Hershey Foods Corp. v. Mars, Inc.

998 F. Supp. 500 (1998)

HERSHEY FOODS CORPORATION, and Homestead, Inc., Plaintiffs,
v.
MARS, INCORPORATED, Defendant.

No. Civ.A. 1:CV-97-1719.

United States District Court, M.D. Pennsylvania.

March 31, 1998.

Alan R. Boynton, Jr., David E. Lehman, McNees, Wallace & Nurick, Harrisburg, PA, Robert D. Rosenbaum, Roberta L. Horton, James T. Walsh, Elissa J. Preheim, Arnold and Porter, Washington, DC, for Hershey Foods Corp., Homestead, Inc.

James W. Evans, Christopher Charles Conner, Mette Evans & Woodside, Harrisburg, PA, Pasquale A. Razzano, Leslie K. Mitchell, Dorothy C. Alevizatos, Fitzpatrick, Cella, Harper & Scinto, New York City, for Mars, Inc.

MEMORANDUM

CALDWELL, District Judge.

Introduction

The plaintiffs are Hershey Foods Corporation and Homestead, Inc. Hershey makes candy and sells it nationwide. Homestead is a related corporate entity that owns many of Hershey's trademarks and trade dress. (We will refer to Hershey and Homestead collectively as Hershey or the plaintiff.) This lawsuit was filed against Mars, Inc., another major candy maker, alleging violations of federal and state law governing competition. The suit stems from Mars' adoption of a trade dress for its peanut butter M & M's that Hershey asserts unlawfully resembles certain elements of its trade dress for its Reese's Peanut Butter Cups.

We are considering Hershey's motion for a preliminary injunction, seeking an order prohibiting Mars from using its current trade dress (or the "trade dress at issue" because Mars freely adopts variations on this dress) on the basis that it dilutes by blurring the Reese's trade dress. Hershey invokes both federal and Pennsylvania law. A hearing was held on the motion on December 23, 1997, January 6, 1998, and January 14, 1998.

The Court has subject matter jurisdiction over plaintiff's federal claim pursuant to 15 U.S.C. § 1121 and 28 U.S.C. §§ 1331 and 1338(a), and supplemental jurisdiction over the state-law claim pursuant to 28 U.S.C. § 1367.

The Law of Dilution

A dilution claim protects a trademark holder's right in the mark's ability to act as a unique identifier of the mark's holder. 3 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 24.70, at 24-118 (4th ed.1997). This cause of action differs from 15 U.S.C. § 1125(a), section 43(a) of the Lanham Act, the federal trademark infringement law, by eliminating the need to show confusion; that is, the need to show that the defendant's mark causes consumers to confuse the defendant's goods with the plaintiff's. See, e.g., Versa Products Co. v. Bifold Company (Manufacturing) Ltd., 50 F.3d 189, 199-200 (3d Cir.1995) (a significant aspect of a section 1125(a) claim is the need to show confusion). Instead, a dilution claim asserts that, while the defendant's mark does not confuse consumers as to the source of the goods, it leads consumers to believe that the plaintiff's mark now identifies two different sources of goods, thus diluting the mark's ability to identify the plaintiff's goods alone. An example of a diluting trademark would be the hypothetical "Dupont shoes." McCarthy, supra, § 24.68 at 24-116.

In other words:

For dilution to occur, the relevant public must make some connection between the mark and both parties. But that connection is not the kind of mental link between the parties that triggers the classic likelihood of confusion test. Rather, the assumption is that the relevant public sees the junior user's mark, and intuitively knows, because of the context of the junior user's use, that there is no connection between the owners of the respective marks. However, even with those who perceive distinct sources and affiliation, the ability of the senior user's mark to serve as a unique identifier of the plaintiff's goods or services is weakened because the relevant public now also associates that designation with a new and different source.

McCarthy, supra, § 24.70, at 24-117 to 118.

Hershey's dilution-by-blurring claim is based on 15 U.S.C. § 1125(c) (1997), a 1996 addition to the Lanham Act, and 54 Pa.C.S.A. § 1124. The two laws are substantively similar. The federal law provides, in pertinent part, as follows:

Remedies for dilution of famous marks

(1) The owner of a famous mark shall be entitled, subject to the principles of equity and upon such terms as the court deems reasonable, to an injunction against another person's commercial use in commerce of a mark or trade name, if such use begins after the mark has become famous and causes dilution of the distinctive quality of the mark....

15 U.S.C. § 1125(c).

A "mark" is defined as "any trademark, service mark, collective mark, or certification mark." 15 U.S.C. § 1127. A "trademark" includes:

any word, name, symbol, or device or any combination thereof—

(1) used by a person,

....

to identify and distinguish his or her goods ... from those manufactured or sold by others and to indicate the source of the goods, even if that source is unknown.

Dilution is defined as:

the lessening of the capacity of a famous mark to identify and distinguish goods or services, regardless of the presence or absence of—

(1) competition between the owner of the famous mark and other parties, or

(2) likelihood of confusion, mistake, or deception.

15 U.S.C. § 1127.

The Pennsylvania law provides as follows:

Likelihood of injury to business reputation or of dilution of the distinctive quality of a mark registered under this chapter, or a mark valid at common law, or a trade name valid at common law, shall be a ground for injunctive relief notwithstanding the absence of competition between the parties or the absence of confusion as to the source of goods or services.

54 Pa.C.S.A. § 1124. It does not define dilution.

The federal statute requires that the mark be "famous" and provides that the court "may consider factors such as, but not limited to" the following in "determining whether a mark is distinctive and famous":

(A) the degree of inherent or acquired distinctiveness of the mark;

(B) the duration and extent of use of the mark in connection with the goods or services with which the mark is used;

(C) the duration and extent of advertising and publicity of the mark;

(D) the geographical extent of the trading area in which the mark is used;

(E) the channels of trade for the goods or services with which the mark is used;

(F) the degree of recognition of the mark in the trading areas and channels of trade used by the marks' owner and the person against whom the injunction is sought;

(G) the nature and extent of use of the same or similar marks by third parties; and

(H) whether the mark was registered under the Act of March 3, 1881, or the Act of February 20, 1905, or on the principal register.

15 U.S.C. § 1125(c)(1).

We read this language as giving the court flexibility in analyzing the enumerated factors so that their relative weight in any given case can be balanced. See Star Markets, Ltd. v. Texaco, Inc., 950 F. Supp. 1030 (D.Haw.1996) (balancing the factors and deciding whether any particular factor either heavily, slightly or merely favored one party).

In deciding whether the defendant's mark dilutes the plaintiff's by blurring, federal courts interpreting section 1125(c) have started from the framework first set forth in Judge Sweet's concurring opinion in Mead Data Central, Inc. v. Toyota Motor Sales, U.S.A., Inc., 875 F.2d 1026 (2d Cir.1989), which Judge Sweet used for interpreting the New York antidilution statute. See Clinique Labs., Inc. v. Dep Corp., 945 F. Supp. 547 (S.D.N.Y.1996); Ringling Brothers-Barnum & Bailey Combined Shows, Inc. v. B.E. Windows Corp., 937 F. Supp. 204 (S.D.N.Y.1996); Star Markets, Ltd., supra; Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development, 955 F. Supp. 605, 615-16 (E.D.Va.1997); WAWA Dairy Farms v. Haaf, 1996 WL 460083 (E.D.Pa.1996).

Judge Sweet used a six-factor test:

1) similarity of the marks

2) similarity of the products covered by the marks

3) sophistication of consumers

4) predatory intent

5) renown of the senior mark

6) renown of the junior mark

875 F.2d at 1035.

He believed there was a positive correlation between blurring and the similarity of the marks, the similarity of the products covered by the marks, the existence of predatory intent, and renown of the senior and junior marks, but that there was a negative correlation if the consumers were sophisticated.

Some commentators have disagreed on whether all of the factors are relevant to a dilution analysis. See McCarthy, supra, § 24:94.1 at 24-163 to 164 (factors 2, 3, 4, and 6 do not appear to be relevant to a blurring analysis); Eric A. Prager, "The Federal Trademark Dilution Act of 1995: Substantial Likelihood of Confusion," 7 Fordham Intell. Property, Media and Entertainment L.J. 121 at n. 51 (1996) (factors 2, 3, 5, and 6 are not relevant). Additionally, in Clinique Labs., supra, the court refused to use predatory intent because Congress had not made it part of the statute. See 945 F.Supp. at 562 n. 22.

