therefore for policy reasons it is thought that one competitor should not be permitted to limit the language of commerce by preventing other competitors from fairly describing their own goods. Id. at 444-45 (footnotes omitted).

There are a variety of tests which have been employed in an effort to distinguish a descriptive mark from a suggestive mark. Id. § 11:21. The one most frequently employed by the courts, and this circuit, is the so-called "imagination" test. Id. at 491-92. As the court stated in General Mills, suggestive marks are those "marks, requiring imagination to reach a conclusion as to the product's nature." 824 F.2d at 625. When this test is employed: "If the mental leap between the word and the product's attributes is not almost instantaneous, this strongly indicates suggestiveness, not direct descriptiveness." 1 J. McCarthy, Trademarks and Unfair Competition § 11:21, at 492.

In this connection Professor McCarthy notes that marks which are laudatory are considered as a condensed form of describing the character or quality of the goods, and are thus regarded as "descriptive:"

Marks which are merely "laudatory" and descriptive of the alleged merit of a product are also regarded as being "descriptive." This includes such marks as GOLD MEDAL, BLUE RIBBON, and the like. Since each tangible product carries with it a "psychic load" of intangible consumer psychological expectations about the product, a mark could be "descriptive" of the product itself or those intangible expectations, or both. Self-laudatory or "puffing" marks are regarded as a condensed form of describing the character or quality of the goods.

Id. § 11:5, at 443 (footnotes omitted).

In this case, I conclude that the "mental leap" between the words and the product's attribute "is almost instantaneous." The mark HEALTHY CHOICE, when associated with foods, while not directly descriptive of a food product, does immediately provide a message of "the healthy characteristics of the good simply from looking at the mark." Worthington Foods, Inc., v. Kellogg Co., 732 F. Supp. 1417, 1435 (S.D.Ohio 1990) (holding that the mark HEARTWISE applied to food did not permit the consumer to "directly cull a message concerning the healthy characteristics of the good simply from looking at the mark") (emphasis added).

In Worthington Foods the court reasoned that while the mark HEARTWISE might convey the message "wise for the heart," consumers must understand the relationship between the low fat, low cholesterol nature of food and the connection between cholesterol and heart disease before fully understanding the message that the product with which the mark was associated had "healthy characteristics." Id. In this case, even the minimal reasoning process described in the Worthington Foods case is not necessary to reach the conclusion that the goods to which the mark is associated are "good for you."

ConAgra seems to believe that since the mark HEALTHY CHOICE does not immediately bring to mind a food product that the mark is therefore suggestive. It is true that the mark HEALTHY CHOICE does not directly suggest the product with which the mark is associated, but the suggestive/descriptive dichotomy has not, as I have understood it, been limited in such a fashion. Rather, the question is whether the term "directly and clearly conveys information about the ingredients, qualities or characteristics of the product or service." 1 J. McCarthy, Trademarks and Unfair Competition § 11:21, at 491 (emphasis added).

In this connection, I was particularly impressed by how ConAgra chose to publicize its HEALTHY CHOICE mark. ConAgra's television commercials (defendant's exhibit 531) feature celebrities discussing ConAgra's HEALTHY CHOICE frozen entrees in terms of their low fat, sodium and cholesterol contents. These commercials end with a celebrity asking the consumer to "make a healthy choice" or "make the healthy choice." I believe it particularly "relevant that the proponent of trademark status itself used the term in a descriptive sense." Id. at 494 (footnote omitted).

My finding is supported by both of the linguistic experts called to testify by the parties. Both classified the HEALTHY CHOICE mark as "laudatory." (filing 95, TR 591:16-22; filing 98, TR 969:5-10).

In this connection, Dr. Shuy, Hormel's linguistics expert, testified that he did a computer search of publications to determine the common usage of the term "healthy choice" (filing 98, TR 933:11-25; 934:1-7). This search revealed in excess of 30 articles (defendant's exhibits 522 and 523) using "healthy choice" which were "used primarily in the context of Healthy [ sic] foods" (filing 98, TR 936:1-2).

Shuy reviewed various commercial brochures and advertising materials, including one put out by the Washington Apple Commission on a program called "Healthy Choices" (filing 98, TR 938:19-24). This document (defendant's exhibit 527) provides a fold-out wall chart for schoolchildren counseling them to "Make a Healthy Choice." (filing 98, TR 939:16-25; 940:1-15). The Washington State apple growers' "Healthy Choices for Kids" program was written up in a Washington Post article, dated June 14, 1991, which indicated that about 3 million schoolchildren throughout the country received this free educational kit (defendant's exhibit 522).

Still further, Mr. Golderman, a ConAgra executive responsible for developing new products under the HEALTHY CHOICE brand, testified that he had been aware for about two years of a nutritional program of General Foods that utilized the slogan or phrase "Make a Healthy Choice" (filing 94, TR 470:18-25; 471:1-3). This slogan appears to be a registered trademark of General Foods (defendant's exhibit 521 at 55, U.S. Registration No. 1,528,219).

This evidence of how the public is likely to perceive the mark HEALTHY CHOICE is customarily received in trademark cases. Best Buy Warehouse v. Best Buy Co., Inc., 751 F. Supp. 824, 826 (W.D.Mo. 1989) (third party advertisements probative of the descriptive nature of "Best Buy"), aff'd, 920 F.2d 536 (8th Cir. 1990).

In further support of its claim that HEALTHY CHOICE was laudatory and descriptive as opposed to suggestive, Hormel introduced a collection of 78 third party trademark registrations (defendant's exhibits 520 and 521). Defendant's exhibit 520 is a collection of 21 "health" and "healthy" marks. Defendant's exhibit 521 is primarily a collection of 57 "choice" marks, some of which also include the "health" formative.

I have discounted, in terms of evidentiary significance, much of this evidence of third party trademark registration because Hormel did not provide significant accompanying proof of the extent and length of the actual use of these other registrations. 1 J. McCarthy, Trademarks and Unfair Competition § 11:26, at 514-16. Moreover, since much of the third party registration evidence provided by Hormel regards third party usage of marks on wholly unrelated products, the probative value of much of Hormel's evidence of third party registration is quite small. Id. at 514 (footnote omitted).

However, these registrations do have some evidentiary significance in showing how various terms, words, prefixes or suffixes of a mark are used in ordinary parlance. Id. at 516. In this sense, they are very much like the printed materials appearing in the public press described above.

For example, consider the case of Nestle Foods Corp. v. Kellogg Co., 6 U.S.P.Q.2d 1145 (T.T.A.B. 1988), where the patent and trademark trial and appeal board was confronted with litigation over TASTER'S CHOICE as applied to coffee and DINER'S CHOICE for tea, and the board considered such other third party marks as BREWER'S CHOICE for coffee and KING'S OWN CHOICE for tea. The board said:

While opposer is perfectly correct that this record does not establish widespread use of other marks in incorporating the word "CHOICE" applied to food products, the certificates of registration introduced by applicant are relevant evidence, when taken together with the dictionary evidence, to the extent that they tend to prove that the word "CHOICE," as a component of a mark applied to food products, is not wholly arbitrary but, rather, has a somewhat laudatory significance as applied to the respective food products of these parties. See Tektronix, Inc. v. Daktronics, Inc., 748 F.2d 675, 189 U.S.P.Q. 693 (C.C.P.A. 1976).

Id. at 1149.

The citation to 748 F.2d 669 regarding Tektronix, Inc., in Nestle Foods is incorrect. The correct citation is Tektronix, Inc., v. Daktronics, Inc., 534 F.2d 915, 917 (C.C.P.A. 1976) (holding that the board was not in error in dissecting the marks by considering 38 third party registrations having the suffix "tronics" or "tronix" where the holder of the mark "Tektronix" opposed registration of the mark "Daktronics"). The Tektronix opinion is cited with apparent approval in Specialty Brands, Inc., v. Coffee Bean Distributors, Inc., 748 F.2d 669, 675 (Fed. Cir. 1984).

Also in one instance there was significant evidence of a third party's usage of a similar mark on similar goods other than on a Hormel or ConAgra product. The HEALTHY RECIPES product, by Sandoz Nutrition Corporation, for shelf-stable microwaveable meals, bearing Registration No. 1,489,764, registered May 24, 1988, (defendant's exhibit 520, at 1) is competitive with ConAgra's HEALTHY CHOICE shelf-stable microwaveable product and Hormel's HEALTH SELECTIONS shelf-stable microwaveable product. A specimen of this product appears as defendant's exhibit 552. Photographs show that the HEALTHY RECIPES product was available in Miami, Florida (defendant's exhibits 619 and 620). Mr. Bunge, a Hormel consumer survey expert, testified that he personally observed the HEALTHY RECIPES product in numerous stores in Chicago, Illinois (filing 96, TR 748:22-25; 749:1-4).

ConAgra responds to the evidence of the HEALTHY RECIPES registration by suggesting that Hormel seeks to "dissect" ConAgra's trademark. In other words, ConAgra does not believe the mark HEALTHY RECIPES is sufficiently similar to ConAgra's mark for comparative purposes. ConAgra was aware of the HEALTHY RECIPES registration. It had concluded that despite the fact that the package on the HEALTHY RECIPES product used a good bit of green, like ConAgra, and the word "Healthy" was used in the mark, the HEALTHY RECIPES mark was not confusingly similar. This was true according to a ConAgra executive because ConAgra did not have the right to exclusive use of the word "Healthy" and the word "Recipes" was not synonymous with the word "Choice" (filing 85, Hughes deposition, TR 89:2-25; 90:1-24).

I agree with ConAgra that "dissection" of a mark based upon minute similarities or dissimilarities is inappropriate as suggested in the case cited by ConAgra, Grandpa Pidgeon's of Missouri v. Borgsmiller, 477 F.2d 586, 587 (C.C.P.A. 1973) (holding that it was inappropriate to dissect two figures of a "Grandpa" on the basis that one was slightly stooped and the other apparently spry). However, I do not believe Hormel suggests the kind of minute scrutiny that ConAgra and the Grandpa Pidgeon's of Missouri court criticize. In this case, Hormel's effort to analyze "health," or for that matter "choice," is essentially an analysis of major constituents of ConAgra's mark, not some relatively insignificant part of the mark. I think a competitor's use of a registered trademark employing the word "Healthy" on competing goods is indicative of the laudatory nature of marks such as those at issue in this lawsuit. Compare Nestle Foods Corp. v. Kellogg Co., 6 U.S.P.Q.2d at 1149 (the word "Choice" was laudatory when applied to the food products of the parties in suit).

I therefore find that the third party registration of HEALTHY RECIPES is of more relevance than the other third party registrations proffered by Hormel, that consideration of the HEALTHY RECIPES registration (or for that matter other third party registrations) does not constitute improper dissection of ConAgra's mark, and that the HEALTHY RECIPES registration, and to a much lesser extent the other third party registrations, strongly suggest that the formatives "Healthy" or "Choice" have "a somewhat laudatory significance as applied to the respective food products of these parties." Nestle Foods Corp., 6 U.S.P.Q.2d at 1149.

Nevertheless, third party use evidence of the kind proffered by Hormel is certainly not determinative; it provides only a small part of the puzzle. General Mills, Inc., 824 F.2d at 626-27.

Finally, Hormel argues that ConAgra has made a judicial admission that ConAgra's mark is weak. In Comsource Independent Foodservice Companies, Inc., et al. v. ConAgra, Inc., No. 1:89-CV-1248 (N.D.Ga), ConAgra was accused of infringing Comsource's trademark HEALTHCHOICE. (Defendant's exhibit 549). In that litigation, ConAgra filed Proposed Findings of Fact and Conclusions of Law wherein it strenuously argued that the HEALTHCHOICE mark was a laudatory mark and therefore descriptive and inherently weak, stating:

The HEALTHCHOICE name is intended to convey a meaning of healthy products which represent a good choice for consumers. The HEALTHCHOICE name thus describes one of the attributes of its products. (Defendant's (sealed) exhibit 535 — ConAgra's Proposed Findings of Fact at p. 4, ¶ 15).

Characterizing the HEALTHCHOICE mark as laudatory, ConAgra asserted that:

A laudatory mark (such as HEALTHCHOICE) is an inherently weak mark and, accordingly, Plaintiffs' rights in the mark must be narrowly construed. (Citations omitted).

(Defendant's (sealed) exhibit 535 — ConAgra's Proposed Conclusions of Law at p. 23, ¶ 15).

ConAgra strenuously argued the weakness of HEALTHCHOICE:

Plaintiffs' mark HEALTHCHOICE is an inherently weak mark and, accordingly, is entitled to limited protection. Plaintiffs' mark can be described as "laudatory." As a noted trademark authority indicated:

Marks which are merely `laudatory' and descriptive of the alleged merit of a product are also regarded as being `descriptive.' This includes such marks as GOLD MEDAL, BLUE RIBBON and the like . . . Selflaudatory [ sic] or `puffing' marks are regarded as a condensed form of describing the character or quality of the goods.

See 1 J. McCarthy, Trademarks and Unfair Competition, § 11:5, at 443 (2d Ed. 1984). Because the mark HEALTHCHOICE is descriptive, it is considered weak and is entitled to limited protection. (Defendant's (sealed) exhibit 535 — ConAgra's Proposed Ultimate Findings of Fact and Conclusions of Law at p. 37, ¶ 44).