When these legal principles are put together, to prove dilution by blurring, the plaintiff must make the following showing: (1) that its mark is famous (as opposed to merely being a valid trademark); (2) that the defendant began using a mark in commerce after the plaintiff's mark became famous; and (3) that the "defendant's use causes dilution by lessening the capacity of the plaintiff's mark to identify and distinguish goods or services." McCarthy, supra, § 24:89, at 24-138. See also Utah Division of Travel Development supra, 955 F.Supp. at 614. In turn, dilution depends on at least how some of the factors set forth in Mead Data Central apply here.^[1]

Findings of Fact

The Parties

1. Hershey Foods Corporation is a corporation organized under the laws of Delaware with its principal place of business in Hershey, Pennsylvania.

2. Homestead, Inc. is a corporation organized under the laws of Delaware with its principal place of business in Newark, Delaware.

3. Homestead manages and owns many of the trademarks and trade dresses that Hershey uses in connection with its confectionery products.

4. Mars, Incorporated is a corporation organized under the laws of Delaware with its principal place of business in McLean, Virginia.

5. The M & M/Mars Division of Mars, Incorporated is an operating division of Mars.

6. Hershey manufactures confectionery products, including chocolate and peanut butter candy. Hershey markets peanut butter cups under the trade name "Reese's" and sugar-shelled peanut butter candy under the trade name "Reese's Pieces."

7. Mars also manufactures and sells candy throughout the United States and in foreign countries. Mars sells varieties of its sugar-shelled chocolate candy under its trademark M & M's, two lower-case "m"s connected by an ampersand.

8. Hershey and Mars have long been the top two candy manufacturers in the United States. Hershey ranks number one and Mars number two in U.S. market share among marketers of confectionery products. Hershey gained the number one rank from Mars in 1988.

9. Reese's peanut butter cup candy and Reese's Pieces candy and the Mars peanut butter M & M's candy compete directly with each other.

The Parties' Products and Trade Dress

A. The Reese's Peanut Butter Cup and Reese's Pieces Products and Trade Dress

10. Hershey's Reese's peanut butter cup candy has been Hershey's best-selling brand for decades. This is a chocolate peanut butter candy in a shallow cup shape with milk chocolate on the outside and peanut butter on the inside.

11. Hershey also markets Reese's Pieces candies, which is peanut butter encased in a candy shell.

12. Reese's peanut butter cup is sold in a package with a particular shade of background orange, the Reese's logo (its trade name in a slanted handwritten style in bright yellow with brown "keylining;" i.e., the yellow letters outlined in brown) and a brown rectangular bar with a sawtoothed top (imitating in two dimensions the top edge of a peanut butter cup) directly underneath the logo. The bar contains the message in bright yellow letters "PEANUT BUTTER CUPS." This is the "complete Reese's trade dress." See plaintiff's exhibit 1.

13. In this case, Hershey claims rights in the portion of that dress which consists of a particular shade of background orange as the principal element in combination with the brown and yellow elements, in a particular arrangement and juxtaposition used to create a distinctive overall impression. This is the "abbreviated Reese's trade dress."

14. The Reese's Pieces trade dress is similar in that it consists of the same shade of background orange, the Reese's logo (its trade name in a slanted handwritten style in bright yellow with brown keylining). However, it differs by having a so-called lozenge (a cartouche-shaped element) beneath the logo. The lozenge is brown with white keylining and has the words "peanut butter" in white letters. See plaintiff's exhibit 2.

B. The Mars' M & M's Products and Trade Dress

15. Mars has been selling its M & M's plain chocolate candies, bite-size individual chocolate candies with a sugar shell, in a package having a solid brown background with its distinctive M & M's logo in white lettering since 1949.

16. Mars began selling its M & M's peanut chocolate candies, a product with a peanut at the center, in the same style package, using a yellow background with the M & M's trademark (or logo) in its distinctive brown lettering style in 1964.

17. Mars began selling its M & M's almond chocolate candies, a product with an almond at the center, in the same style package using a tan background with the brown M & M's logo in 1987.

18. Mars began selling its M & M's chocolate candies in a mint chocolate variety in the early 1990's. The packaging for the M & M's mint chocolate candies uses a green background color with the M & M's logo.

19. Mars owns Federal trademark registrations for its M & M's variety packages Nos. 1,266,509, 1,272,035, and 2,005,935.

20. All of Mars' M & M's varieties packaging have included a display of cascading individual M & M's candy pieces (the "lentils") since at least as early as 1994. These candy pieces are "cultural icons" and highly recognizable by the public. They serve to identify the M & M's brand candies to consumers.

21. The design elements of the M & M's packages for these varieties, including the distinctive lettering style, the presence of the individual candy pieces or "lentils," and the lozenge indicating the variety situated below the M & M's trademark, are consistent throughout the range of these M & M's branded products.

22. The different varieties of M & M's brand of chocolate candies are together the largest selling brand of candy in the United States. In the last four years, Mars has sold approximately 4 billion dollars worth of M & M's brand candies in packaging bearing a display of the "cultural icon" images of the individual M & M's candies bearing the letter "m".

The M & M's Peanut Butter Variety

A. Background and Concept for the Peanut Butter M & M's Product

1. Peanut Butter M & M's: 1988-1993

23. In November 1988, senior management of Mars approved a proposal to commence test marketing of a peanut butter M & M's variety. The proposal stated: "The product [the proposed new M & M's variety] marries the benefits of two best selling brands—"M & M's" Chocolate Candies and Reese's Peanut Butter Cups—to grow M & M/Mars share of the chocolate candy market."

24. A test package for the Peanut Butter M & M's product employed a red background color and a yellow M & M logo, keylined in brown. The lozenge underneath the logo had a yellow background containing brown letters stating "Peanut Butter."

25. In 1990, Mars began to introduce Peanut Butter M & M's in phases. The initial package for the Peanut Butter M & M's product employed a red background color and a yellow M & M brand name, keylined in brown. Contrary to the test package, the lozenge had a brown background containing yellow letters stating "Peanut Butter." Plaintiffs have referred to this package as "Trade Dress 1."

26. After some initial success, by 1993 the product was "a disappointment." Sales of Peanut Butter M & M's fell from a peak of $78 million in 1991 to $40 million in 1992 to $34 million in 1993.

2. 1993-1994 "Relaunch"

27. Notwithstanding disappointing sales, Mars had confidence that peanut butter M & M's could succeed and, in 1993, Mars developed a plan to "relaunch" Peanut Butter M & M's.

28. The relaunch plan was to reformulate the product and market it in a revised trade dress that would begin entering retail markets in 1994. The background color of the trade dress was changed from red to "a more orange red." The revised package also included a yellow "violator" (words in yellow on the left side of the package) that stated: NEW RECIPE/GREAT NEW TASTE! MORE PIECES! Plaintiffs have referred to this package as "Trade Dress 2."

B. Hershey's Objections to Trade Dress 2

29. Hershey became aware of Trade Dress 2 in the spring of 1995.

30. Hershey objected to Trade Dress 2 because it employed a color combination "dangerously close" to that used on the REESE'S Trade Dress.

31. In an April 27, 1995 letter, Hershey's counsel, Mr. Snyder, wrote to Mars' in-house counsel, expressing Hershey's concerns. Mr. Snyder specifically asked that Mars discontinue the use of the yellow violator.

32. In correspondence in June of 1995, Mars indicated its plans to remove the yellow violator, but reserved the right to use yellow violators in the future. Hershey similarly reserved the right to challenge such use by Mars in the future.

33. Subsequently, Mars removed the yellow violator. Plaintiffs have referred to the altered package, after removal of the violator, as "Trade Dress 3."

C. Mars' PTO Application and the Parties' Negotiations Regarding Same

34. On October 3, 1995, the Patent and Trademark Office ("PTO") published a notice that Mars had filed an application to register its Trade Dress 3, but describing the colors broadly as "orange" (for the background color) and brown (for the M & M mark).