ConAgra's defense in the Comsource litigation was based, in part, upon numerous third party trademark applications and registrations for marks including the "health," "healthy" and/or "choice" formatives. ConAgra admitted that both the HEALTHCHOICE mark and its own HEALTHY CHOICE mark were entitled to only narrow rights. ConAgra argued:

Not only is Plaintiffs' mark inherently weak as noted above, the numerous third-party uses of the terms "HEALTH," "HEALTHY" and "CHOICE" further narrow each of the parties' rights with respect to their respective marks. (emphasis added) (Citations omitted).

(Defendant's (sealed) exhibit 535 — ConAgra's Proposed Ultimate Findings of Fact and Conclusions of Law, at p. 38, ¶ 47).

ConAgra responds by suggesting that when it was sued by Comsource, ConAgra reasonably asserted that the Comsource mark was entitled to only narrow protection because the mark had received little attention, nominal promotion, and it was being utilized essentially within the food service industry, such as hospitals and nursing homes. Plaintiff's post-trial brief at 19-20. Accordingly, ConAgra argues that, since Comsource's use preceded ConAgra's, ConAgra was forced to use the mark PROFESSIONAL CHOICE for the food service industry. Id. at 20. ConAgra then argues that it saw the value of the HEALTHCHOICE mark for all food products, retail and food service, and, accordingly, ConAgra purchased Comsource's rights for 3/4 of a million dollars. ConAgra concludes by arguing the Comsource litigation establishes only that the mark HEALTHY CHOICE is a very valuable mark.

The distinctions that ConAgra makes regarding the Comsource litigation do not make much of a difference. Obviously, ConAgra thought its mark was valuable, as ConAgra points out. But this does not, in and of itself, mean very much. Clearly, laudatory and descriptive marks may be valuable precisely because they preclude, if they are enforced, other competitors from using a common method of describing the subject product. Still further, while the fact that Comsource sought to market its products primarily within the food service industry is relevant, thus suggesting the context in which ConAgra might argue that "narrow" protection should be accorded the competing mark, this distinction does not explain ConAgra's contention that the HEALTHCHOICE mark was "intended to convey a meaning of healthy products which represent a good choice for consumers" resulting in a "laudatory mark," which is " inherently [a] weak mark" (emphasis added). It would appear that the mark is laudatory, and ConAgra so argued, whether in the context of the fairly narrow market such as food service, or in the broader market in which ConAgra and Hormel now joust. Stated differently, it seems to me that ConAgra admitted that its mark, and Comsource's mark, are "puffing" marks, and marks which "puff" would appear to do so regardless of market definition.

In summary, applying the first prong of the first SquirtCo test, I conclude that ConAgra's mark is relatively weak because it is more laudatory and descriptive than it is suggestive. I come to this conclusion for the following reasons: (a) the laudatory and descriptive use of the HEALTHY CHOICE mark in ConAgra's advertising; (b) the testimony of both of the expert linguists that HEALTHY CHOICE is a laudatory term; (c) the third party use and registrations indicating that the mark is laudatory and descriptive rather than suggestive, and (d) ConAgra's admissions of the laudatory and descriptive nature of the HEALTHCHOICE mark made in the Comsource litigation.

(2)

The second prong of the "strength" test requires an examination of the marketplace recognition value of the mark. 1 J. McCarthy, Trademarks and Unfair Competition § 11:25, at 509-10. In this circumstance a mark can start off weak, but become strong in the marketplace, such as through substantial expenditure of the promoter's resources for advertising. On the other hand, a mark may be quite distinctive in the abstract, but never realize potential in the marketplace.

According to ConAgra it has invested in excess of 200 million dollars in the HEALTHY CHOICE brand name (plaintiff's (sealed) exhibits 225-228). This includes about 194 million dollars in marketing, trade, and sales expenditures, and about 20 million dollars in plant, property and equipment. Since September of 1989, HEALTHY CHOICE products have been advertised on national television using various celebrities (defendant's exhibit 531). Promotion of the HEALTHY CHOICE line to consumers has also included freestanding inserts in newspapers, which often provide coupons which allow consumers to purchase the product for a reduced price (plaintiff's exhibits 163 and 248). ConAgra employs at its headquarters 22 to 25 full-time research and development personnel, 25 full-time marketing personnel, 35 full-time sales personnel, 2 full-time financial personnel, and 8 full-time market research personnel devoted to the development and marketing of the HEALTHY CHOICE line (filing 94, TR 407:22-25; 408:1-21). These numbers do not include people employed at the production plants (filing 94, TR 408:22-25).

Gross sales of HEALTHY CHOICE food products for the fiscal year ending in May of 1988 totalled approximately 21 million dollars (plaintiff's (sealed) exhibit 225), increasing to 202 million dollars in May of 1990 (plaintiff's (sealed) exhibit 226), increasing to 352 million dollars in May of 1991 (plaintiff's (sealed) exhibits 227 and 228), and projected at 500 million dollars for the fiscal year ending May of 1992 (plaintiff's (sealed) exhibit 224). HEALTHY CHOICE products are currently being sold in about 26,000 to 28,000 retail outlets nationwide and about 23% of all the households in the country have tried HEALTHY CHOICE products, which, according to a ConAgra executive, is the highest level of household penetration that any prepared food product has achieved (filing 94, TR 404:22-25; 405:1-13).

Advertising surveys indicate that 47% of a general sample of adults were aware of the brand name HEALTHY CHOICE (filing 94, TR 347:19-22). ConAgra has expended more than 50 million dollars on a national advertising campaign for HEALTHY CHOICE (filing 94, TR 348:1-4). ConAgra's advertising efforts have resulted in its chief executive officer receiving the Marketing Man of the Year Award from Advertising Age, and the ConAgra executive in charge of the HEALTHY CHOICE brand was awarded Brand Manager of the Year Award from Food and Beverage Marketing (filing 94, TR 349:14-25).

Hormel responds by suggesting that ConAgra is in effect "hiding the ball." Hormel argues, and the pretrial conference order confirms, that the point of competition at issue in this case is "shelf-stable products" and not foods in general or refrigerated foods specifically (order on final pretrial conference, filing 48, at ¶ D "Controverted and Unresolved Issues"). Moreover, as Hormel points out, the entire focus of this case has been on the likelihood of confusion between ConAgra's shelf-stable cup products and Hormel's shelf-stable cup products. From this point, Hormel then argues, among other things, that it is appropriate to judge whether the HEALTHY CHOICE brand has been successful in the sense of brand awareness in the shelf-stable market, and not the general food market or the refrigerated market. Hormel asserts that if the inquiry is so properly limited, ConAgra has virtually no brand awareness in the shelf-stable market.

In support of its argument, Hormel points out that the chief executive officer of ConAgra testified that the HEALTHY CHOICE success has been limited to frozen foods only (filing 92, TR 73:8-17). Hormel argues that ConAgra's own documents show that, despite whatever success ConAgra has achieved in the frozen food area, such success will not transfer over to the shelf-stable market because the shelf-stable market is comprised primarily of consumers who use canned food, not consumers who use frozen foods.

In the frozen dinner and entree category for which ConAgra believes it has achieved success, its February 8, 1991, consumer research report shows that "[r]elative to competitive advertising, advertising for Healthy Choice has had about average success in terms of reach in creating a memorable impression." (Plaintiff's (sealed) exhibit 190 at 22).

For example, a ConAgra promotion plan labelled "AGNSS" for the period June 1, 1990, through September 30, 1991, states in pertinent part that frozen food buyers will not be purchasers of shelf-stable products:

The current user base of shelf stable products is comprised primarily of consumers of canned food. They are very price-value conscious and are also drawn to shelf stable for the convenience it offers. They are heavy users of coupons.

Buyers of frozen food are not being drawn to this category in great numbers at this time. Heavy buyers of frozen food do not even walk down the canned food aisle in the grocery store so have little exposure to the category. Their need for convenience is being satisfied through their frozen food purchases. Also, they tend to be less price conscious than their canned food buying counterpart.

. . . . .

An important characteristic of our target audience is that nutrition is not of major concern. Also, it has been determined that our target audience is driven by convenience and price-value.

(Defendant's (sealed) exhibit 569 at 3 and 6 (stamp # s 00026242 and 00026245)) (Emphasis in the original).

In a similar marketing plan dated May 8, 1990, ConAgra recognized the same point that purchasers of frozen food would not be the shelf-stable purchasers and would not see or consider the shelf stable product:

HEALTHY CHOICE frozen users (and frozen users in general) do not normally use canned foods, so they will not see/consider the product in the store. HEALTHY CHOICE frozen users are older than the shelf stable category user. ConAgra does not want to cannibalize sales of frozen.

Consumers purchase S.S. based upon convenience, but convenience is not a point of difference for HEALTHY CHOICE. (Convenience will draw people to the category, but not specifically to HEALTHY CHOICE.)

Consumers in the S.S. category are less concerned about their health and about the health aspect of canned food than are consumers in the frozen category.

(Defendant's (sealed) exhibit 571 at p. 3 (stamp # 00028108)).

At the time suit was filed, HEALTHY CHOICE shelf-stable products were being marketed only in four test markets — Denver, Colorado, Des Moines, Iowa, Omaha, Nebraska, and Miami, Florida. By comparing profit and loss statements of ConAgra it is fairly easy to see that ConAgra has expended relatively insignificant resources in terms of advertising in the shelf-stable category (compare plaintiff's (sealed) exhibits 225, 226, and 227 dealing with HEALTHY CHOICE frozen foods, and plaintiff's (sealed) exhibit 228 which deals with HEALTHY CHOICE shelf-stable food) (filing 94, TR 456:6-15). For example, advertising expenditures for HEALTHY CHOICE during fiscal year 1991 totaled approximately 44 million dollars (plaintiff's (sealed) exhibit 227, run 7-09-91, total advertising), while advertising expenditures for the shelf-stable market during fiscal year 1991 totaled approximately $761,000 (plaintiff's (sealed) exhibit 228, run 7-09-91, total advertising).

When one endeavors to analyze the strength of the mark in the marketplace, one is endeavoring to assess "degree of recognition in the minds of the relevant customer class." 1 J. McCarthy, Trademarks and Unfair Competition § 11:1, at 434 (emphasis added) (footnote omitted). Thus, as Professor McCarthy indicates, "[m]any cases decide a trademark conflict for the junior user by saying that while plaintiff's mark is strong in its own market, it is weak in defendant-junior user's market." Id. § 11:24 at 506 (footnote omitted).

Even if one were to assume that ConAgra's mark was strong for frozen foods, it would be inappropriate, as a matter of fact, to assume that the mark is also strong in the shelf-stable part of the market. Since the evidence in this case suggests that shelf-stable product consumers are not likely to be the same consumers who purchase frozen food products, ConAgra's reliance upon market penetration in the frozen food section of the supermarket, and related monetary expenditures, does not establish that its mark HEALTHY CHOICE is likewise "strong" in the shelf-stable area of the supermarket. While one might assume that there might be some "transfer of equity" of the mark HEALTHY CHOICE as applied to frozen foods to the mark HEALTHY CHOICE as applied to shelf-stable food products, such an assumption cannot be quantified in a meaningful sense. Therefore, such an assumption cannot serve as persuasive evidence of the strength of the HEALTHY CHOICE mark as applied to shelf-stable food products.

ConAgra argues that the constant references by Hormel to the HEALTHY CHOICE frozen dinners "as competitive product" in the development of HEALTH SELECTIONS establish that Hormel's shelf-stable products compete with ConAgra's frozen food product, citing various exhibits, mostly generated by Hormel's advertising agency. Plaintiff's post-trial brief at 23 (citing plaintiff's exhibits 34, 35, 36, 38, 39, 43, 44, 45, 46, 48 and 49). I have reviewed these exhibits and I do not agree that they prove that Hormel thought frozen products and shelf-stable products were fungible in the minds of consumers. Moreover, these exhibits certainly do not establish that ConAgra's mark was well recognized in the shelf-stable market.

In conclusion, (a) given the recent introduction of the HEALTHY CHOICE mark into the shelf-stable section of the grocery store, (b) given the fact that the mark HEALTHY CHOICE as applied to shelf-stable food products was only being test marketed at the time of this litigation, (c) given the fact that ConAgra's marketing expenditures for the HEALTHY CHOICE mark in the shelf-stable market have been relatively insignificant, and (d) given the fact that purchasers of frozen foods are not likely to be purchasers of shelf stable products, I conclude that the mark HEALTHY CHOICE in the relevant market — the shelf-stable market — is not strong.

ConAgra points out that the competition in the shelf-stable market is competition with respect to essentially identical products. I do not in any way discount the importance of this point. Rather, I believe the question of competitive proximity should be considered under the third SquirtCo factor. The fact that the goods are in competitive proximity does not serve, however, to prove that the senior user's mark is "strong" in the relevant market.

B. SIMILARITY OF THE MARKS

The second factor to be considered is the similarity of the marks. In this regard, "the use of identical, even dominant, words in common does not automatically mean that two marks are similar." General Mills, 824 F.2d at 627 (OATMEAL RAISIN CRISP for breakfast cereal not confusingly similar to APPLE RAISIN CRISP for breakfast cereal). Rather, one must examine the "similarities of sight, sound, and meaning" to determine the "overall impression created by the marks." Id. One does "not merely compare individual features." Id.

ConAgra concedes that this is not a "trade dress" lawsuit. Plaintiff's post-trial brief at 5. However, ConAgra argues, and Hormel does not disagree, that when assessing the likelihood of confusion between two trademarks, similarity between the trademarks "must be assessed in terms of its effect upon prospective purchasers, and among the factors which could reasonably be expected to have an impact on, and be remembered by, potential consumers, the overall packaging concept is especially important." Lever Bros. Co. v. American Bakeries Co., 693 F.2d 251, 257 (2d Cir. 1982). Thus, as Professor McCarthy has stated, "the total effect of `background' such as shape, color and design of packaging and containers will be compared to see if that background is likely to intensify or dilute any confusion between marks." 2 J. Thomas McCarthy, Trademarks and Unfair Competition § 23:18, at 99 (2nd ed 1984) (citing, David Sherman Corp. v. Heublein, Inc., 340 F.2d 377 (8th Cir. 1965)).

In this connection I note that in an infringement action, "the defendant's background trade dress may be sufficient to distinguish the goods even though, in the abstract, the marks might be said to be confusingly similar." Id. at 101. And, very importantly, if "both plaintiff and defendant sell their products in a trade dress which is in the public domain and free for all to use . . . plaintiff cannot use this similarity as a ground for likelihood of confusion." Id. (footnote omitted).

(1)

I begin with a comparison of the two products as they would appear to the consumer — by sight. This comparison may be seen, by way of example, by comparing defendant's exhibits 536 and 545, respectively a HEALTHY CHOICE "micro cup" and a HEALTH SELECTIONS micro cup.

ConAgra marketed a micro cup with a "sleeve" overwrap. This sleeve was initially used on HEALTHY CHOICE cups because ConAgra wanted the packaging to be close to the frozen entree packaging and because ConAgra was having problems in reproducing a photograph onto the foam micro cup. In December of 1990, ConAgra decided to discontinue the use of the cardboard sleeve. ConAgra still uses the "sleeve" product and will do so until its supplies are exhausted. ConAgra also uses the product without a sleeve. If one wanted to examine ConAgra's HEALTHY CHOICE product without a sleeve, one could examine plaintiff's exhibit 270. Whether one compared defendant's exhibit 536, the HEALTHY CHOICE micro cup with a sleeve, and defendant's exhibit 545, the HEALTH SELECTIONS micro cup, or whether one compared plaintiff's exhibit 270, the HEALTHY CHOICE micro cup without a sleeve, with defendant's exhibit 545, the HEALTH SELECTIONS micro cup, the factual setting is about the same. Except as indicated in the text, I attach no independent significance to the different HEALTHY CHOICE presentations.

ConAgra's microwaveable shelf-stable products display the HEALTHY CHOICE mark in white print on a dark green cup label and sometimes, where the sleeve product is used, on dark green cardboard oversleeves. There is some fairly significant yellow coloration used on the ConAgra label. The oversleeve product partially obscures the microwaveable cup from view when displayed on a grocery store shelf, but the cap of the cup is green and the cap can be partially observed. On the ConAgra product, the figure of a jogger is displayed between the word HEALTHY and the word CHOICE. ConAgra's name is not used. ConAgra's HEALTHY CHOICE mark is printed on its product labels with the initials "H" "C" in a letter size larger than the lettering used in the remainder of the words.

Hormel's HEALTH SELECTIONS microwaveable cup has a predominantly white background color. Hormel does not use a figure or symbol. The Hormel "house" trademark using the word HORMEL is shown in logo form in red print framed in green and black border directly over and associated with the green-colored mark HEALTH SELECTIONS. The HEALTH SELECTIONS cup has a green cap and uses green as a background for the presentation of certain information, such as calories and fat. Hormel's HEALTH SELECTIONS mark is printed and displayed in uniform size in relatively tall and narrow capital letters.

I conclude that from the perspective of "sight" the trade dress is not confusingly similar, for among other reasons, the following: (a) the fairly prominent use of the HORMEL logo and name compared to ConAgra's use of the symbol of a jogger without ConAgra's name; (b) the green background with fairly prominent yellow coloration on the ConAgra product compared with the white background with fairly prominent green coloration of the Hormel product; (c) the difference in type style (particularly ConAgra's capitalization of "H" and "C") and color for the marks "HEALTHY CHOICE" and "HEALTH SELECTIONS." And, even if there are similarities, the similarities are in the use of the color green and the word "Health" or its derivatives. ConAgra has conceded that it has no proprietary right to the color green (filing 99, TR 1330:16-25; 1331:1-9). ConAgra has also conceded that it has no proprietary right to the word "Healthy" or its derivatives (defendant's exhibit 540, at ¶ 3 "Consent Decree"); filing 85, Hughes deposition, TR 89:2-25; 90:1-24).

This conclusion is buttressed by the recently concluded litigation involving ConAgra and Campbell Soup Co., Campbell Soup Co. v. ConAgra, Inc., No. 90-4965 (D.N.J.). In that matter, ConAgra alleged that the trademark HEALTHY REQUEST used by Campbell Soup Co. for health-oriented soup products was confusingly similar to the HEALTHY CHOICE mark. ConAgra specifically alleged that Campbell's HEALTHY REQUEST mark was "similar in appearance and identical in meaning" to HEALTHY CHOICE (defendant's exhibit 547). Notwithstanding this alleged similarity, ConAgra agreed to the entry of a consent decree which allowed Campbell to continue to market its product and use its mark, HEALTHY REQUEST, subject to certain conditions including (a) that the type style used for the HEALTHY REQUEST mark must be different than that used for HEALTHY CHOICE, (b) that Campbell not use a figure confusingly similar to the jogger on the HEALTHY CHOICE logo, (c) that Campbell not use the color green as the "predominant color" in the background of its labels, and (d) that Campbell would use the name "Campbell's" in its promotional and advertising material (defendant's exhibit 540, ¶ 1 "Consent Decree"). In my judgment, Hormel's HEALTH SELECTIONS product satisfied the principle concerns expressed by ConAgra in the consent decree involving Campbell Soup Co.

(2)

Both ConAgra and Hormel commissioned their linguistics experts to conduct a phonology or sound analysis of the respective marks HEALTHY CHOICE and HEALTH SELECTIONS. Dr. Geis, ConAgra's expert, testified that "differences" in "sounds exist to signal differences in meaning." (filing 96, TR 619:1-5). In fact, Geis believed that the fundamental proposition in linguistics was that differences in sounds exist to signal differences in meaning (filing 96, TR 619:1-5). Dr. Shuy, Hormel's expert, also agreed that phonology is very important for understanding because "[it is] the noise you make that meaning is conveyed through." (filing 98, TR 954:12-18).

Dr. Geis acknowledged that there is a difference in sound between the mark HEALTHY CHOICE and HEALTH SELECTIONS, but he believed it appropriate to "place special weight on the initial segments" while not suggesting "that the ends are unimportant." (filing 95, TR 564:15-20) Dr. Shuy came to just the opposite conclusion, finding that the mark HEALTH SELECTIONS has 13 sounds, 9 of which differ from the sounds found in the mark HEALTHY CHOICE (filing 98, TR 953:21-25; 954:1-5).

If phonology is indeed as important as the linguists seem to suggest, I was more persuaded by Dr. Shuy's straightforward testimony than I was by Dr. Geis' somewhat rambling testimony. Moreover, from a "common sense" point of view, it seems to me that the mark HEALTH SELECTIONS sounds significantly different from the mark HEALTHY CHOICE.

In this connection, it should also be noted that ConAgra has acknowledged in the Campbell Soup litigation that "it does not have the exclusive right to use the word "healthy" or its "derivatives" (defendant's exhibit 540, at ¶ 3 "Consent Decree"). Thus, in terms of sound (or for that matter sight or meaning) the similarities between "healthy" and "health" in the respective marks are in the public domain for all to use. A ConAgra executive confirmed this point (filing 85, Hughes deposition, TR 89:2-25; 90:1-24).

(3)

According] to Dr. Geis, ConAgra's linguistics expert, the marks HEALTHY CHOICE and HEALTH SELECTIONS "have virtually identical meanings" (filing 95, TR 595:2-4). Indeed, even Dr. Shuy, Hormel's expert, agreed that the words "choice" and "selection" are regarded in dictionaries as synonyms, that he is not aware of any dictionary that does not regard the words as synonyms, and that Roget's International Thesaurus, a good source of synonyms according to Shuy, regards "selection" as a synonym for "choice" (filing 98, TR 997:8-25; 998:1-13). Dr. Shuy believed, however, that the word "choice" required an animated agent/subject (i.e., an actor) whereas "selections" does not, and the word "choice" refers to a limited array, whereas "selections" refers to a larger array (filing 98, TR 957:17-25; 958:1-25; 959:1-25; 960:1). But even Dr. Shuy, not intentionally, used the synonym "choices" when endeavoring to define the difference between "choice" and "selections":

That is the principle I'm getting at here is that the word "choice" has a more limited array of possibilities connected with it than does the word "selection" and even more so the word "selections" which indicates an even broader array of choices is available. An unfortunate use of word there (filing 98, TR 959:21-25; 960:1) (emphasis added).

Since every single dictionary and thesaurus brought to my attention (plaintiff's exhibits 251, 309 310) regard "selection" as a synonym for "choice", notwithstanding the qualifications suggested by Dr. Shuy, I conclude that the mark HEALTH SELECTIONS means essentially the same thing as the mark HEALTHY CHOICE.

Dr. Shuy also suggested differences between the marks based on such things as syntax, among others. Normally speaking, however, "meaning" can be determined from standard dictionary definitions. 2 J. McCarthy, Trademarks and Unfair Competition § 23.8, at 67 (footnote omitted).

C. COMPETITIVE PROXIMITY

The products in this case compete with each other "head-on." As noted by the Eighth Circuit Court of Appeals in SquirtCo:

Where the products are closely related, less similarity in the trademarks is necessary to support a finding of infringement. (citing David Sherman Corp. v. Heublein, Inc., 340 F.2d 377, 380 (8th Cir. 1965)). Competitive proximity is one factor to be considered, even though infringement may be found in the absence of direct competition. H.M.H. Publishing Co. v. Turbyfill, 330 F. Supp. 830 (M.D.Fla. 1971).

628 F.2d at 1091.

D. INTENT TO CONFUSE

As the United States Court of Appeals for the Eighth Circuit stated in SquirtCo, "[i]ntent on the part of the alleged infringer to pass off its goods as the product of another raises an inference of likelihood of confusion, but intent is not an element of a claim for trademark infringement." Id. at 1091. In so stating, it must be recognized that the place of intent in this analysis is to determine whether there is likely consumer confusion, and the effort in determining the defendant's state of mind is only an aid to the goal of determining consumer confusion, not an end unto itself. 2 J. McCarthy, Trademarks and Unfair Competition § 23:35, at 155.

ConAgra alleges that the evidence supports a finding of intent on the part of Hormel to trade upon the good will that ConAgra had established in its HEALTHY CHOICE mark. ConAgra's argument is essentially twofold. First, ConAgra argues that Hormel had made the decision to use the mark HEALTHY SINGLES on its line of shelf-stable microwaveable cup products, and that decision was reversed immediately after Hormel received word that ConAgra was introducing a line of shelf-stable products under the mark HEALTHY CHOICE. Second, ConAgra argues that there are a string of "coincidences," which are in fact not coincidences, establishing that Hormel intended to unfairly use ConAgra's good will in the HEALTHY CHOICE mark. I shall examine these contentions in turn.

(1)

ConAgra argues that Hormel decided to use its HEALTH SELECTIONS mark, instead of its HEALTHY SINGLES mark, after it learned that ConAgra was entering the shelf-stable market. This argument is predicated on the following facts.

On September 18, 1990, Brent Payne, a Hormel product manager dealing with the HEALTH SELECTIONS product line, wrote Hormel's advertising agency indicating that Mr. Bross, Payne's boss, wanted to move forward quickly on Hormel's HEALTHY SINGLES brand (plaintiff's exhibit 116). On that same date Payne wrote Ms. Brennan, who was responsible for Hormel's label design, and indicated that "upper management" wanted to move forward with the HEALTHY SINGLES product line, and he asked that Hormel's advertising agency prepare a mockup of the labels for the HEALTHY SINGLES brand (plaintiff's exhibit 117). On September 19, 1990, Hormel sent an "intent-to-use" application for HEALTHY SINGLES to the United States Patent and Trademark Office by overnight mail, for products matching the description of the HEALTH SELECTIONS product line (plaintiff's exhibit 307). On September 20, 1990, Hormel's district manager, Dale Heine, sent a telex to Rick Bross and Brent Payne stating, "As I notified you today, HEALTHY CHOICE has 8 items in the micro line and the same 8 items in a 15 oz. can . . . we will try and get samples and send when possible." (plaintiff's exhibit 119). On September 20, 1990, Ms. Brennan corresponded with Mr. Payne, indicating that the brand name for the micro cup line would be "HEALTH SELECTIONS" (plaintiff's exhibit 120).

Hormel responds that the uncontroverted evidence establishes that Hormel chose HEALTH SELECTIONS as the brand name for the subject line, over HEALTHY SINGLES, before it learned of ConAgra's introduction of its HEALTHY CHOICE shelf-stable product line. Hormel relies upon the following evidence.