35. Hershey was concerned that, because Mars did not limit the shade of colors referred to in its PTO application, the application would, if granted, allow Mars to use in the future a shade of orange similar to that used on the Reese's trade dress. Hershey communicated these concerns to Mars and demanded that Mars withdraw its PTO application or revise it to limit the shade of orange claimed by Mars. In response, counsel to Mars asserted that the shade of orange used in Mars' Trade Dress 3 was "distinctly different" from that used in the Reese's trade dress. The parties subsequently commenced negotiations of a possible settlement agreement relating to the shades of orange which each would be able to use to "avoid unnecessary litigation,"

36. Pending these negotiations, Mars agreed to extend the time within which Hershey could file objections to Mars' PTO Application.

37. The negotiations continued until April 1997, but were unsuccessful. Hershey promptly thereafter filed an objection at the PTO to Mars' PTO Application.

38. Hershey initiated a resumption of the negotiations in early June 1997.

39. During the parties' negotiations, Mars was on full notice that Hershey objected to Mars' claim that Mars could adopt a color scheme similar to that used on the Reese's trade dress.

D. 1996: Mars' New Marketing Strategies

40. After the relaunch of Peanut Butter M & M's in 1994, the product's performance continued to be disappointing. Sales for 1994 were down and sales for 1995 remained flat despite the "relaunch."

41. Peanut Butter M & M's continued to be a disappointing product for Mars and its survival was in doubt. Mars was in need of a strategy to stabilize sales of Peanut Butter M & M's to prevent further declines which might terminate its viability as a product.

42. In August 1996, a new brand manager was appointed for the Peanut Butter M & M's variety. She was assigned the task of advertising, promotion and packaging for the variety.

43. The brand manager targeted Reese's Pieces as one area where M & M's peanut butter candies could grow.

E. New Package Design Project

44. In October 1996, the brand manager prepared a "Packaging Design Brief," a formal document that identifies the objectives of the package redesign project. This document was signed by the brand manager and her superior.

45. Under a section entitled "Legal Constraints," it was written that "Background color of the `M & M's' Peanut Butter pack cannot get any closer to the shade of orange on Reese's Pieces packs."

46. Mars sent the Packaging Design Brief to its packaging design firm.

47. The brand manager was familiar with the Reese's package and was aware at the time she wrote the Packaging Design brief that something had gone on between Hershey and Mars regarding the color orange.

48. The legal constraint on the Package Design Brief was subsequently removed. Under the section entitled "Review Packaging Brief," the brand manager wrote "Note amendments—removal of legal constraint on color of Peanut Butter package." Id. The packaging design firm was made aware of the removal.

49. Mars adopted the shade of background orange for its current trade dress for M & M's Peanut Butter candy to copy the orange background of the Reese's trade dress.

50. Mars began making commercial use in commerce of the new Peanut Butter M & M's package, the trade dress at issue in this case, in late June 1997.

51. The new package was distributed on a rolling basis, gradually replacing the prior package.

52. Plaintiffs learned of the trade dress at issue around October 25, 1997, and filed this suit on November 10, 1997.

The Current M & M's Peanut Butter Trade Dress and Its Contrast With the Abbreviated Reese's Trade Dress

53. The current M & M's peanut butter trade dress (although Mars may vary it from time to time) uses the same orange background as used by Hershey in its Reese's trade dress, a brown M & M's logo slightly off center to the left of the package, cascading M & M's candies on the right of the package and a tannish yellow lozenge keylined in brown containing the words "Peanut Butter" in brown. See plaintiff's exhibit 1.

54. The current M & M's peanut butter trade dress does not use certain design elements that Hershey uses on its complete Reese's trade dress. It does not use: yellow letters, the Reese's name, slanted lettering for the M & M's trademark, cursive script lettering, brown keylining around the M & M's logo, a brown peanut butter cup design and it does not use a brown bar of any kind with yellow letters. The M & M's package uses a tannish yellow flavor lozenge and Reese's package does not.

Duration and Extent of Use of the REESE'S Trade Dress

55. With the exception of the sawtoothed bar representing a peanut butter cup, the Reese's trade dress has been used for more than half a century, since 1947 or earlier, in connection with Reese's peanut butter cup candy.

56. This form of the Reese's trade dress was first used by the H.B. Reese Candy Company, which initially sold Reese's Peanut Butter Cups, and Hershey has continued this use since it acquired the H.B. Reese Candy Company in the 1960s. The sawtoothed bar first appeared in about 1974.

57. Hershey uses the Reese's trade dress with modifications on numerous products in addition to Reese's peanut butter cups. These include: Reese's Pieces candy; Reese's Crunchy Cookie Cups candy; Reese's Nutrageous candy; Reese's Peanut Butter Chips; Reese's Shell Ice Cream Topping; Reese's Sprinkles; and Reese's Peanut Butter. The same orange color pervades all of the Reese's confectionery products.

58. Hershey sells hundreds of millions of dollars of products under the Reese's trade dress annually. Estimated factory sales to trade customers of Reese's peanut butter cups alone in 1997 totaled over $450 million; Hershey's estimated factory sales of all products bearing the Reese's trade dress during the same period totaled over $600 million.

59. Hershey's estimated retail sales of its products bearing the Reese's trade dress in 1997 totaled over $800 million.

60. Since 1990, Hershey's estimated factory sales of goods bearing the Reese's trade dress have exceeded $4 billion. Since 1990, Hershey's estimated retail sales of goods bearing the Reese's trade dress have exceeded $5.2 billion.

61. The Reese's peanut butter cup is currently the number one selling candy product in grocery stores. Reese's peanut butter cups have been among the top handful of all confectionery products in the United States for a number of years. In each year from 1990 to 1997, estimated retail sales of Reese's peanut butter cups alone totaled $400 million or more.

62. In addition to the many products bearing the Reese's trade dress that Hershey sells, other companies use a form of that trade dress under license from Hershey in marketing their own products. These include: Reese's Peanut Butter Puffs cereal, which is sold by General Mills; Reese's Peanut Butter Cups Ice Cream, which is sold by Friendly's; and Breyer's Good Humor ice cream that also contains pieces of Reese's Peanut Butter Cups.

Duration and Extent of Advertising and Publicity for the Reese's Trade Dress

63. Advertisements featuring some form of the Reese's trade dress have been printed and disseminated for decades, since at least as early as 1954.

64. As early as the 1960s, Hershey circulated print advertising alerting consumers to "Look for Reese's Peanut Butter Cups wherever candy is sold, in the familiar orange wrapper" and to "Look for the bright orange pack."

65. In addition, Hershey's television advertisements close with a shot of the candy package displaying the Reese's trade dress. Hershey's current television advertising campaign, which runs on all of the national networks and many independent stations, actually opens with a large screen in the shade of orange used in the Reese's trade dress, with elements of yellow and brown, and closes with a package shot and the orange background again.

66. The Reese's trade dress also features prominently in magazine advertising for Reese's peanut butter cups candy. For example, certain of Hershey's ads are set against a field of the bright orange color used in the Reese's trade dress, with yellow and brown elements.

67. Other advertising and promotional materials for the Reese's products also feature the Reese's trade dress. These include billboard advertisements displaying a massive field of orange and point of sale materials, such as cardboard cards placed on shelves at retail establishments.

68. Hershey devotes millions of dollars to advertising for the Reese's products each year. For example, during the period from 1985 to 1997, Hershey spent at least $5 million on advertising per year for Reese's peanut butter cups alone. Hershey spent approximately $23 million in advertising for Reese's products in 1997, and has an advertising budget for the same amount in 1998. Since 1990 alone, Hershey has spent more than $121 million in advertising on Reese's products.

Channels of Trade in Which the Reese's trade dress is Used

69. The Reese's peanut butter cups product bearing the complete Reese's trade dress is sold nationwide.

70. The complete Reese's trade dress in some modified form appears on products marketed in broad classes of trade to consumers of all ages. Reese's products are sold in grocery stores, drug stores, mass merchandise stores, vending machines, concession stands, theaters, and by fund raisers. The greatest volume of Reese's peanut butter cups is sold in grocery stores.

Whether the Reese's trade dress is Registered

71. Plaintiffs have obtained a trademark registration with the U.S. Patent and Trademark Office for a mark that includes the complete Reese's trade dress as well as the word "REESE'S." (Registration No. 1,574,070).

Plaintiff's Survey Evidence

Dr. Jacoby's Survey

72. Jacob Jacoby, Ph.D. designed, conducted and testified with respect to the plaintiff's survey evidence.

73. The survey used respondents nationwide and attempted to determine (i) whether the abbreviated Reese's trade dress had achieved secondary meaning; (ii) whether the abbreviated Reese's trade dress is famous; and (iii) whether the current M & M's peanut butter trade dress is diluting the abbreviated Reese's trade dress.