On September 19, 1990, Rick Bross, grocery products director of marketing, called a meeting with certain marketing managers and representatives of Hormel's advertising agency (filing 66, Bross (sealed) deposition, TR 172:1-19). The purpose of the meeting was for Bross to make a final decision as to the name on the new shelf-stable product and he had reached the conclusion that he would recommend the name HEALTH SELECTIONS over the name HEALTHY SINGLES by the time of the meeting (filing 66, Bross (sealed) deposition, TR 172:7-15). Brent Payne, the product manager who reported to Bross, had thought up the name HEALTHY SINGLES for this product line, and he favored it and said so at the meeting (filing 75, Payne (sealed) deposition, TR 36:6-25; 37:1-3; 47:6-25; 48:1-25; 49:1-25; 50:1-23). According to Payne, "There were more in favor of HEALTH SELECTIONS than HEALTHY SINGLES" (filing 75, Payne (sealed) deposition, TR 49:11-14).

Bross felt that the name HEALTHY SINGLES was far too limiting in that there was an implicit suggestion in the name that the product only had single servings and there was a suggestion that it would only appeal to consumers who were unmarried (filing 66, (sealed) Bross deposition, TR 176:4-12). Although Payne still liked the name HEALTHY SINGLES, he "understood those concerns and I accepted them [as] very legitimate" (filing 75, Payne (sealed) deposition, TR 43:14-15). As a result, either on September 19, 1990, or early the next day, September 20, 1990, Bross made his recommendation to the group vice president, and the group vice president agreed and authorized the use of the name HEALTH SELECTIONS (filing 66, Bross (sealed) deposition, TR 177:12-25; 178:1-25; 179:1-25; 180:1-25; 181:1-25; 182:1-25; 183:1-7). Bross was confident his meeting with the group vice president was prior in time to his receipt of the telex from Heine about ConAgra's expansion into the shelf-stable area of the supermarket (filing 66, Bross (sealed) deposition, TR 198:13-16).

I do not think the foregoing evidence proves that Hormel intended to trade upon ConAgra's name so as to confuse the public. First, Mr. Bross testified that he did not know of ConAgra's entry into the shelf-stable market until after he had communicated his decision to the group vice president. Second, Hormel had applied for and been granted the mark HEALTH SELECTIONS in March, 1990. Product development of the line of products ultimately marketed under the HEALTH SELECTIONS mark began in earnest in June of 1990, and on August 17, 1990 (defendant's exhibit 510), Hormel filed an additional application which had the effect of broadening the scope of goods and encompassing the entire existing product line for the registration of HEALTH SELECTIONS. Third, it would make little sense to base a brand name on the supposed name HEALTHY SINGLES when there had yet to be trademark approval, particularly when one had an already approved trademark in the form of HEALTH SELECTIONS. In this connection, I was especially convinced by the testimony of Eric Brown, a Hormel executive, when he stated:

I think it's important to understand that as competitors who are entering this marketplace, there's a time to move forward and to get the product on the shelf. And at this time, in the fall of September — or in the fall of the year here, we were ready to proceed. We had formulations coming out of our research and development area. And we had an approved, registered trademark available for our use that, in fact, was a better name [than HEALTHY SINGLES] for us to use than the name being proposed by a product manager. Its almost obvious, from a management standpoint, that you would move with a known commodity, a trademark that's already got registration, and you can get on the market in the next few months to move with that trademark, makes so much more sense than to start an entire new trademark application procedure in a long process of trying to approve a name that we don't even know whether or not will make it. The [HEALTH SELECTIONS] name was available for our use and we used it.

(filing 96, TR 701:15-25; 702:1-10)

(2)

ConAgra next argues that certain "coincidences" amounted to a pattern of conduct indicative of bad faith. I shall briefly discuss each of these allegations in turn. I conclude that these "coincidences" are not evidence of an intent to confuse the public.

First, ConAgra notes that Hormel employed the same design firm that designed the HEALTHY CHOICE frozen dinner packaging. It appears, however, that the design firm had had a fairly long-standing relationship with both Hormel and ConAgra and the design firm did not distinguish between ConAgra or Hormel in terms of which was the more important client (filing 61, Skartvedt deposition, TR 16:5-25; 17:1-6). The witness for the design firm testified that no reference was made to ConAgra's HEALTHY CHOICE product during the design firm's development of the HEALTH SELECTIONS cup project (filing 61, Skartvedt deposition, TR 78:14-17), except for general discussions about all competitors' products, such as to review the back of panels for the claims that various competitors were making regarding such things as amounts of cholesterol (filing 61, Skartvedt deposition, TR 83:1-13). The design firm witness specifically testified that the design firm intended to make the Hormel packaging distinctive so as to distinguish the Hormel product from that of other competitors (filing 61, Skartvedt deposition, TR 159:3-16).

Next, ConAgra argues that Mr. Bross of Hormel thought of the mark HEALTH SELECTIONS in February or March of 1989, which was close in time to when ConAgra first announced its HEALTHY CHOICE line. As I read the testimony of Mr. Bross, however, he is uncertain as to when he thought of the name except to say that it was sometime before March of 1989 (filing 66, Bross (sealed) deposition, TR 107:14-25; 108:1-25; 109:1-5).

Third, ConAgra argues that Hormel conceived the mark HEALTH SELECTIONS before Hormel had done any development work with regard to the product which the mark would be applied to. Hormel responds by stating there is nothing unusual about this procedure and that in any event the HEALTH SELECTIONS development was delayed for legitimate business reasons due to Hormel's prior introduction of its TOP SHELF products in 1986, its KID'S KITCHEN products in 1987, and its MICRO CUP line in 1988 and "[s]o it was very much a part of our plan, but it was just a matter of getting it all done" (filing 96, TR 671:22-24).

Still further, ConAgra argues that Hormel "coincidentally" selected the same shade of green as that used by ConAgra on its HEALTHY CHOICE product. However, it appears that, except for nutritional data, the basic design for the HEALTH SELECTIONS package was completed sometime prior to October 26, 1990 (filing 67, Brennan (sealed) deposition, TR 104:6-11). It appears that Hormel first obtained samples of HEALTHY CHOICE shelf-stable microwaveable cup products sometime around December 6, 1990 (plaintiff's exhibit 166).

The design firm evidently did examine a HEALTHY CHOICE frozen dinner package, not the microwaveable cup product, but then only for the purpose of examining nutritional information (filing 61, Skartvedt deposition, TR 82:22-25; 83:1-11). The design firm witness did not recall seeing a HEALTHY CHOICE shelf-stable product at Hormel's office or at the design firm's office (filing 61, Skartvedt deposition, TR 84:10-15).

Still further, the HEALTHY CHOICE product (plaintiff's exhibit 536), containing a fair amount of green, and the HEALTH SELECTIONS product (defendant's exhibit 545), also containing a fair amount of green but less than the HEALTHY CHOICE product, can profitably be compared with the HEALTHY RECIPES product (defendant's exhibit 552), which also contains a significant amount of green as part of its labeling. To the extent that trade dress is an appropriate consideration, I have considered trade dress in connection with the second SquirtCo factor. However, to the extent ConAgra claims that the selection of the color green is indicative of an intent to confuse, I think the HEALTHY RECIPES package, which uses a substantial quantity of green, indicates that no fair inference can be drawn from the selection of the color green by Hormel. ConAgra has conceded that the HEALTHY RECIPES trademark as it appears on its package is not confusingly similar to the HEALTHY CHOICE mark (filing 85, Hughes deposition, TR 89:2-25; 90:1-24).

Fifth, ConAgra complains that Hormel actively monitored HEALTHY CHOICE product sales and referred to HEALTHY CHOICE packaging in the package design phase of Hormel's HEALTH SELECTIONS product line and this consequently evidences bad faith. Hormel responds by stating that Hormel, like ConAgra, needed to, and did, keep abreast of products and activities of other food companies. Indeed, in one of the marketing plans of ConAgra for HEALTHY CHOICE shelf-stable meals there is evidenced a keen awareness on the part of ConAgra of the competitive positions, including advertising revenues expended, of ConAgra's competitors, including Hormel (defendant's exhibit (sealed) 567, at 7-8). Insofar as Hormel's reference to ConAgra's packaging is concerned, the evidence indicates, as described above, that Hormel examined HEALTHY CHOICE frozen dinners, not the micro cups, and then only for the purpose of examining nutritional information claims.

Finally, ConAgra argued that Hormel has "down-sized" the Hormel house logo HORMEL on the HEALTH SELECTIONS label. This indicates to ConAgra an intent to mislead. I have compared the products suggested by ConAgra (plaintiff's exhibits 267 and 272). Plaintiff's exhibit 267 is a HEALTH SELECTIONS cup manufactured by Hormel and plaintiff's exhibit 272 is another product in cup form manufactured by Hormel. ConAgra is correct in that the Hormel house logo HORMEL on plaintiff's exhibit 272 is larger than the Hormel logo on plaintiff's exhibit 267.

Hormel responds by suggesting that each cup, for each product line, had unique design priorities. For example, I note that plaintiff's exhibit 275, a KID'S KITCHEN cup, manufactured by Hormel, also has a down-sized Hormel logo when compared with plaintiff's exhibit 272, another cup product put out by Hormel. In fact, if one compares the HEALTH SELECTIONS cup (plaintiff's exhibit 267), the Hormel cup product upon which ConAgra relies (plaintiff's exhibit 272), and the Hormel KID'S KITCHEN cup, (plaintiff's exhibit 275), the Hormel logo on each product is prominently displayed, although on each cup the size of the logo varies.

Indeed, it appears that the design priorities for the label of the HEALTH SELECTIONS cup given to the design firm by Hormel were as follows, and in this order: (a) brand name; (b) Hormel identifier/logo; (c) appetite appeal shot; (d) variety identifier/net weight statement; and (e) nutritional claims (defendant's exhibit 598). Mr. Brown for Hormel testified that it is "a matter of policy that the Hormel logo will appear prominently" on virtually every product manufactured by Hormel, with very few exceptions, such as where Hormel uses a licensing agreement to market another product, as in the case of ChiChi Salsa (filing 96, TR 661:15-25; 662:1-8). I do not think the size of the Hormel logo on the relevant product is indicative of an intent to confuse the public.

In this regard, I note that when ConAgra settled with Campbell's it required Campbell's to use its Campbell's logo, but permitted Campbell's to use any size so long as the logo was "readily visible" (defendant's exhibit 540 at 1, ¶ 1(a)). Certainly the Hormel logo is "readily visible" on the HEALTH SELECTIONS cup. In fact, it is placed in the center of the cup just above and centered between the words HEALTH and SELECTIONS (plaintiff's exhibit 267).

In fact, Dr. Jacoby, plaintiff's consumer survey expert, may have inadvertently proven the prominence of the Hormel logo. As discussed in more detail later, Dr. Jacoby endeavored to test consumer confusion. His study appears as plaintiff's exhibit 255. In group 4, which was Miami, Florida, consumers were shown a HEALTH SELECTIONS product and asked to buy the specific product (and two others) that had been shown to the consumer by an interviewer who had given the consumer 7 dollars to make the various purchases. In group 4 the consumers were shown a HEALTH SELECTIONS product, but only HEALTHY CHOICE was carried in the supermarket (plaintiff's exhibit 255 at 8 (stamp # 10)).

Despite the fact that the consumer was to purchase only "the specific products that I just showed you" (plaintiff's exhibit 255 at 1 (stamp # 033) (main questionnaire)), 91 percent of the respondents (134 of 147) brought back a Hormel product, which was not a HEALTH SELECTIONS product since it was not on the shelf (plaintiff's exhibit 255, at table 5, (stamp # 057)). The purchase of another Hormel product would appear to show the prominence of the Hormel logo. Indeed, Dr. Jacoby acknowledged the possible prominence of the Hormel logo: "They were shown Hormel and we meant them to pay attention to HEALTH SELECTIONS. They may have paid attention to Hormel." (filing 93, TR 243:16-25; 244:1-2) (emphasis added).

In summary, I do not believe the claimed "coincidences" suggested by ConAgra establish an intent on the part of Hormel to confuse the public.

E. ACTUAL CONFUSION

Both ConAgra and Hormel introduced expert witness testimony and consumer surveys in an effort to determine whether consumers would actually be confused by the Hormel and ConAgra products. ConAgra's evidence takes the form of expert witness testimony and a consumer survey and report conducted by Dr. Jacob Jacoby and a follow-up study by Dr. Leon Kaplan. Hormel introduced a consumer survey by Mr. John Bunge. Hormel also produced the expert testimony of Dr. Ivan Ross, who was engaged to examine ConAgra's surveys. There is no question that each of the four expert witnesses were well qualified.

Before addressing the consumer surveys, and the positions relative to each survey, it is well to make a number of observations. Since the subjective mental associations and reactions of prospective purchasers are often at issue in a trademark case, consumer surveys, designed to determine the existence of likelihood of confusion, are frequently conducted and received into evidence. 2 J. McCarthy, Trademarks and Unfair Competition § 32:46. As Professor McCarthy has indicated, while all properly conducted surveys share some of the same attributes, id. at § 32:50, "one must keep in mind that there is no such thing as a `perfect' survey." Id. at 779. Thus:

Like any scientific method related to statistics in the social sciences, every survey, no matter how carefully constructed and conducted, has some potential flaws somewhere. The proper approach is to view such evidence with some understanding of the difficulty of devising and running a survey and to use any technical defects only to lessen evidentiary weight, not to reject the result out-of-hand.

Id. (citing SquirtCo, 628 F.2d at 1091: "In evaluating survey evidence, technical deficiencies go to the weight to be accorded them, rather than to their admissibility.").

But as Professor McCarthy has noted, while survey evidence has been deemed important, "the courts are careful in opinions not to place all their evidentiary eggs in the survey basket." Id. § 32:55, at 791 (footnote omitted). Indeed:

Probably, a part of a lingering judicial skepticism about survey evidence can be laid at the feet of parties and their attorneys who, in a desperate search for some kind of evidence, offer, with a straight face, a haphazard, self-serving "survey." It may not be surprising that many judges view such purported "scientific evidence" with distaste. They know that the techniques of testing and sampling buyer reactions have been developed to a fairly high degree of accuracy. . . . As weak arguments detract from even a strong case, a weak survey may detract from even the strongest case of trademark infringement or unfair competition.

Id. at 791.

With these observations in mind, I turn to the various surveys and the critiques of the parties.

(1)

The Jacoby study, entitled Healthy Choice and Health Selections: Assessing the Likelihood of Marketplace Confusion appears as plaintiff's exhibit 255. I shall only describe the pertinent portions of this survey.

There are a host of potential criticismsthat can be directed at any survey. Because I find all of the surveys were sufficiently well done to merit consideration, I will not describe the many foundational prerequisites that each of the surveys satisfied in order to gain admissibility. I will also not attempt to discuss each criticism; rather, I shall discuss only those criticisms which seem especially pertinent.

The Jacoby study was designed to try to examine confusion from three perspectives, to-wit: as to product, as to source, and as to sponsorship. To test for these various possibilities, a consumer survey was designed wherein potential respondents were intercepted on their way into a supermarket that carried one or both of the HEALTHY CHOICE or HEALTH SELECTIONS products. After being "screened" for eligibility and agreeing to participate, the respondents were shown three products (one of which was either the HEALTHY CHOICE or HEALTH SELECTIONS product) and asked to go into the supermarket and purchase these three products (with monies provided by the interviewer) while doing their regular shopping. The respondents were asked to return with the products, at which time the consumers would be asked questions.

The main questionnaire proceeded as follows: Individuals consenting to participate were shown three products — French's Mustard, Heinz Tomato Paste and either HEALTHY CHOICE or HEALTH SELECTIONS. Respondents were told to look at the products as they would if they saw them in the supermarket and were thinking about buying them.

When the respondents indicated they were finished looking at the products, the interviewer put the products away and asked each respondent to purchase the three products while doing his or her normal shopping in the supermarket. The respondent was given 7 dollars to buy the products. They were also told that they would receive 10 dollars when they returned and answered some questions. Respondents were instructed that if the supermarket was out of the particular brand they had seen, "please don't substitute another. Just come back with the change."

When the respondents returned to the interviewer, the specific brand names of the products were recorded. The products were recorded in response to what became known as question 2. After all the products were taken from each respondent, the respondent was asked various additional questions.

The respondent was shown both products, HEALTHY CHOICE and HEALTH SELECTIONS, and asked if the two products "came from the same company or from different companies." This was known as question 3a. The respondents were given four choices: same company; different company; no opinion; or don't know.

Those respondents who did not know or had no opinion were told the products came from different companies. Then those respondents, and those other respondents who said the products came from different companies, were asked whether the different companies were connected or associated with each party in any way. This was known as question 3c.

These respondents had four choices: as probably connected or associated in some way; probably not connected or associated in any way; no opinion; or don't know. Those who said the products were connected were asked "what makes you say that." This question was known as question 3d.

Finally, all respondents, except those who said the products came from the same company in question 3a, were asked: "Would you say that in order to put out its version of microwaveable entree, the company that came second": probably did have to get a license or ask permission from the company that came first; probably did not have to get a license or ask permission from the company that came first; you don't have an opinion on that; or don't know. This last question was known as question 4.

The survey was administered to five groups in different geographic locations of the country. Fifty consumers in group one were shown HEALTHY CHOICE and the supermarket carried only HEALTHY CHOICE. The 50 consumers were then interviewed. One hundred fifty consumers in group two were shown HEALTHY CHOICE and only HEALTH SELECTIONS was carried by the supermarket. The 150 consumers were then interviewed. Thirty consumers in group three were shown HEALTHY CHOICE and the supermarket carried both HEALTHY CHOICE and HEALTH SELECTIONS. The 30 consumers were then interviewed. One hundred fifty consumers in group four were shown HEALTH SELECTIONS and the supermarket carried only HEALTHY CHOICE. The 150 consumers were then interviewed. Thirty consumers in group five were shown HEALTH SELECTIONS and the supermarket carried both HEALTHY CHOICE and HEALTH SELECTIONS. The 30 consumers were then interviewed.

Dr. Jacoby analyzed the resulting data. Dr. Jacoby purported to analyze three types of confusion.

The first type of confusion was so-called "product confusion" evidenced by responses to the initial shopping task referred to as question 2. Upon returning from the supermarket, the brand names of the products purchased by each respondent were written down for each of the three products they were asked to purchase. It was not possible to determine product confusion for respondents in group one, inasmuch as these people were shown HEALTHY CHOICE and then sent into a supermarket that carried only HEALTHY CHOICE. If a person brought back the "wrong product" — HEALTHY CHOICE or HEALTH SELECTIONS — he or she was counted as confused. Confusion ranged from 6% to 84%, with the weighted average across the four groups being about 40%.

The second type of confusion that Dr. Jacoby attempted to isolate was what he called "source confusion." This effort was made by asking question 3a: "These are 2 different brands of food that can be microwaved. Do you think that they come from the same company or from different companies, or don't you know?" In this circumstance, after the respondents had returned with their shopping task completed, and the products were taken from the respondents, the respondents were shown both HEALTHY CHOICE and HEALTH SELECTIONS. They were asked question 3a. If the response was "same company," the respondent was counted as confused. According to Dr. Jacoby, across the five groups, source confusion ranged from 15% to 36%, with the weighted average being about 23%.

The third area Dr. Jacoby attempted to test for was what he called confusion as to association and sponsorship. This was in reality two different, but related, types of confusion, which for analytical purposes was counted as one type of confusion and the data aggregated.

As to "association" confusion, respondents not exhibiting "source confusion" in response to question 3a either said or were told (in question 3b) that the two products "came from different companies." With this knowledge, the respondents were then asked (in question 3c): "Would you say that the companies that put out these products are probably connected in some way, are probably not connected in some way, or you don't have an opinion on that?" Those responding affirmatively to question 3c were then asked question 3d: "What makes you say that?" According to Dr. Jacoby, "most replied with an answer that was either `trademark or trade dress significant' — that is, the answer revealed that the respondent believed that the companies that put out the two brands were associated because of either the name and/or the trade dress." (plaintiff's exhibit 255, at 23 (stamp # 025)). Examples Dr. Jacoby gave were "`because they look exactly alike and are the same product,'" `"the packaging makes you believe that,'" and "`they look the same.'" ( Id.)

As to "sponsorship" confusion, all respondents (excluding those who said "same company" in response to question 3a) were also asked (in question 4) essentially whether the company that came second probably did not have to get a license or ask permission from the company that came first; probably did have to get a license or ask permission from the company that came first; or respondent did not have an opinion. If the person responded that one had to get a license or ask permission from the company that came first, the respondent was counted as confused.

Confusion as to association and sponsorship, according to Dr. Jacoby, was then determined by a proportion of all respondents in each group who answered one or both of these questions in a manner that indicated that they believed that either an association was present (question 3c) for either trademark or trade dress reasons (question 3d) or sponsorship was necessary (question 4). Association and sponsorship confusion ranged, according to Dr. Jacoby, from 14% to 42%, with the weighted average being about 23%.

Dr. Jacoby then endeavored to determine the "net" confusion. To preclude double counting, according to Dr. Jacoby, respondents who exhibited two or more types of confusion were only counted as being confused once. The weighted average across the five groups was about 69%, with the weighted average for groups one, three, four and five being 58% (group two being excluded because percentage of confusion shown with regard to question 2 being unusually high at 84%). Reproduced in tabular form, the results of Dr. Jacoby's study were as follows:Group Base Product — Q.2 Source — Q.3a Sponsorship Q.3c/4 NET # % # % # % # %

Table 1 Type of Confusion 1 48 NA NA 12 25 20 42 32 67 2 149 125 84 22 15 19 13 130 87 3 28 3 11 10 36 4 14 14 50 4 147 9 6 41 28 42 29 84 57 5 27 2 7 8 30 6 22 14 52 ---- ---- ---- ---- WEIGHTED AVERAGE: 39.6 23.3 22.8 68.7 (Plaintiff's Exhibit 255, at 20 (stamp # 022)).

(2)

I turn now to examine the criticisms advanced by Hormel with regard to Dr. Jacoby's study. Hormel raises essentially two criticisms of Dr. Jacoby's study. First, Hormel argues that the primary defect in Dr. Jacoby's survey, which defect renders the results meaningless, is the failure to inquire into each respondent's state of mind after a response was made to various survey questions. In other words, Hormel argues that with regard to all of the questions, Dr. Jacoby should have asked: "What makes you say that?" The second criticism of Dr. Jacoby's survey was that it did not contain a "control." I agree with Hormel that the criticisms advanced with regard to Dr. Jacoby's study are serious flaws in the study. However, while these flaws call into question the results, they do not in my opinion totally invalidate this study. Rather, I conclude that Jacoby's study does point to some level of confusion, but that level of confusion may not be caused by trademark infringement.

The "[w]hat makes you say that?" question is a typical question designed to isolate, and therefore explain, the real thought processes of the respondent who evidences confusion. In other words, this type of question is designed to determine whether a person is confused for relevant trademark reasons or for some other unrelated and therefore irrelevant reason. Well-designed studies typically employ the "[w]hat makes you say that?" question or some variation. Indeed, Professor McCarthy indicates that such a question is now a part of a standard survey format. 2 J. McCarthy, Trademarks and Unfair Competition § 32:50, at 778 ("What makes you think so?").

Dr. Ivan Ross, hired by Hormel to critique the ConAgra studies, essentially testified that the Jacoby study was meaningless for failure to ask the "state of mind" question because it was his belief that consumer behavior alone does not explain what motivated the consumer and thus the behavior, without a state of mind question, is inherently ambiguous (filing 99, TR 1129-1142; 1202-1211). Even Dr. Kaplan, another ConAgra expert, admitted that the state of mind question should have been asked following questions 3a and 4 in order to determine whether any respondents were confused for trademark relevant reasons:

Q. So there are people in the control group, as well as in the main group, who are confused for nontrademark reasons?

A. That's right.

. . . . .

Q. Isn't that the very reason for asking the "why" question, you want to know why both groups are confused for non-trademark reasons?

A. We have no disagreement . . . the "why" question would have been an asset and should have been asked, certainly in 3A and 4; I don't think we really disagreed on that.

(filing 98, TR 1086:15-25; 1087:1-2).

The significance of failing to ask the state of mind question is evident from the responses to the one and only state of mind question which was asked following question 3c (whether the companies were connected or associated). Dr. Jacoby recorded the responses, and these responses appear at Table 9 (plaintiff's exhibit 255, Table 9, p. 9 (stamp # 060)). Only one respondent out of 399 was confused because HEALTHY CHOICE and HEALTH SELECTIONS had "same name/similar name." Of the remaining 28 other respondents who were counted as "trademark confused," it is unclear whether they were confused for trademark reasons alone, partially for trademark reasons, or not at all for trademark reasons. Some of the respondents were counted as confused because of "packaging." Others of the respondents were confused because they responded "they look alike/they look the same/look similar." Others were counted as confused because they responded "colors look the same." Still others were confused because they responded "both say healthy/both say health." Still one other was confused because of "same style."

Inasmuch as this is not a trade dress lawsuit, packaging alone cannot constitute trademark infringement. ConAgra obviously has no proprietary right to the color green, for example. ConAgra has also admitted that it has no proprietary right to the word healthy or derivations of the word health. If the respondents were confused by such things as the packaging, or the color green, or the allegedly healthy attributes of the products, but not the mark, there is no likelihood of confusion for trademark related reasons.

This is not to say that trade dress is irrelevant. It is only to say that similarity in trade dress alone (particularly where some of the dress is in the public domain) will not establish a likelihood of confusion where it is claimed that there is trademark infringement. The relevant question is whether the offending mark, with or without trade dress, plays a part in the consumer's confusion.

Even if all 28 of those counted as "trademark confused" because of the packaging considerations were properly added to the one person who was confused because of the similarity in name, the resulting percentage of confusion is only about 7.3%. This level of "confusion" is not large enough to support a finding that consumers generally are or will be confused by the offending mark.