74. Dr. Jacoby surveyed consumers of (i) candy containing chocolate, peanuts, almonds or peanut butter in a sugar-coated shell; (ii) chocolate and peanut butter candies; or (iii) other kinds of peanut butter candies. Dr. Jacoby surveyed two distinct groups of such candy consumers.

75. To conduct the survey, Dr. Jacoby created two stimuli, a representation of the abbreviated Reese's trade dress and a representation of the M & M's peanut butter trade dress at issue. The stimuli depicted their respective colors, as well as the proportion, arrangement and juxtaposition of those colors. The trade name "Reese's" was deleted and the sawtoothed top of the peanut butter cup shape which appears below the word mark was modified to a rectangular shape. The "M & M's" logo and the design mark consisting of "lentils" or "cascading M & M's" were deleted from the M & M's peanut butter stimulus. The lozenge beneath the "M & M's" logo was retained in the representation of the M & M's peanut butter package.

76. One group was shown the Reese's stimulus and the other the M & M's stimulus.

77. Both groups were also shown representations of six other brands; specifically, Snickers, Butterfingers, Nestle' Crunch, Skor, Goldenberg's Peanut Chews and Milka. Milka is sold only in Europe. As a control, the two groups were also shown a representation of an "inverted Snickers" package, a package with all the design elements of a Snickers but with different colors than the actual Snickers package.

78. In all of the representations, including the two test representations and all seven comparison representations, the designation "BRAND X" (in upper case, slightly slanted block letters) was used in place of the word marks or brand names in the colors of the real words, and repeating letters "ABC" were used to replace other textual elements.

79. The Reese's stimulus preserved the symmetrical pattern of the actual Reese's package; the "Brand X" logo was centered, where it would be on the real package. The logo was the same size, its type style was slanted, and the number of letters in Brand X equaled the number of letters in "Reese's."

80. The Mars stimulus did not have certain identifying characteristics present on the real M & M's peanut butter package. It substituted a slanted Brand X logo for the upright two-letter M & M's logo. It did not have the cascading M & M's candy pieces, which normally appear on the right side of the package. In the absence of the M & M's, the Brand X logo was not asymmetrical but symmetrical, centered on the stimulus rather than slightly off to the left.

81. Certain identifying characteristics were left on the comparison brands. The Milka stimulus retained all graphic illustrations and design elements of the actual package, including an illustration of a cow in the exact same form and location as on the commercial package. The Goldenberg's Peanut Chews representation retained a blue star in the same shape and location along with the words "bite size" inside the star. The Skor stimulus did not retain the picture of the candy bar that appears on the left hand side of the commercial package. However, the Brand X logo on that stimulus remained asymmetrical; it was not centered so that it would fill the space that would have been occupied by that picture.

82. The objective of the survey with respect to one group of respondents was to determine whether the Reese's trade dress had achieved secondary meaning and fame. The objective of the survey with respect to the second group of respondents was to test the impact of the M & M's peanut butter color scheme contained on the current trade dress (the trade dress at issue).

83. In the case of both groups of respondents, an interviewer first asked a series of questions in a "screener" to confirm that the individuals qualified as consumers of the relevant products and met other standard criteria for participation in such surveys. The interviewers then explained the procedure and asked a series of questions with respect to each of the representations of packages which were presented in random order. They were asked, "If you think you know, what brand of candy comes in this package?" After asking this question with respect to each of the representations, the interviewer questioned each respondent again with respect to each of the representations of packages to determine the reason for the answers to the first question. Specifically, the interviewer asked, "What, in particular makes you think it is (brand previously named)?"

84. The survey was conducted in 19 geographically dispersed markets spread across the four U.S. Census Regions.

85. The survey results with respect to the representation of the Reese's trade dress were as follows: 94% of the respondents identified the Reese's stimulus as Reese's and only Reese's, and slightly less than 88% of those respondents indicated color as the reason for the Reese's identification; two respondents misidentified products other than the M & M's product as the Reese's trade dress and indicated color as the basis for the misidentification.

86. With respect to the M & M's stimulus, the results were as follows: 7% correctly identified the representation of the M & M's peanut butter package as M & M's, 51% misidentified it as Reese's and slightly less than 49% indicated color as the reason for the misidentification.

87. Eighty-eight per cent correctly identified the Butterfinger representation, 81% correctly identified the Snickers representation, 15% correctly identified the Skor representation, 6% correctly identified the Goldenberg's peanut chews representation and 0% correctly identified the Milka representation.

88. With respect to the representations used with both groups for comparison purposes: there were only two respondents who identified any of the comparison representations as Reese's based on color.

89. On the basis of the survey results, Dr. Jacoby formed the opinion (i) that the Reese's trade dress has achieved a high degree of secondary meaning; (ii) that the high degree of secondary meaning establishes that the Reese's trade dress is indeed famous; and (iii) that the M & M's peanut butter color scheme is dilutive of the Reese's trade dress.

90. The stimuli were shuffled before being presented to each respondent. Therefore, one or both of the Milka and Goldenberg's stimulus was shown to at least 2/3's of the respondents in the M & M's group before the M & M's stimulus. The presence of design elements on these stimuli would suggest to respondents that corresponding design elements on other stimuli would be retained if they existed. Thus, when exposed to the M & M's stimulus they might incorrectly understand that that package had no design elements.

91. The Skor stimulus would also have been shown to at least 1/2 of the respondents prior to the M & M's stimulus, thus improperly suggesting that the Mars stimulus was an accurate portrayal of a centered brand name rather than the asymmetric logo that M & M's really used.

92. Dr. Jacoby's study provides no evidence that the ability of the real Reese's package to serve as a unique identifier of the Reese's goods has been weakened or lessened because the public now associates that real package with M & M's.

Dr. Ross' Pilot Survey

93. In rebuttal of the defendant's critique of the Jacoby survey, plaintiff introduced a pilot survey or pretest conducted by Ivan Ross, Ph.D. which counsel for plaintiff commissioned in advance of this litigation, once again, to address (i) whether the Reese's trade dress had achieved secondary meaning; (ii) whether the Reese's trade dress is famous; and (iii) whether the M & M's trade dress at issue is diluting the Reese's trade dress, and other issues.

94. Dr. Ross surveyed consumers of "candy." His survey was not limited to consumers of peanut or peanut butter candy or candy sold in a candy shell.

95. Dr. Ross surveyed three separate groups of candy consumers with about 30 persons to a group. Each group viewed a single representation of a candy package, either (i) the Reese's trade dress; (ii) the color scheme of the M & M's trade dress at issue; or (iii) a previous version of the M & M's peanut butter color scheme with a reddish background which was used when the product was first introduced in 1988 until about 1994.

96. Dr. Ross also employed representations of the packages which removed all source identifying indicia other than color and the arrangement, proportion and juxtaposition of colors. In doing so, Dr. Ross substituted a series of four "X"s, i.e., "XXXX," appearing in standard type style (without any slant or other stylization) for the brand names to represent the colors utilized in the actual brand names.

97. In his screening questions, Dr. Ross used the word "candy" only, not "peanut" or "peanut butter," to identify appropriate respondents. He thus interviewed respondents who had eaten any type of candy, not just peanut butter candy.

98. Dr. Ross' survey results with respect to the Reese's trade dress representation indicated that 91% of the respondents identified that representation with Reese's products (either alone or together with other products) and 81% of the respondents identified the representation with Reese's products exclusively. Dr. Ross' results with respect to the representation of the M & M's trade dress at issue indicated a 6% correct identification of the package and a 53% misidentification of the package as Reese's.

99. On the basis of the survey results, Dr. Ross concluded: (i) that the Reese's trade dress has achieved a high degree of secondary meaning; (ii) that the high degree of secondary meaning establishes that the Reese's trade dress is famous; and (iii) that the color scheme of the M & M's trade dress at issue is dilutive of the Reese's trade dress.

100. Dr. Ross' survey is an insufficient basis on which to enjoin Mars' use of the current trade dress.

Other Trade Dress Aspects of this Case

101. Even were it true that the colors orange or yellow or brown were functional, the combination of those colors in the particular manner represented by the Reese's trade dress that creates the unique overall appearance achieved by that combination is not functional.

102. The color orange does not function to communicate peanut butter flavor.

103. The actual color of peanut butter is a shade of brown or tan, not orange.

104. The color orange appears on the packages of various food products on the market that contain no peanut butter flavor.

105. The packaging of some peanut butter candies and peanut butter products does not have the color orange.

106. There are at least several examples of trade dress similar to the abbreviated Reese's trade dress in the food industry alone. See Defendant's exhibits 1, 2, 8, 13, 17, 41, 46 and 48.

Nonsurvey Evidence Bearing on Dilution

107. The parties' marks are not similar. Looking at only the colors and their juxtaposition on the dress, the plaintiff uses a bright yellow for its trade name, keylined in brown, with a brown rectangular bar underneath with yellow lettering in the bar. The defendant uses a solid brown for its logo, not yellow with brown keylining, and the logo extends over a different area of the package. Its lozenge reverse the colors of the Reese's rectangular bar and uses a tannish yellow.

108. The consumers are not sophisticated. The parties' products are inexpensive, and targeted to children between the ages of 8 and 17 years of age.

109. The Reese's trade dress is highly renowned by the relevant consuming public.

110. The M & M's logo is also highly renown but was not the subject of this motion.

111. Mars developed the trade dress at issue to mimic the background orange on the Reese's trade dress.

Conclusions of Law

1. The abbreviated Reese's trade dress is not famous or distinctive under the federal or Pennsylvania dilution laws.

2. The M & M's peanut butter color scheme does not dilute the abbreviated Reese's trade dress.

3. Hershey has not demonstrated a likelihood of success on the merits, an essential factor to obtain preliminary injunctive relief.

Discussion

Mars' Retroactivity Argument

We deal first with a threshold issue. Section 1125(c) was enacted on January 16, 1996. Mars argues that applying that section here is forbidden as a retroactive application of the law because Mars has used orange, brown, and yellow for many years beginning in the 1940s and used them again on its peanut butter variety in 1994 and on certain fun size packs of candy in the 1990s as well. Moreover, despite Hershey's counsel's objection in April 1995 that the color orange used on the peanut butter variety was almost identical to Reese's orange, Mars continued to use that color in combination with brown and yellow without further objection. Hence, Mars' use of these colors predates the federal prohibition of dilution and cannot be covered under this law.

In opposition, the plaintiff argues that it seeks only injunctive relief against Mars' trade dress, not damages for Mars' use, so that there is no retroactive bar to applying federal law here. See, e.g., Fuente Cigar, Ltd. v. Opus One, 985 F. Supp. 1448 (M.D.Fla.1997). Additionally, Mars' current trade dress is materially different from the dress introduced in 1994 as to the background orange and the use of yellow.

After review of the exhibits submitted by the parties, and without deciding the first point, we agree with the plaintiff as to the second point. Mars' current trade dress is materially different from previous versions. Thus, there is no retroactivity problem here; the federal law is being applied to conduct occurring after its enactment.

Standard For Granting a Preliminary Injunction

Hershey seeks a preliminary injunction, so it must satisfy the following standard:

A preliminary injunction is appropriate when a party demonstrates the following: (1) that he is likely to suffer irreparable injury in the absence of injunctive relief; (2) that he is likely to prevail on the merits. See Schulz v. United States Boxing Ass'n, 105 F.3d 127, 131 n. 6 (3d Cir.1997). In addition, the district court may consider the injunction's effect on the adverse party and other interested parties, as well as the general public interest. See Bradley v. Pittsburgh Bd. of Educ., 910 F.2d 1172, 1175 (3d Cir.1990); Arthur Treacher's Fish & Chips Inc. v. A & B Management Corp., 689 F.2d 1137, 1143 (3d Cir.1982).

Anderson v. Davila, 125 F.3d 148, 159 (3d Cir.1997). The movant must show both irreparable injury and a likelihood of success. See Schulz v. United States Boxing Ass'n, 105 F.3d 127, 131 n. 6 (3d Cir.1997). We will address the likelihood of success on the merits by examining the strength of the plaintiff's dilution claim.

The Reese's Trade Dress and Arguments Against Protectability

The parties dispute what should be included in the Reese's trade dress. We must therefore specify the exact nature of the dress, or the portions of it, that the plaintiff can protect before analyzing the dilution claim.

As the defendant points out, in most cases, the Reese's Peanut Butter Cup trade dress would consist of its overall appearance, including all the elements of the dress. "Trade dress is a complex composite of features' and `[t]he law of unfair competition in respect to trade dress requires that all of the features be considered together, not separately." American Greetings Corp., 807 F.2d at 1141 (quoting SK&F, Co. v. Premo Pharmaceutical Labs., 481 F. Supp. 1184, 1187 (D.N.J. 1979), aff'd, 625 F.2d 1055 (3d Cir.1980))." Duraco Products, Inc. v. Joy Plastic Enterprises, Ltd., 40 F.3d 1431, 1439 (3d Cir.1994) (internal quotation marks omitted); Freixenet, S.A. v. Admiral Wine & Liquor Co., 731 F.2d 148, 149-50, 153 (3d Cir.1984)(treating the plaintiff's trade dress for its wine bottle as including all the elements of the wine bottle's appearance, including its frosted black color, the foil covering the cork, and the characteristics of the two labels on the bottle).

Thus, the trade dress could be described as a particular shade of background orange, the Reese's logo (its trade name in a slanted handwritten style in bright yellow with brown "keylining;" i.e., the yellow letters outlined in brown) and a brown bar with a sawtoothed top (imitating in two dimensions the top edge of a peanut butter cup) directly underneath the logo with the bar containing the message in bright yellow letters "2 peanut butter cups." This is the complete Reese's trade dress.

However, the plaintiff asserts that certain portions of this trade dress, the abbreviated trade dress, have achieved distinction in their own right and are therefore entitled to trade dress protection. They maintain that this trade dress "consists of a unique shade of orange as the principal element in combination with and in juxtaposition to the colors yellow and brown, in an arrangement creating a distinctive overall appearance." (Supporting brief at p. 5; plaintiff's proposed finding of fact no. 8, second sentence). Contrary to the defendant's contention, Hershey asserts it is not protecting the mere use of these colors on the package but the "particular arrangement ... including the relative proportions of each color, the placement of each color on the package and the juxtaposition of each color in contact with the others." (plaintiff's reply brief at p. 1) (footnote omitted)(emphasis in original).

This description of the trade dress omits the trade name, the slanted handwritten style of the trade name, the sawtoothed top of the brown bar, and the words inside the bar. The defendant asserts that the plaintiff cannot abbreviate the trade dress in this fashion, that Hershey must include all the elements of the dress for the purpose of its dilution claim.

We disagree with the defendant, although we confess to some serious doubts. To begin with, most of the cases it has cited to support this argument, Wm. Wrigley Jr. Co. v. City Sweats, Inc., 41 U.S.P.Q.2d 1380 (N.D.Ill. 1996); Sunbeam Products, Inc. v. West Bend Co., 39 U.S.P.Q.2d 1545, 1549 (S.D.Miss. 1996); Golden Bear Int'l, Inc. v. Bear U.S.A., Inc., 969 F. Supp. 742 (N.D.Ga.1996), did not directly address the issue. Even Merriam-Webster, Inc. v. Random House, Inc., 35 F.3d 65 (2d Cir.1994), where the Second Circuit's rejection of a state dilution claim was based on examination of the entire trade dress, including the logo and trade name, did not do so. In other words, the plaintiffs in those cases were satisfied to litigate their claims by use of the entire trade dress. In the instant case, Hershey asserts that it has established rights in certain elements of its trade dress.

Second, trade dress must be treated like a trademark for the purposes of the Lanham Act. See Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 112 S. Ct. 2753, 120 L. Ed. 2d 615 (1992) (section 1125(a) protects trade dress from infringement). A trademark is defined broadly as "any word, name, symbol, or device or any combination thereof...." 15 U.S.C. § 1127. A certain combination of colors is protectable. Id. As the plaintiff also points out: "The fact that the owner of a mark uses it in association with accessory symbols or words does not deprive him of what the public recognizes as a mark. In an early case, Judge Hand recognized that a seller is entitled to protection of as many separate marks as the public in fact recognizes as trademarks...." McCarthy, supra, § 7:2, at 7-5 (footnote omitted). As the Seventh Circuit stated, a trademark:

need not be the name of the brand; it need not even be a word; it can be a slogan, a symbol, a combination of words and symbols, an ornamental feature, a distinctive shape, or something else intended to remind the consumer of the brand. See 1 McCarthy, supra, ch. 7.