Professor McCarthy has observed that: "The lowest reported figure is 8.5%, which the court found to be `strong evidence' of a likelihood of confusion where other evidence was also strongly supportive." 2 J. McCarthy, Trademarks and Unfair Competition § 32:54, at 785 (citing Grotrian, Helfferich, Schulz, Th. Steinweg, Nachf v. Steinway Sons, 365 F. Supp. 707, 716 (S.D.N.Y. 1973), modified, 523 F.2d 1331 (2nd Cir. 1975)). In the United States Court of Appeals for the Eighth Circuit, the lowest reported percentage has been between 10 and 12. Mutual of Omaha Ins. Co. v. Novak, 836 F.2d 397, 400 (8th Cir. 1987), cert. denied, 488 U.S. 933, 109 S.Ct. 326, 102 L.Ed.2d 344 (1988). The United States Court of Appeals for the Seventh Circuit, reviewing cases involving low percentage results of surveys endeavoring to establish likelihood of confusion, found that a figure of 7.6% is "a factor weighing against infringement." Henri's Food Products Co. v. Kraft, Inc., 717 F.2d 352, 358-59 (7th Cir. 1983).

In addition to the 8.5% who were confused as to names, 7.7% of those surveyed perceived a business connection between the two companies. 365 F. Supp. at 716.

ConAgra relies upon Wendy's lnt'l, Inc., v. Big Bite, Inc., 576 F. Supp. 816, 823-24 (S.D.Ohio 1983). At the preliminary injunction stage of the proceedings, and concluding therefore that it would not analyze in detail the "de minimis" question, the Wendy's court found that defendant's survey which was slanted in favor of finding no likelihood of confusion, but which produced confusion at the rate of 7% for all age groups and 21% for the most relevant age group, should be credited for purposes of issuing a preliminary injunction. I do not think that the Wendy's case is persuasive precedent for ConAgra's position that levels of confusion in the 7% range are enough because: (a) the court did not consider the de minimis question; (b) the court's view was expressed at the preliminary injunction stage of the proceedings; and (c) the level of confusion in the most relevant consumer age group was 21%.

The significance of the "[w]hat makes you say that?" question is readily apparent. When Dr. Jacoby used the question, the level of confusion decreased dramatically, and, depending upon the interpretation of the responses, either reduced the level of confusion to virtually nothing (one respondent out of 399) or the question injected serious doubts about whether the respondents were confused in part or in whole for trademark (as opposed to trade dress only) relevant reasons.

It might reasonably be argued that there was no necessity of asking "[w]hat makes you say that?" after asking question 2. As ConAgra points out in its brief, plaintiff's post-trial brief at 32 n. 8, in regard to question 2, 100% of the respondents correctly purchased the Hunt's Tomato Paste, and 95% of the respondents correctly purchased the French's Mustard. (Plaintiff's exhibit 255, at Tables 3 4 (stamp # s 55 56)). This would, from ConAgra's point of view, suggest that there was no reason to assume that consumers misunderstood what was required of them during the purchase task, and therefore no reason to ask "[w]hat makes you say that?" or some similar question.

But I agree with Hormel that the state of mind question would have been helpful on question 2 because the shopping instructions first asked the respondents to "look at these three products," "keep these particular products in mind" and after the products had been displayed for a short time period, the consumer was told to "buy the specific products that I just showed you" but only "the particular brand I showed you" (plaintiff's exhibit 255, In Store Study Main Questionnaire at 1 (stamp # 33)). Both HEALTHY CHOICE and HEALTH SELECTIONS were brand new in the markets, unlike the Hunt's and French's products. Without a state of mind question after the purchasing task, it is difficult to know whether the respondents understood the directions.

Indeed, I believe there is fairly strong reason to believe the respondents may not have understood the purchase task. In group 4, which was Miami, Florida, consumers were shown a HEALTH SELECTIONS product and asked to buy that specific product (and the two others) that had been shown to consumers by an interviewer who had given the consumers 7 dollars to make the various purchases. In group 4 the consumers were shown a HEALTH SELECTIONS product, but only HEALTHY CHOICE was carried in the supermarket (plaintiff's exhibit 255 at 8 (stamp # 10)).

Despite the fact that the consumer was to purchase only "the specific products that I just showed you" (plaintiff's exhibit 255 at 1 (stamp # 033) (main questionnaire)), 91 percent of the respondents (134 of 147) brought back a Hormel product, which was not a HEALTH SELECTIONS product since it was not on the shelf (plaintiff's exhibit 255, at table 5, (stamp # 057)). Although Dr. Jacoby did not count these individuals as "confused" for trademark reasons, these respondents were "confused" as to their task. Thus, if the balance of the respondents who were counted as "confused" for trademark reasons also misunderstood their task to be to bring back a similar, but not identical, product, the results of question 2 are meaningless. A "state of mind" question would have cured this ambiguity in the data.

Even if I were to accept ConAgra's argument in regard to question 2, I do not believe this is necessarily helpful to ConAgra. It must be remembered that it was not possible to test for product confusion (question 2) for respondents in group 1, inasmuch as these people were shown HEALTHY CHOICE and then sent into a supermarket that carried only HEALTHY CHOICE. Thus, one is left with groups 2, 3, 4 and 5. In group 2, an astounding 84% of those responding were marked as confused and even Dr. Jacoby indicated that this number "appeared unusually high relative to the other groups." (Plaintiff's exhibit 255, at 21 (stamp # 23)). Thus, Dr. Jacoby, although not totally discounting group 2, indicated that if one were to place group 2 aside, "the evidence from each of the other groups reveals that some proportion of consumers (on the order of 7%) are likely to experience product confusion." (Plaintiff's exhibit 255 at 21-25 (stamp # s 23 24)). As suggested previously, a 7% level of confusion is normally not considered legally significant. Thus, even if I were to agree with ConAgra, which I do not, that the state of mind question need not have been asked with regard to question 2, I am still not persuaded that question 2 is suggestive of significant levels of confusion.

The second major flaw pointed out by Hormel had to do with Dr. Jacoby's failure to use a "control." As Professor McCarthy has noted, "there is often 'general background noise' in survey figures." 2 J. McCarthy, Trademarks and Unfair Competition § 32:54, at 784. Since "there will always be some interviewees who are bored, hurried or just plain contrary," their responses "must be filtered out through control questions." Id. In this case, a control question would have involved using a control product, that is, a product that is a non-infringing product which is similar to the products at issue. For example, if an admittedly non-infringing product produced the same or similar levels of confusion, then one would assume that the survey is measuring only "general background noise."

During the pretrial preparation stage, ConAgra took the deposition of Dr. Ivan Ross, and after that deposition, where Dr. Ross criticized the failure of Dr. Jacoby's study to use a control, Dr. Jacoby admitted to Dr. Kaplan, his colleague, that "it represented an omission, not having the control" (filing 98, TR 1020:7-8). Even Dr. Jacoby admitted that the purpose of a control is to get a sense of what the general error rate was and he did not have that for shelf-stable microwaveable products:

Q. And accordingly — what's the purpose of a controlled [ sic] product like Professor Ross envisions?

A. As I mentioned, to be able to get a handle on what's the general error rate. Now, I got that with respect to using the Hunt's and the French's, but I didn't get it with respect to another product in that specific category, namely the diet-conscious shelf-stable microwaveable, and it would have been nice to have done that, and that's one of the points where I agree with Dr. Ross, in hindsight, that would have been nice.

(filing 92, TR 130:8-19).

Thus, it seems to me that Dr. Jacoby's study is also flawed for failing to use a control product. It is quite likely that at least some of Dr. Jacoby's results reflect "background noise" which would have been "screened out" had he used a control product, meaning a non-infringing product that would have been competitive with HEALTHY CHOICE and HEALTH SELECTIONS. Since no control was used, the Jacoby study, standing alone, must be significantly discounted.

(3)

ConAgra, evidently sensitive to the criticisms of Dr. Ross about the failure to use a control in the Jacoby study, commissioned Dr. Kaplan to do a supplemental survey which sought to use Jacoby's survey methodology but to add a control product (plaintiff's exhibit 311). Dr. Kaplan's survey was conducted in Dallas, Texas, and Miami, Florida, and used approximately 100 respondents. The Jacoby survey methodology was employed with the exception that the Kaplan study respondents were shown only HEALTHY CHOICE and a control, LIGHT BALANCE, before entering the store. The stores did not carry HEALTH SELECTIONS or it was removed from the shelves prior to the control test. The respondents were then asked the same questions as the respondents in the Jacoby study.

Dr. Kaplan then calculated the "net confused" as was done in the first study. He obtained a weighted average of 28% total as being confused, down from 68% in Jacoby's study, and then subtracted his (Kaplan's) net confused number of 28% from Jacoby's number of 68% for a net confused of approximately 40% when a control was used. Introducing the control to question 2 of the Jacoby study did not change the results. Introducing the control to question 3a of the Jacoby study reduced the confusion level to 8.7% or a net difference of 14.6% between the Jacoby study and the Kaplan study. Introducing the control to questions 3c/d did not affect the Jacoby study and confusion remained at 7.3%. Introducing the control to question 4, the permission question, totally eliminated permission (sponsorship) confusion.

It should be noted that Dr. Kaplan separated questions 3c/d from question 4 for analytical purposes. Dr. Jacoby had simply lumped them together. Hence, Kaplan had four types of confusion whereas Jacoby had only three, but Jacoby's third factor was in reality a composite of association and permission (sponsorship) types of confusion.

The results of Dr. Kaplan's study are produced below:(City) Base Product Company Associated Permission Q.s2,3a, 3c/d,4

Type of Confusion Group Q.2 — Wrong Q.3a — Same Q.3c/d Q.4 Net % % % % % 1 (MIA, DVR) 48 NA 25 15 38 67 2 (DLS) 149 84 15 1 12 87 3 (DLS) 28 11 36 11 4 50 4 (MIA) 147 6 28 10 24 57 5 (DLS) 27 7 30 11 11 52 WEIGHTED ----- 39.6 23.3 7.3 19.0 68.7 AVERAGE: Control 4 (DLS) 42 0 2 0 12 14 5 (MIA) 50 0 14 0 26 40 WEIGHTED ----- 0 8.7 0 19.6 28.3 AVERAGE: NET 39.6 14.6 7.3 ----- 40.4 DIFF. (Plaintiff's exhibit 311, at 6).

(4)

Hormel responds to the Kaplan study by making three primary points. First, Hormel notes that Dr. Kaplan did not attempt to deal with the "state of mind" question, which Hormel believes is the most serious flaw in the Jacoby study. Second, Hormel notes that the use of the control by Dr. Kaplan totally eliminated any confusion dealing with permission (sponsorship) confusion. Finally, Hormel argues that the control that Dr. Kaplan used — LIGHT BALANCE — was not a good control because it did not control either for the words HEALTHY — HEALTH or for the color green. As for the first two points made by Hormel, they are indeed true, and at this juncture they need no further analysis. I thus turn to the question of whether LIGHT BALANCE was an adequate control.

In his supplemental survey, Dr. Kaplan set up two "cells" or control groups utilizing LIGHT BALANCE, a microwaveable shelf-stable product in a bright blue cup as a "control" product. Fifty respondents in each control group were asked the same questions that were asked in Jacoby's original survey. Dr. Ross, Hormel's expert who was retained to critique the Jacoby study, had testified that a control product should be a non-infringing product as similar as possible to the products at issue. Dr. Ross would have used the HEALTHY RECIPES product as a control because:

It appears to have more of the elements in common with the alleged to be infringing product of Hormel than does the LIGHT BALANCE product for a number of reasons, but certainly color and name. And, again, name is the primary issue, as I understand our issues, which make it a much better control.

(filing 99, TR 1147:2-9).

It is true that the LIGHT BALANCE product is fairly different from either the HEALTHY CHOICE product or the HEALTH SELECTIONS product in that it is essentially all bright blue and its name, in the sense of sight, sound and meaning, is quite different (compare the LIGHT BALANCE product (plaintiff's exhibits 278 and 279) with the HEALTH SELECTIONS product (plaintiff's exhibit 267 and defendant's exhibit 576), and the HEALTHY CHOICE product (defendant's exhibit 536)). The HEALTHY RECIPES product (defendant's exhibit 552) is much closer in both name and coloration (and also because it has a cardboard package similar to the HEALTHY CHOICE sleeve) to both the HEALTH SELECTIONS and HEALTHY CHOICE products.

Dr. Kaplan admitted that had he known of the HEALTHY RECIPES product he would have used it as a control:

Q. Let me show you what's been marked as Defendant's Exhibit 552. It's a product called Healthy Recipes. Are you familiar with this product at all, Dr. Kaplan?

A. No. I'm not. Not until you just showed it to me.

Q. Let's operate on the assumption, Dr. Kaplan, that ConAgra has looked in and looked hard at this product and decided that there's no trademark confusion problem, okay? That's the assumption we're operating under.

A. Okay.

Q. So that in that respect there's no problem in using Healthy Recipes as a control; right?

A. That's correct.

Q. Healthy Recipes, of course, has the word "healthy" in it?

A. It certainly does.

Q. Light Balance doesn't?

A. That's true.

Q. Healthy Recipes has green — the color green on the package?

A. Yes, it does.

Q. White lettering and so forth. It's fair to say, is it not, Dr. Kaplan, that the Healthy Recipes product is a lot more similar to either Health Selections or Healthy Choice than is Light Balance?

A. I think it's fair to say — yeah, I think that's a fair statement. At least just based on my observation.

Q. So if you had known about that product, in order to have a more effective control, you certainly would have utilized that as a control rather than Light Balance, is that fair to say?

A. Yes.

(filing 98, TR 1055:20-25; 1056:1-25; 1057:1-6).