Blau Plumbing, Inc. v. S.O.S. Fix-It, Inc., 781 F.2d 604, 609 (7th Cir.1986).

In the instant case, the plaintiff asserts that "something else" besides its complete trade dress has come to signify its brand to the consuming public. We cannot say as a matter of law that it should be precluded from trying to protect this particular combination of elements as a mark. We note that in First Brands Corp. v. Fred Meyer, Inc., 809 F.2d 1378 (9th Cir.1987), the Ninth Circuit addressed the plaintiff's claim based on infringement of a trade dress consisting of a yellow, F-style jug for antifreeze, and only after deciding that the jug had not acquired a secondary meaning did the court also look to the labels of the competing brands to determine that there would be no likelihood of confusion.

Mars has also argued that the trade dress is not protectable because the colors are functional in that they distinguish the product and they stimulate appetite.

Functional elements are not protectable. See American Greetings Corp. v. Dan-Dee Imports, Inc., 807 F.2d 1136 (3d Cir.1986). However, the overall combination of features may be nonfunctional and thus protectable. Id. The plaintiff seeks to protect not just the colors orange, brown and yellow, but peculiar shades of these colors in the combination found on its trade dress. We therefore reject the defendant's argument.

The Dilution Claim

A. Fame.

As noted above, section 1125(c)(1) lists certain factors we must consider on the fame element of a blurring claim.

1. Factor A, the degree of inherent or acquired distinctiveness of the mark, and Factor F, the degree of recognition of the mark in the trading areas and channels of trade used by the marks' owner and the person against whom the injunction is sought.

The plaintiff has combined these factors for analysis and we will follow its lead. To support its trade dress's claim of distinctiveness and recognition by candy consumers, the plaintiff relies on the nationwide survey it commissioned from Professor Jacob Jacoby of New York University. Using a stimulus that deleted the Reese's trade name and certain other aspects of its trade dress, so that the portions of its trade dress at issue in this case could be tested, Professor Jacoby found that 94% of those surveyed recognized it. Based on these results, Dr. Jacoby testified that in his opinion the abbreviated Reese's trade dress has achieved a high degree of secondary meaning and that it is famous.

In opposition, the defendant attacks the stimulus used to test the trade dress. Mars contends that the stimulus, in addition to the colors as they appear on the Reese's peanut butter cup package, used other elements of the trade dress. These purportedly included using "Brand X," which had the same number of letters as in Reese's, slanting Brand X at the same angle as Reese's and using almost the same script style for Brand X as Reese's, putting the Brand X in the same location as it appears on a real Reese's package, and reproducing the sawtoothed bar as a rectangular form of the same size and color. Mars contends that these additional elements meant that the stimulus did not test at all for the claimed trade dress and gave the survey respondents additional clues for the source of the dress.

Second, Mars relies on the results for the stimulus used to represent the M & M's peanut butter trade dress. Since this stimulus left off certain specific elements of that trade dress, the cascading M & M's on the right side of the package and the geometric shape of the M & M's logo itself, the defendants contend that it is a truer test of the abbreviated Reese's trade dress, not the M & M's dress. Only 49* of the respondents surveyed for the Mars stimulus identified it as Reese's. Mars contends that this is not a sufficient percentage for fame or distinctiveness, as opined by Dr. Jacoby.

We disagree with both of the defendant's arguments. Its first argument erroneously asserts that the Reese's stimulus incorporates other elements of the trade dress. A review of the stimulus reveals the defendant's error. The slant is simply not important and using "Brand X" is proper because it roughly equals the space taken up by the trade name. Otherwise, the stimulus does not use the same script style nor does it have the sawtoothed illustration of a brown peanut butter cup.

The second argument erroneously assumes that Hershey is testing merely for the combination of colors. It is not; the colors must correspond to their placement on the trade dress. The M & M's stimulus does not have the same correspondence and cannot be used to defeat the results of testing the more accurate stimulus.

Hence, we conclude that the Hershey's trade dress has satisfied Factor A and F for assessing fame. The dress has acquired distinctiveness and it is recognized by purchasers of both Hershey and Mars products.

2. Factor B: the duration and extent of use of the mark in connection with the goods or services with which the mark is used.

The defendant does not attempt to contest this factor, and it is easily satisfied. As the plaintiff essentially argues in its supporting brief at page 10, a portion of the Reese's trade dress, the background orange and the brown keylined yellow, has been in use for half a century. Hershey's factory sales of goods bearing the Reese's trade dress or a modified version of it were approximately $628 million in 1997 and have exceeded $4 billion since 1990, and estimated retail sales since 1990 are more than $5.2 billion. The Reese's peanut butter cups, on which the trade dress is principally used, has been among the top handful of confectionery products in the United States in recent years, and for many years has been the number one or two selling confectionery item in this country.

3. Factor C: the duration and extent of advertising and publicity of the mark.

To satisfy this factor, plaintiff points out that it has spent more than $120 million nationwide advertising its Reese's products since 1990 alone. Hershey argues that, since these advertisements stress the trade dress, see Findings of Fact 63-67, frequently including reproductions of the package for the Reese's peanut butter cups, its advertising expenditures support a finding that its mark is famous.

In opposition, the defendant contends that mere advertising expenditures alone, without an advertising strategy calling attention to the trade dress, is insufficient, citing First Brands Corp., supra.

We do not believe that the advertisements must call attention to the dress but do believe that this factor is not especially strong here. It is true that the recent print advertisements feature the dress, but they strongly emphasis the background orange, not the overall color scheme, since the orange is used as the background of the advertisements. Additionally, the earlier print ads specifically call attention to the orange package. In these circumstances, the plaintiff's advertisements satisfy factor C, but weakly.

4. Factor D: the geographical extent of the trading area in which the mark is used.

This factor favors the plaintiff. The defendant does not contest that the mark appears on Hershey's products nationwide.

5. Factor E: the channels of trade for the goods or services with which the mark is used.

This factor favors the plaintiff. The mark appears on Hershey's products wherever candy is sold.

6. Factor G: the nature and extent of use of the same or similar marks by third parties.

The defendant argues that this factor does not favor the plaintiff because numerous third-party food companies use marks that are similar to the Reese's abbreviated trade dress by employing a similar combination of the colors orange, brown, and yellow.

The plaintiff counters in two ways. First, Hershey argues that the trade dress the defendant relies on is not the same or similar to the Reese's trade dress at issue in this case. Second, it is not sufficient for the defendant merely to point out third-party uses of certain elements of the contested trade dress. Hershey contends that Mars must also establish that such third-party use confuses consumers as to source. As it quotes from AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1537 (11th Cir.1986) (footnote omitted), "third party use of one or more suggestive or arbitrary elements of a plaintiff's trade dress renders that trade dress indistinct only if the third party use is so extensive and so similar to the plaintiff's that it impairs the ability of consumers to use the trade dress of the products to identify their source." As further support, plaintiff also cites Paramount Pictures Corp. v. Dorney Park Coaster Co., 698 F. Supp. 1274, 1280 (E.D.Pa.1988), and Jaret Int'l. Inc. v. Promotion in Motion, Inc., 826 F. Supp. 69, 75 (E.D.N.Y.1993).

We reject the plaintiff's argument. First, these three cases deal with infringement under section 1125(a), and while the framework for analyzing the distinctiveness of a mark can be borrowed from such cases for use in a section 1125(c) case, see Mead Data Central, supra, 875 F.2d at 1030 (citing P.F. Cosmetique, S.A. v. Minnetonka, Inc., 605 F. Supp. 662 (S.D.N.Y.1985)), it is obvious that the relevant passages are dealing with confusion, the essential element of an infringement claim but irrelevant to a dilution claim. Second, and in a related point, because we deal here with section 1125(c) and the statutory criteria for determining fame, we need only follow the statute, and if it simply requires us to determine the nature and extent of the use of the same or similar marks by third parties, we need not go further and determine confusion. Other courts have not imposed this additional requirement in connection with Factor G. We note also that they have looked outside the relevant market for similar marks. See Sports Authority, Inc. v. Abercrombie & Fitch, Inc., 965 F. Supp. 925, 941 (E.D.Mich.1997) ("authority" used outside relevant market); Trustees of Columbia University v. Columbia Healthcare Corp., 964 F. Supp. 733, 750 (S.D.N.Y.1997) ("Columbia" is used outside the medical or healthcare industry); Star Markets Ltd. v. Texaco, 950 F. Supp. 1030, 1035-36 (D.Haw. 1996) ("Star" has been used outside the supermarket industry); Knaack Mfg. Co. v. Rally Accessories, Inc., 955 F. Supp. 991, 998, 1002, 1005 (N.D.Ill.1997).