ConAgra responds by suggesting that the HEALTHY RECIPES product was often found in different places in the supermarket and was not truly competitive. However, as noted previously, photographs show that the HEALTHY RECIPES product was available in Miami, Florida (defendant's exhibits 619 620). Mr. Bunge, a Hormel survey expert, testified that he personally observed the HEALTHY RECIPES products in numerous stores in Chicago, Illinois (filing 96, TR 748:22-25; 749:1-4). ConAgra had seriously examined the HEALTHY RECIPES product to determine whether it was infringing and had concluded that it was not infringing because of the use of the word "recipes" (Hughes deposition, TR 90:10-24). Under these circumstances, I believe the HEALTHY RECIPES product was sufficiently competitive that it could be used as an appropriate control.

I conclude that the HEALTHY RECIPES product would have been a far better control than the LIGHT BALANCE product for the reasons stated by Dr. Ross. This of course does not mean the Kaplan study is irrelevant, it only means that the control was not effective to screen for the color green and the word "healthy" or "health", understanding that ConAgra has no proprietary rights to the color green or the word "healthy" or its derivatives.

(5)

I am then left with the Jacoby study, supplemented by the Kaplan study. Ultimately, I conclude that the Jacoby study, as supplemented by the Kaplan study, is indicative of some likelihood of confusion, but the cause of the confusion cannot be ascertained with any particular degree of reliability. I arrive at this result for the following reasons.

As to group 2, and question 2, dealing with "product" confusion, of the Jacoby study, the 84% level of confusion is too high to be believed when compared with about an average 79% level of confusion for groups 3, 4 and 5 on the same question in the Jacoby study or control cells 4 and 5 in the Kaplan study. There simply is no explanation why levels of confusion would be so much greater in group 2 when compared with the other groups. If one disregards group 2, for which Kaplan did not attempt to introduce a control, for purposes of examining question 2, one is left with a level of confusion somewhere in the area of 7% as indicated by both Jacoby and Kaplan. As stated previously, a 7% level of confusion has normally been considered de minimis.

It should be remembered that no test was possible for group 1 because these people were shown HEALTHY CHOICE and then sent into a supermarket that carried only HEALTHY CHOICE. Moreover, Dr. Kaplan's control cells were only used for groups 4 and 5, where confusion in the Jacoby study was 6% and 7%, and thus no effort was made to introduce a control for groups 2 or 3, where confusion in the Jacoby study was 84% and 11% (filing 98, TR 1049:5-19).

By use of a control, the Kaplan study reduced the level of confusion with respect to question 3a, same company, from 23.3% to 14.6%. While this reduction may not appear to be great at first glance, one recognizes that by use of a relatively poor control confusion was reduced by about 38% from what it had been in the Jacoby study. This in turn suggests how sensitive these studies are to irrelevant factors.

Concerning sponsorship confusion dealing with association (questions 3 c and d), one sees a level of confusion of 7.3% in both the Jacoby study and Kaplan study, which, as indicated previously, is de minimis. Moreover, as noted previously, when the state of mind of the respondents was probed by asking "[w]hat makes you say that?", in only one response is there clear evidence that the offending mark either alone or in conjunction with the trade dress (as distinguished from the trade dress alone) was confusing.

Insofar as question 4 was concerned, dealing with sponsorship confusion in the Jacoby study, the Kaplan control totally eliminated any confusion. Once again this emphasizes the sensitivity of these types of surveys to irrelevant factors.

The overall impact of the Kaplan study, using a poor control, was to reduce confusion in the Jacoby study by about 41%. This seems to confirm, yet again, the sensitivity of these surveys to irrelevant factors, and, as a result, the need to be quite careful when determining what the surveys actually show.

Thus, after the Kaplan control, the levels of confusion stand at about 7% for question 2, 14.69% for question 3a, and 7.3% for question 3c/d. I am persuaded that: (a) had the state of mind question been asked in question 2 and question 3a the responses would have been similar to the responses received in regard to question 3 c/d, (b) the responses received in regard to question 3c/d do not clearly establish confusion for trademark relevant reasons, and (c) had the HEALTHY RECIPES control been used, the levels of confusion would have been substantially reduced.

(6)

Hormel, like ConAgra, also commissioned a consumer survey. This survey (defendant's exhibit 533) was prepared by Mr. John Bunge.

Bunge's survey was conducted in shopping malls in the cities of Baltimore, Kansas City, New Orleans and San Francisco. Respondents, after they were screened, were shown a display of products made up of 21 shelf-stable items, including two each of the products at issue, two HEALTHY RECIPES products, two LIGHT BALANCE products, two Hormel TOP SHELF cartons, two Kraft products, and various other products (defendant's exhibit 533 at 7). According to Bunge, he selected the 21 products after visiting ten grocery stores in the Chicago, Illinois, area, thereby determining, to Bunge's satisfaction, that the products in the display typically represented the products in the shelf-stable section of the stores he visited (filing 96, TR 747:9-25; 748:1-25; 749:1-18).

Each of the respondents in Hormel's survey was seated at the product display, with 21 products in view, and asked in question 6 to state which of the following "three statements you think is correct: all of these products are put out by the same company or organization; some of these products are put out by the same company or organization; all of these products are put out by different companies or organizations." Respondents were not given a specific option to say "don't know," although if a respondent stated that he or she did not know, the interviewer evidently recorded the answer.

If the respondent answered that all were put out by the same company, or that all were put out by different companies, or that he/she did not know the answer, the interview was completed, except for demographic information. If the respondent thought some were put out by the same company or organization, he/she was asked which ones and why he/she came to that conclusion. Reasons were recorded verbatim and probed for completeness and clarity.

Following these questions, whatever product the respondent was talking about was removed from the exhibit and the respondent was asked, "do you think any of the other products are put out by a different single company or organization than the one you were just talking about?" If the answer was "no" or "don't know," the interview was completed, except for demographic information. If the answer was "yes," the respondent was again asked which ones and why he/she came to that conclusion. Reasons were recorded verbatim and probed for completeness and clarity. This procedure was repeated until there was a single product or no product left in the exhibit or until the respondent did not think any of the remaining products were put out by a different company or organization or did not know. Lastly, demographic information was obtained.

The results of Mr. Bunge's study, from his perspective, are as follows:

1. Only one of 425 respondents exclusively grouped together the two Healthy Choice products and the two Health Selections products as being put out by the same company or organization. (Table 9, page 22 and Verbatim Comments, Resp. # 1002, page 35)

2. Only 22 of 425 respondents (5.2%) grouped together the two Healthy Choice products and the two Health Selections products, with each other or in combination with other products, in any fashion. (The two Health Selections with the other 6 Hormel products and Healthy Recipes, or Lunch Bucket, or Light Balance, as well as the Healthy Choice, etc.) (Table 9, pages 22-25, Verbatim Comments, pages 35-38)

3. A total of 56 respondents grouped together only the two Healthy Choice products with the two Healthy Recipes products. This compares to the single respondent who grouped together the Health Selections with the Healthy Choice products. (Table 9, page 22)

4. A total of 89 of the 425 respondents (20.9%) group together Healthy Choice products and Healthy Recipes products, with each other or in combination with other products, as being put out by the same company or organization. This compares with the 22 respondents who grouped together Healthy Choice with Health Selections in any fashion. (Table 9, pages 22-25, Verbatim Comments, pages 39-50)

5. Even among the subgroup of those 272 respondents who think "some" of the 21 products are put out by the same company or organization, the 22 grouping together the Healthy Choice products and the Health Selections products, with each other or in combination with other products, in any fashion is only 8.1% while the 89 grouping together the Healthy Choice and Healthy Recipes, with each other or in combination with other products, in any fashion is 32.7%. (Table 9, pages 22-25 and calculations)

6. By far, the vast majority of respondents who thought some of the products were put out by the same company or organization correctly grouped together, as single groupings, the various sources. Some 51.5%, of all respondents and 80.5%, of those who said "some," grouped together Hormel products only and with no others. Similarly, 47.5% of all and 73.5% of those saying some grouped together Chef Boyardee products only and with no others. For the two Kraft products, the figures were 45.4% and 71.0%. (Table 9, page 20-21)

7. Of the 22 respondents who grouped together Health Selections and Healthy Choice products, with each other or in combination with other products, in any fashion, 14 at the same time included Healthy Recipes in the grouping. Reading the reasons why, regarding these source groupings, for all 22 respondents, indicates a variety of reasons and strength for the recognition of the Hormel name and consistency of name on packages. While some mention is made of similarity of green color and name similarity, also strong is the fact that Hormel makes several products and the others are "light," or "diet" or "healthy." (Verbatim comments, pages 35-38.

8. Among the 22 respondents who group together Health Selections and Healthy Choice, with each other or in combination with other products, in any fashion, very few specifically give name similarity as a reason, apart from a concept of Healthiness (Verbatim Comments, pages 35-38), but among the 89 who grouped together Healthy Choice with Healthy Recipes, at least 32 stress the name similarity, the name or word "Healthy or healthy, etc. (Verbatim comments, pages 39-50)

9. Among all 425 respondents in this research, only three unique groupings other than true single source groupings were mentioned by more than as many as 10 respondents. These two groupings were the 56 respondents who put together the Healthy Choice and the Healthy Recipes products, the 19 respondents who put together the two Kraft products with the two Light Balance products (both predominantly blue) and the 11 respondents who put together Chef Boyardee and Lunch Bucket Products. (Tables 9 and 10, pages 20-31)

(Defendant's exhibit 533 at 12-13).

(7)

ConAgra, like Hormel, is able to vigorously attack consumer surveys. I shall note only the most serious of these attacks.

ConAgra also has additional objections to the Bunge survey, such as the following. First, ConAgra argues that the respondents were not confined to prospective purchasers. About 75% of the respondents, however, affirmed that they had purchased single serving, non-frozen, microwaveable meals or entrees in the past three months, and the remaining 25% were "very likely" or "somewhat likely" to purchase such products. Second, ConAgra argues that Bunge should have limited the products surveyed to health-oriented, microwaveable products, but the evidence is that most, if not all, of the products are competitive whether health-oriented or not. Third, ConAgra criticizes Bunge for not conducting the survey in a grocery story setting, rather than in a shopping mall. Yet, all of the survey experts agreed that consumer surveys are often conducted in shopping malls. Next, ConAgra argues that Bunge should have conducted his surveys in the test markets. However, Bunge's rationale for not doing so — test markets are likely to have extensive promotional activities which could adversely affect the results of a consumer test — seems convincing. In any event, except for the objections mentioned in the text, none of them seriously undermine the results of the Bunge survey.

ConAgra first points out that Mr. Bunge did not allow respondents an explicit opportunity to state that the respondent did not know the answer to the question. The impact of not permitting the respondent to say "don't know" is obviously to force the respondent to guess. Dr. Ross, the Hormel expert employed to criticize Dr. Jacoby's study, indicated that he always used a "don't know" option for studies done for legal purposes (filing 99, TR 1106:11-25; 1214:1-2). While Dr. Ross said that he believed as an empirical matter, based upon specific research he had done, that the failure to use a "don't know" option would result in a random and uniform distribution of forced answers, I believe ConAgra's point is well taken. While there may be a uniform distribution of forced answers given the absence of a "don't know" question if the sampling size is large enough, there is no empirical evidence in the record to suggest that the absence of a "don't know" response in this case would be irrelevant. To put the matter differently, while one might assume that statistically forced answers would be distributed uniformly if enough respondents were forced to give answers, we do not know what the effect on this particular study is of the failure to give a "don't know" option. Accordingly, the study must be viewed with caution. But, as Hormel points out, this is not a fatal flaw. SquirtCo, 628 F.2d at 1089-1091 n. 4.

Second, and most significant in my mind, ConAgra points out that respondents were permitted to examine the 21-product display for as long as they wished. In my mind, this viewing environment does not replicate the environment in which consumers normally function. They do not sit and stare at products, for purposes of answering questions, for substantial periods of time when they are shopping. According to Bunge, the respondents were given however long they wished to examine the products and he remembered that it was "reported to me that some interviews took 20, 25 minutes" with the average interview being probably 8 to 10 minutes (filing 97, TR 840:6-25; 841:1-5).

I agree with ConAgra that this procedure casts serious doubts about the validity of the Bunge study. Without some limitation as to time, the Bunge study in essence could turn into a "reading contest" where the consumer is given leave, and in fact is invited, to make a minute inspection of the various products. This is not what consumers actually do. As Professor McCarthy has indicated, "the closer the survey context comes to marketplace conditions, the greater the evidentiary weight it has." 2 J. McCarthy, Trademarks and Unfair Competition § 32:48, at 771. Recognizing the very difficult task that Mr. Bunge had before him, I believe the "reading contest" argument advanced by ConAgra is a compelling argument. Although I do not believe it to be a fatal flaw, the "reading contest" problem cautions a skeptical view of the Bunge survey results.

(8)

What remains to be determined, then, is the impact of the Jacoby study, supplemented by the Kaplan study, viewed in light of the Bunge study. Ultimately, I think all of the studies confirmed that there is some likelihood of confusion, but the confusion does not appear to be caused for trademark relevant reasons.

Insofar as I am concerned, Bunge's survey produced the following pertinent results:

(a) One out of 272 respondents who answered question 6 grouped two HEALTHY CHOICE products with the two HEALTH SELECTIONS products;

(b) Twenty-two respondents out of 272 grouped the two HEALTHY CHOICE products with the two HEALTH SELECTIONS products in combination with other products;

(d) Eighty-nine respondents grouped the two HEALTHY CHOICE products with the two HEALTHY RECIPES products in combination with other products;

(e) The recorded verbatim responses revealed that few, if any, respondents grouped the HEALTHY CHOICE products with the HEALTH SELECTIONS products because of similarity of the name;

(f) The dominant factor in grouping the HEALTHY CHOICE, HEALTH SELECTIONS and HEALTHY RECIPES products together in some combination appears to have been similarity in packaging or the "healthy" attributes of the product;

(g) The house logo HORMEL was often recognized by the survey respondents.

In fact, Bunge's study, in many important respects, is consistent with Jacoby's study, supplemented by Kaplan's study. For example, in the Jacoby study, for question 2, the purchasing task, the levels of confusion for groups 4 and 5 were 6% and 7%, respectively. Kaplan's study used a "control" only for groups 4 and 5. Thus, as to question 2, the levels of confusion, where a control was used, remained at 6 and 7% respectively. If one totally ignored verbatim responses obtained both by Jacoby, regarding questions 3c and d, and by Bunge, Bunge's 8.1% level of confusion (22 out of 272 respondents grouping HEALTHY CHOICE and HEALTH SELECTIONS in combination with other products) compares favorably to the 6% and 7% levels of confusion found in groups 4 and 5 in the Jacoby/Kaplan study.

Moreover, if one simply compared the levels of confusion found by the Jacoby study, as supplemented by the Kaplan study, regarding questions 3c and d where Drs. Jacoby and Kaplan found 7.3% confusion with the level of confusion found by Bunge at 8.1% (once again ignoring the verbatim responses), the levels of confusion appear to be similar. As noted previously, levels of confusion at 6, 7 or even 8% are "de minimis."

More importantly, when one takes the responses from the Jacoby study produced in response to questions 3c and d and compares the verbatim responses obtained from the Bunge study, there is a consistency in the responses by consumers. Seldom if ever in either the Jacoby study or the Bunge study do any of the consumers suggest that similarity in the marks alone or the similarity in the marks together with the similarity in trade dress are the factors confusing the consumers. Rather, in analyzing the verbatim responses it appears that consumers may be confused by such things as packaging alone, particularly the color green or the allegedly healthy attributes of the product, but not the similarity in marks, either alone or in conjunction with the trade dress. As noted previously, trade dress alone does not establish likelihood of confusion where it is claimed that there is trademark infringement. The question is whether the junior mark, with or without trade dress, plays a part in consumer confusion. By looking at the verbatim responses produced as a result of Dr. Jacoby's study and Mr. Bunge's study, in very few instances is there confusion as a result of the similarity of the marks, with or without trade dress as a consideration.

Accordingly, I am persuaded that the Bunge study, particularly giving appropriate consideration to the verbatim responses, is not inconsistent with the Jacoby study, supplemented by the Kaplan study. There is confusion, but the true level of confusion is fairly small, and whatever confusion exists does not exist for trademark relevant reasons.

ConAgra would add the 6 or 7% level of confusion found in the Jacoby study regarding question 2, where a control was used, to the 8.1% level of confusion found by Bunge to arrive at confusion of between 14% and 15%. Adding the results of the two studies together to arrive at aggregate confusion is incorrect. Both the Jacoby study and the Bunge study tried to measure the same thing — confusion — but in different ways. The impropriety of adding these results can be illustrated by an example. If the ConAgra method were adopted, a plaintiff could conduct 100 surveys and if each study showed a level of confusion of 1%, a trier of fact would be forced to conclude that 100% of consumers would be confused. The better view would be to average the results. If one averaged the level of confusion found by Jacoby in question 2, regarding groups 4 and 5 (where Kaplan used a control) with the level of confusion found by Bunge, and ignored the verbatim responses, the average level of "confusion" is about 7.4%.

F. DEGREE OF CARE

There is no question but that the goods at issue are relatively inexpensive. This would tend to suggest consumers would be less careful, and therefore more likely to be confused, by any similarity in the marks. However, as ConAgra's documents suggest, consumers of shelf-stable products "are very price-value conscious" (defendant's (sealed) exhibit 569 at p. 3). This would suggest that although the goods are inexpensive, some degree of care might be expected of consumers. Finally, in the shelf-stable area of the supermarket "nutrition is not of major concern" (defendant's (sealed) exhibit 569 at 6). This would suggest that the use of the laudatory term "healthy" or "health" will not be a factor bearing on the degree of care exercised by many consumers of shelf-stable products. While the price of the goods involved suggests that consumers will not exercise a high degree of care, they will also not be whimsical in their purchases either.

G. CONCLUSION

The resolution of the infringement issue does not hinge upon a single factor, but involves careful consideration and weighing of all of the relevant factors. SquirtCo, 628 F.2d at 1091. In arriving at my decision, I have carefully considered and weighed each of the relevant factors.

On the one hand, Hormel's mark literally means the same as ConAgra's mark, the products are closely related from a competitive standpoint, and the products are relatively inexpensive. These factors weigh in favor of ConAgra.

On the other hand, ConAgra's mark is neither strong nor distinctive on the spectrum of marks or in the relevant marketplace, the marks, including their trade dress, appear to be different, the marks sound different, there was no intent on the part of Hormel to pass off its goods as the product of ConAgra, and there was no persuasive evidence of significant actual confusion. These factors weigh in favor of a finding for Hormel.

Notwithstanding the fact that I have carefully considered and weighed each and every factor, I wish to briefly make two points which especially convince me that judgment should be in favor of Hormel.

First, ConAgra's mark is a "laudatory" mark. In one sense, ConAgra has endeavored to appropriate unto itself the exclusive right to call its products "healthy." Consumers are not likely to be confused when confronted by competing "puffing" marks. Moreover, as a matter of policy, ConAgra should be give only narrow protection for its mark since the enforcement of ConAgra's mark would limit the language of commerce by preventing competitors from fairly describing their own goods.

By this statement I certainly do not mean to impute a bad or illegal motive to ConAgra.

Second, the scientific evidence presented to assist me, and which helps a judge to avoid improper subjective assessments, does not support ConAgra's assertions. The consumer surveys establish that: (a) there was very little confusion for trademark relevant reasons among the respondents, and (b) what confusion for trademark relevant reasons existed was de minimis.

Accordingly, I find as a matter of fact and conclude as a matter of law that Hormel's mark HEALTH SELECTIONS does not so resemble ConAgra's mark HEALTHY CHOICE that use of Hormel's mark will be likely to cause confusion or mistake, or to deceive in the shelf-stable market. Therefore, judgment on the infringement action must be for Hormel and against ConAgra.

III.

Hormel makes two arguments in its Section 43(a) Lanham Act false advertising counterclaims. First, Hormel argues that in its print advertisements ConAgra is falsely representing that its HEALTHY CHOICE shelf-stable products meet United States Department of Agriculture policy guidelines for low sodium and low calories. Second, Hormel contends that in its label and print advertisements ConAgra is falsely representing the amounts of cholesterol and sodium contained in its HEALTHY CHOICE canned entree products.

Hormel argues that ConAgra's claims are explicitly false and, therefore, Hormel need not prove consumer deception since consumer deception is proved by explicit falsity without evidence of actual consumer reaction. See, e.g., Sandoz Pharmaceutical Corp. v. Richardson-Vicks, 902 F.2d 222 (3rd Cir. 1990). Therefore, Hormel did not produce any evidence of consumer confusion. Presuming for the moment that Hormel's legal argument is correct, I conclude as a matter of fact that ConAgra's claims in its advertisements and on its labels are not explicitly false.

A. USDA CLAIM

Hormel argues that print advertisements have appeared in newspapers since on or about June 30, 1991, which state in pertinent part that ConAgra's products are "U.S. inspected and passed by Department of Agriculture [establishment number]" (Defendant's Exhibit 512). The same print advertisements represent that HEALTHY CHOICE products are "low in calories, fat, cholesterol and sodium." Such an advertisement was published. (defendant's exhibit 512). The essence of Hormel's claim is that ConAgra's products do not comply with the United States Department of Agriculture's Sodium Labeling Guidelines, Policy Memo 49C (June 14, 1984), and Lite and Similar Terms, Policy Memo 071A (March 31, 1985).

As ConAgra points, however, its advertisements do not claim that ConAgra's products comply with the policy guidelines. Compared to competitive products such as Hormel's, ConAgra's products are in fact low in fat, cholesterol, and sodium, (filing 59, deposition of Michael Trautschold, TR 81:19-25; 82:1-23). For example, if one compares the HEALTHY CHOICE turkey chili with bean product (plaintiff's exhibit 269) with the Hormel chili product (plaintiff's exhibit 271) which Hormel claims contains "Less Salt," the Hormel product has 845 milligrams of sodium per 7.5 ounce serving, whereas the HEALTHY CHOICE product contains 560 milligrams of sodium per 7.5 ounce serving.

The fact that the ConAgra product may not comply with specific Department of Agriculture labeling guidelines does not make ConAgra's claims literally false or misleading. ConAgra makes no explicit claims that its products comply with USDA labeling guidelines. Hormel does not contend that as a general matter the statement "inspected and passed by Department of Agriculture" is untrue. Accordingly, judgment must be for ConAgra. Sandoz Pharmaceutical Corp. v. Richardson-Vicks, 902 F.2d at 921 ("We hold that it is not sufficient for a Lanham Act plaintiff to show only that the defendant's advertising claims of its own drug's effectiveness are inadequately substantiated under FDA guidelines; the plaintiff must also show that the claims are literally false or misleading to the public.").

Therefore, I need not reach the question of whether or not the USDA guidelines are applicable in this case. Among other things, ConAgra argues that USDA guidelines cannot regulate advertising, as opposed to the regulation of labels, since the guidelines do not explicitly pertain to advertisements, as opposed to labels, and the authorizing statutes, 21 U.S.C. § 601(o) and 607, apply only to labels and not advertisements generally.

B. CANNED PRODUCTS CLAIM

The second argument advanced by Hormel is that ConAgra is falsely representing on HEALTHY CHOICE canned entrees the amount of cholesterol and sodium in the products. Hormel's argument is based on the statement on a HEALTHY CHOICE beef stew label that the product contains 35 milligrams of cholesterol and 540 milligrams of sodium. Hormel contends, and it is true, that the levels of cholesterol and sodium displayed on the HEALTHY CHOICE can label are for a 7.5 ounce serving whereas the can actually contains 15 ounces of the product. Defendant's exhibit 513 is an example of the label at issue.

ConAgra's response asserts the following, all of which are true. First, the label on its 15-ounce cans contains the phrase "Serving Suggestion." Second, the back of the can label clearly indicates that the beef stew contains 35 milligrams of cholesterol and 540 milligrams of sodium per 7.5 ounce serving. Finally, ConAgra points out that the United States Department of Agriculture examined and approved the label on its 15-ounce can of beef stew. (filing 86, Hordvik Deposition, TR 29:19-23).

As a "serving suggestion," the claims on the ConAgra label appear literally true. Stated differently, there is no evidence that the claims are literally false. Therefore, judgment must be for ConAgra.

C. CONCLUSION

Hormel's Section 43(a), 15 U.S.C. § 1125(a), arguments under the Lanham Act are predicated upon this court finding that ConAgra claims "are explicitly false." (Hormel's post-trial findings of fact and conclusions of law at p. 89, ¶ 52). It is upon this basis that Hormel contends that there is no need for "extrinsic evidence regarding how the claims are interpreted by consumers." Id. Having found that both of ConAgra's assertions are not explicitly false, I must therefore enter judgment for ConAgra since there is no evidence in the record that consumers would likely be confused by the ConAgra claims.

IV.

In accordance with this memorandum opinion, including findings of fact and conclusions of law, judgment shall be entered dismissing ConAgra's complaint against Hormel and dismissing Hormel's counterclaims against ConAgra.

Beyond the remedy of dismissal of ConAgra's complaint, I find and conclude that Hormel is not entitled to further relief, such as a declaratory judgment or injunctive relief. Furthermore, to the extent that I have discretion to award attorneys fees to either Hormel or ConAgra, I decline to award attorneys' fees.

Conagra, Inc. v. Geo. A. Hormel Co.

Summary

Opinion

MEMORANDUM OPINION, INCLUDING FINDINGS OF FACT AND CONCLUSIONS OF LAW

I. A.

B.

JURISDICTION AND VENUE

PARTIES

HEALTHY CHOICE

HEALTH SELECTIONS

II.

A. THE STRENGTH OF THE MARK

(1)

(2)

B. SIMILARITY OF THE MARKS

(1)

(2)

(3)

C. COMPETITIVE PROXIMITY

D. INTENT TO CONFUSE

(1)

(2)

E. ACTUAL CONFUSION

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

F. DEGREE OF CARE

G. CONCLUSION

III.

A. USDA CLAIM

B. CANNED PRODUCTS CLAIM

C. CONCLUSION

IV.

Conagra, Inc. v. Geo. A. Hormel Co.

Conagra, Inc. v. Geo. A. Hormel Co.

Case Details

CitationsCopy Citation

Citing Cases

Citations