Finally, the cases are factually distinguishable. Paramount Pictures Corp. dealt with a word mark, "Top Gun," not trade dress. AmBrit and Jaret Int'l dealt with the plaintiffs' complete trade dress (including texture, colors, logos) and the defendants in those cases were defending by pointing out that various elements of the dress, but not the complete dress, appeared on third-party packages. The courts noted that such third party use of one or more suggestive or arbitrary elements of a plaintiffs trade dress could render the plaintiffs dress indistinct only if the third-party use met the test of sufficient similarity.

In the instant case, however, the plaintiff has detached itself from its complete trade dress. It is claiming a protectable mark in an abbreviated trade dress, a particular shade of background orange, bright yellow lettering of an unspecified style, and a rectangular bar below the lettering containing more yellow lettering of some unspecified kind. This type of trade dress claim allows greater leeway in determining what is a similar dress. We believe the defendant has pointed out at least several examples of similar dress in the food industry alone. See Defendant's exhibits 1, 2, 8, 13, 17, 41, 46 and 48.

7. Factor H: whether the mark was registered under the Act of March 3, 1881, or the Act of February 20, 1905, or on the principal register.

The Reese's abbreviated trade dress is not federally registered. Federal registration is not a requirement to protect a mark, but a failure to register counts against a finding of fame. A person would be expected to register a famous mark. See McCarthy, supra, § 24:92, at 24-158.

Weighing the Factors

Factors A, B, D, E and F favor the plaintiff. Factor C does so, but weakly. Factors G and H favor the defendant.

As noted above, however, we do not consider the fame analysis to be a mere arithmetical exercise. We would expect most businesses hoping to succeed under the federal dilution act to be able to satisfy factors A, B, C, D, E and F. Many companies doing business nationwide should be able to.

The important factors for us are G and H. The trade dress Hershey seeks to protect here, a particular combination of orange, brown and yellow, is similar to enough third-party marks in the food industry that we hesitate to preliminarily enjoin Mars' use of the color scheme on its M & M's peanut butter dress. Despite the evidence from the consumer survey, since plaintiff is not relying on its trade name, logo, and sawtoothed bar, the dress does not seem worthy of protection as famous. Cf. P.F. Cosmetique, S.A. v. Minnetonka Inc., 605 F. Supp. 662, 668 (S.D.N.Y.1985) (hair-care product trade dress was weak when it used elements conveying a "clinical image" common in the industry).

Factor H is also significant because Hershey has registered its complete trade dress and has attempted to register its background orange but has never attempted to register the trade dress at issue here.

We conclude that the abbreviated Reese's trade dress at issue in this case has not been shown to be famous and Hershey has failed to establish the first element of its dilution claim.

B. Assuming that Fame Had Been Established, Whether Mars' Use Causes Dilution.

We will assume for the purpose of further analysis that plaintiff established the fame of its trade dress. We proceed to examine whether it has satisfied the remaining elements of a dilution claim. The plaintiff relies on Dr. Jacoby's survey evidence (buttressed in rebuttal by a survey conducted by Dr. Ivan Ross) and, in the alternative, the Mead Data Central six-factor test.

1. The Survey Evidence.

As noted in our findings of fact, Dr. Jacoby surveyed two different groups of consumers of peanut butter candy. He showed a representation of the Reese's peanut butter cup trade dress to one group and a representation of the Mars peanut butter M & M's trade dress to the second group. The major change from the actual trade dress was that these representations did not contain the word mark or logo of either candy. They were simply identified as "Brand X," in the same slanted printed style.

Both groups were also shown representations of six other brands; specifically, Snickers, Butterfingers, Nestle' Crunch, Skor, Goldenberg's Peanut Chews and Milka, a product sold only in Europe. As a control, the two groups were also shown a representation of an "inverted Snickers" package, a package with all the design elements of a Snickers but with different colors than the actual Snickers package. These representations similarly did not contain trade names, word marks, or logos, and also were simply identified as "Brand X," in the same slanted printed style.

Fifty-one per cent of the group that was shown the M & M's Peanut Butter representation mistakenly identified it as a Reese's product. Slightly less than 49% of those persons cited color as the reason for their misidentification. On this basis, Dr. Jacoby concluded that the M & M's peanut butter color scheme is diluting the distinctiveness of the Reese's abbreviated trade dress because a substantial number of people are seeing a representation of the M & M's package and connecting it to Reese's.

Mars has attacked this survey on the following grounds. First, it is fatally flawed because it does not compare the real trade dresses as they appear in commerce. Mars argues that since the statutory language only allows a trademark holder to equitable relief against "another person's commercial use in commerce" of a diluting mark, 15 U.S.C. § 1125(c), it must be the marks that are used in commerce that are compared, not a fabricated representation that omits crucial elements of the trade dress. Thus, Dr. Jacoby's survey is useless.

We reject this argument. First, we have already decided that a trademark holder can seek to protect a combination of elements making up less than the complete trade dress if, for whatever reason, the holder believes that combination has acquired secondary meaning. Second, the quoted statutory language was not intended to impose an evidentiary standard on a trademark holder as to how to test the distinctiveness of the mark or whether it has been diluted. As the plaintiff points out, it was intended to protect noncommercial use of the mark from a dilution claim, such as use of the mark in commercial speech or consumer product reviews. See McCarthy, supra, § 24:97.2, at 24-171.

Second, Mars attacks the methodology by criticizing the questions asked of the respondents. Defendant asserts that respondents should have been asked a "top of the mind" question, an open-ended question that merely inquired of the respondents what the stimulus reminded them off, rather than the specific question about who may have made the candy whose package was represented. Mars submits it was "unprecedented" for Dr. Jacoby to have used a standard secondary meaning test (usually used to test for confusion of source) to test for dilution.

There is no standard criteria for surveying for dilution. We think that the question Dr. Jacoby used was not objectionable since he allowed the respondents the option of saying they did not know.

Third, Mars asserts that the stimulus used for the abbreviated Reese's trade dress improperly included elements of the complete Reese's trade dress and the stimulus used for the M & M's trade dress improperly deleted elements of the actual trade dress. Defendant also argues there were defects in the preparation of the stimuli for the other brands and the results were accordingly distorted.

We note that the Reese's stimulus did use elements other than those involving the colors orange, yellow and brown in a certain juxtaposition; specifically, the symmetrical pattern of the actual Reese's package and the number of letters in Brand X is equal to the number of letters in "Reese's," and the sawtoothed illustration of a brown peanut butter cup was reproduced as a rectangular form of the same size and color.

The M & M's trade dress did not have the M & M's mark and substituted a slanted Brand X logo although the M & M's logo is only two letters printed in upright lower case lettering. It did not have the cascading M & M's candies on the right side of the package. The Brand X logo was centered in the middle of the stimulus rather than slightly off to the left.

As to the other brands, the Skor stimulus did not retain the picture of the candy bar that appears on the left hand side of the commercial package, but, unlike the M & M's stimulus, the Brand X logo on that stimulus was not extended to fill the space that would have been occupied by that picture. The Goldenberg's Peanut Chews stimulus kept the blue star in the same shape and location as the actual package and the words "bite size" were retained inside the star. The Milka stimulus retained the illustration of a cow in the exact same form and location as on the commercial package.

According to Mars, the significance of these aspects of the survey is that the absence of visual clues on the M & M's stimulus, like the cascading M & M's, with the presence of analogous clues on the other brand stimuli, could have misled respondents into concluding that the M & M's stimulus was a representation of the Reese's trade dress. We agree with this conclusion and discount the Jacoby survey.

The plaintiff has attempted to buttress the Jacoby survey with a survey done by Dr. Ivan Ross to assist Hershey's counsel in preparing the case. It was not intended as evidence in the case-in-chief and was only introduced in rebuttal to counter the defendant's criticism of the stimuli used in the Jacoby survey. The Ross survey is described in Findings of Fact 93-99. According to the plaintiff, Dr. Ross obtained essentially the same results as Dr. Jacoby even though he did not use the same potentially confusing stimuli for other brands.

Mars objects to the Ross survey for the same reasons it objected to the Jacoby survey and adds that Dr. Ross has miscalculated some findings. We have rejected most of the defendant's objections to the stimuli used in the Jacoby survey, and it appears that because of the nature of the stimuli used in the Ross survey, they were not misleading in the same sense as the objectionable portions of the Jacoby stimuli.

However, we choose not to rely on the Ross "pretest" to enjoin the defendant. As noted, the plaintiff did not use this survey in its case-in-chief, and the plaintiffs other evidence has not impressed us sufficiently that we should issue preliminary injunctive relief in these circumstances; a preliminary motion seeking to protect an abbreviated trade dress made up only of colors in a certain arrangement.

2. The Mead Data Central Six-Factor Test.

To reiterate, the Mead Data Central six-factor test is as follows:

1) similarity of the marks

2) similarity of the products covered by the marks

3) sophistication of consumers

4) predatory intent

5) renown of the senior mark

6) renown of the junior mark

875 F.2d at 1035.

As noted above, some commentators have questioned the relevance of all of these factors to a dilution claim. In our view, only the second factor is irrelevant. While the quintessential example of dilution is the use of a famous mark on a product not made by the trademark holder, dilution can apply to competitors, as the federal and state statutes make clear, so whether the products are similar or not adds nothing to the analysis. We turn then to the remaining factors.

Under the first criterion, the marks need not be identical, see Mead Data Central, supra, 875 F.2d at 1035, but they should be substantially similar, B.E. Windows Corp., supra, 937 F.Supp. at 211; Elvis Presley Enterprises Inc. v. Capece, 950 F. Supp. 783, 798 (S.D.Tex.1996); American Express Co. v. CFK, Inc., 947 F. Supp. 310, 317 (E.D.Mich. 1996), and "the more similar the marks, the higher the likelihood of dilution." Mead Data Central, supra, 875 F.2d at 1035.

Relying, among other cases, on Merriam-Webster, Inc. v. Random House, Inc., 35 F.3d 65 (2d Cir.1994), the defendant argues that the marks are not substantially similar because the real trade dress of the Reese's Peanut Butter Cup and the Mars' peanut butter M & M's must be compared, not just the elements that the plaintiff claims make up a distinctive trade dress on their own. Merriam-Webster is strong support for the defendant since in that case the Second Circuit borrowed from the federal law dealing with confusion under section 1125(a) to decide under the New York dilution statute whether the trade dresses of two dictionaries were similar. The court of appeals decided they were not by comparing the complete trade dress, including logos, trade names and graphics.

Following Merriam-Webster here, the defendant argues that:

Here the parties real packages are distinctly different, the impressions conveyed by the combination of the respective logos, the quite different names or marks, the different graphics, including the cultural icons of the M & M's® candy pieces and/or M & M's® Characters preclude a finding of similarity.

(Defendant's opposing brief at p. 16) (emphasis in original).

Consistent with its approach of avoiding the logos and concentrating only on the colors and their relative positions on the trade dress, the plaintiff asserts that the marks are very similar:

sophisticated testing found the principal color of both packages to be virtually the same color [orange]. Both parties' trade dresses also employ brown, which is in direct contact with the orange background. Moreover, the Reese's trade name is written in yellow letters key-lined in brown; the "lozenge" on the Mars' package is similarly yellow and is similarly key-lined in brown. Indeed, both packages include the contrasting colors of yellow and brown in an oblong shape centered beneath the brand name. On both packages, the orange background extends to the edges of the rectangular face of the package, with no border or other outlining.

(Plaintiff's supporting brief at pps. 15-16) (brackets added).

Examining the complete trade dress is an important part of the analysis of an infringement claim under section 1125(a) because consumer confusion as to the source of the product is an essential element of that claim. See Versa Products Co., supra. Thus, if trade names and logos identify the source, and do so prominently enough, there can be no confusion. See Bristol-Myers Squibb Co. v. McNeil-P.P.C., Inc., 973 F.2d 1033 (2d Cir.1992).

Assuming that we were correct in deciding that a trade dress claim can be based on something less than the complete trade dress, it is more difficult to discern what purpose examining the trade names and logos would have in a dilution case. Confusion is not an element of a dilution cause of action. In the dilution context, the consumer understands there are two different sources; the focus is instead on the consumer's connection of the same mark with the different sources.

Nonetheless, we need not resolve this issue. Even if we look only at the aspects of both trade dresses the plaintiff identifies, we conclude they are not substantially similar. Looking at only the colors and their juxtaposition on the dress, the plaintiff uses a bright yellow for its trade name, keylined in brown, with a brown rectangular bar underneath with yellow lettering in the bar. The defendant uses a solid brown for its logo, not yellow with brown keylining, and the logo extends over a different area of the package. Its lozenge reverse the colors of the Reese's rectangular bar and uses a tannish yellow. The overall look is different.

The only element the trade dresses have in common is the background orange, but the plaintiff is not claiming protection for orange alone on its preliminary injunction motion.

The next criterion is sophistication of the consumers. This factor works differently in a dilution claim than an infringement claim. Because confusion is an element of an infringement claim, the less sophisticated the consumers, the more likely there is infringement. See Versa Products Co., supra, 50 F.3d at 214. However, the reverse is true in a dilution claim because the more sophisticated the consumers the more likely they will recognize that the mark has become associated with separate sources. See McCarthy, supra, § 24:70, at 24-120 ("dilution presupposes no mental confusion over affiliation or connection, but rather a state of mind that recognizes independent sources and affiliation").

In the instant case the plaintiff argues that its primary customers are in the age group from eight- to 17-years-old, that they are unsophisticated, and that this factor favors Hershey.

However, based on our approach to the sophistication issue, this factor favors the defendant rather than the plaintiff. An unsophisticated consumer would confuse the products, and would not realize that the same mark was being blurred by use on what they knew were two different sources.

The next factor is predatory intent. In the infringement context, the Third Circuit has decided the defendant's intent to copy the plaintiff's trade dress is a weak indicator of infringement. See Versa Products Co., supra, 50 F.3d at 206. Even this weak indicator does not favor the plaintiff. In the instant case, the defendant did not intend to copy the abbreviated trade dress, just the background orange.

The next factor is renown of the senior mark and it favors the plaintiff. We accept for the purposes of this motion that Dr. Jacoby' survey has established the renown of the plaintiff's mark, defined by the plaintiff as its trade dress in the colors of the Reese's package in a certain juxtaposition, and that this factor favors plaintiff. "The greater the fame of the mark, the greater the likelihood that consumers would make mental associations between the famous mark and similar junior marks." Utah Division of Travel Development, 955 F.Supp. at 620. See also Mead Data, 875 F.2d at 1038; B.E. Windows, 937 F.Supp. at 213.

The final factor is the renown of the junior mark. This factor is relevant to a blurring claim because when both the junior and senior marks are well known, it is easier for the junior mark, as an already recognized source, to adopt certain elements of the senior mark, and thus cause consumers to identify the mark with two different sources. As argued by the plaintiff in relying on Utah Division of Travel Development, supra: "Presumably, the potential for dilution is greater where the gap in renown is smaller." 955 F.Supp. at 621 (emphasis added). Here, the plaintiff argues that:

both the REESE'S Trade Dress and the M & M's logo are famous. Indeed, because of the fame of the M & M's brand name, the color scheme which for decades has been associated uniquely with Reese's will quickly be associated with M & M's as well. Consumers will therefore no longer be able immediately to recognize Reese's products by the color scheme alone. This is the essence of dilution.

(Plaintiffs supporting brief at p. 18).

The problem with the plaintiff's analysis is that it relies on the M & M's word mark and logo to establish the renown of the M & M's peanut butter trade dress. This is improper in a case where the plaintiff is claiming that something less than that name and logo is blurring its abbreviated trade dress in its colors and their juxtaposition.

We conclude that the plaintiff has failed to satisfy our modified Mead Data Central test for blurring.

Conclusion

Since the plaintiff has failed to establish a likelihood of success on the merits, an essential element for granting a preliminary injunction, we will deny the motion seeking preliminary injunctive relief.

NOTES

[1] Our analysis of the federal law will apply to the state dilution claim as well.