1308 Gerald R. Weber, Atlanta, GA, Paul Alan Levy, Public Citizen Litigation Group, Washington, DC, for Plaintiff.

Claudia T. Bogdanos, Partha P. Chattoraj, Robert L. Raskopf, Quinn, Emanuel, Urquhart, Oliver & Hedges, LLP, New York, NY, Kenneth R. Ozment, John M. Bowler, Troutman Sanders, Atlanta, GA, for Defendant.

ORDER

TIMOTHY C. BATTEN, Sr., District Judge.

This action arises from the contention of Defendant Wal-Mart Stores, Inc. that its registered `trademarks "WALMART"; "WAL-MART"; and "WAL★MART"; its registered word mark "ALWAYS LOW PRICES. ALWAYS"; and its "wellknown smiley face mark" were infringed by Plaintiff Charles Smith's anti-Wal-Mart merchandise. Smith petitions the Court to declare his activities legal so that he may resume them without fear of incurring liability for damages; Wal-Mart counterclaims for an award of ownership of Smith's Wal-Mart-related domain names, an injunction precluding Smith from making commercial use of any designation beginning with the prefix "WAL," and an award of nominal damages. Both parties pray for costs and attorneys' fees.

Pending before the Court are Smith's motion for summary judgment [76], Wal-Mart's motion for summary judgment [77], Smith's motion in limine to exclude Wal-Mart's expert witness evidence [78], and Wal-Wart's motions in limine to exclude evidence from Smith's two rebuttal expert witnesses [81, 82].

I. Background

Wal-Wart Stores, Inc., which had approximately $283 billion in gross domestic revenue in fiscal year 2008,[1] sells retail goods and services through a large chain of nearly 6500 physical stores and its Internet site, www.wal-mart.com. The company also owns mad operates additional domain names, including www. walmartstores.com and www.walmartfacts. com, that link to the www.wal-mart.com website.

The company owns and has continuously used the well-known WAL-MART trademark 1309 and service mark in the United States for retail department store services since 1962 and has longstanding registered trademark rights in the marks. WAL-MART and WALMART are used alone or in conjunction with Wal-Mart's blue five-pointed star. Wal-Mart also owns a trademark registration in the word mark "ALWAYS LOW PRICES. ALWAYS."

The registered WAL-MART marks are usually displayed in Wal-Mart's blue, block-letter font, and when the word "ALWAYS" is used at the end of the phrase "ALWAYS LOW PRICES. ALWAYS," it is displayed in a red, italicized font, placed at approximately a forty-five degree angle after the horizontal blue, block-letter phrase "ALWAYS LOW PRICES." The company also often uses a yellow "smiley face" in conjunction with its registered marks.

Wal-Mart uses its marks extensively on its buildings, advertising, community support programs, and in association with its credit card, vision care, vacation planning, pharmacy and other services.

Smith is an avid and vocal critic of Wal-Mart. He believes that Wal-Mart has a destructive effect on communities, treats workers badly, and has a damaging influence on the United States as a whole — an influence so detrimental to the United States and its communities that Smith likens it to that of the Nazi regime. With the goals of stimulating discussions about Wal-Mart and getting others of like mind to join him in expressing strongly negative views about Wal-Mart, Smith created various designs and slogans that incorporated the word "Walocaust,"[2] a word Smith invented by combining the first three letters of Wal-Mart's name with the last six letters of the word "holocaust."

Smith created four basic Walocaust designs.[3] One design depicted a blue stylized bird modeled to resemble a Nazi eagle grasping a yellow smiley face in the same manner that a Nazi eagle is typically depicted grasping a swastika. Above the bird image, the word "WAL★OCAUST" was printed in a blue font comparable to that commonly used by Wal-Mart. Two designs were text only: one design read, "I ? WAL★OCAUSTH They have FAMILY VALUES and their ALCOHOL, TOBACCO and FIREARMS are 20% OFF"; and another design read, "WAL★OCAUSTH Come for the LOW prices[,] stay for the KNIFE fights." The fourth was a graphical design that depicted the word WAL★OCAUST on a Wal-Mart-like storefront that also included the Nazi eagle image, a poster advertising family values and discounted alcohol, tobacco and firearms, and other images commenting negatively on Wal-Mart.

Smith does not claim any exclusive right to his Wal-Mart-related creations; in fact, he says that he would like to see the general public use the terms freely. He hoped that the word "Walocaust" would become such a commonly used term to describe Wal-Mart that it might eventually appear in the dictionary.

In late July 2005, to help draw attention to his Walocaust concept and his views 1310 about Wal-Mart in general, Smith arranged for some of his designs to be printed on t-shirts and other items like mugs, underwear, camisoles, teddy bears, bumper stickers and bibs that could be purchased through www.CafePress.com.[4] He also placed text on his CafePress account home page that included harsh statements about Wal-Mart, such as "Walocaust: The World is Our Labor Camp. Walmart Sucks" and

Say hello to the Walocaust, say hello to low prices, say hello to child labor, say hello to unpaid overtime, say hello to 60 hour work weeks, say hello to low pay, say hello to poverty[.] Say hello to the Walocaust, say goodbye to health insurance, say goodbye to weekends, say goodbye to vacation, say goodbye to retirement, say goodbye to living indoors[.] The Walocaust: coming soon to your occupation. A real web site is coming soon. Contact: Walocaust@ yahoo.com[.]

Although CafePress offered the option to open a "basic shop" at no charge, which would have allowed Smith to sell his items at cost, Smith instead chose to pay $6.95 per month for a "premium account," which offered several automated functions that allowed him to set up a website without knowing HTML code. This enabled Smith to display on his CafePress website his products, his other designs, and content more fully expressing his views about Wal-Mart. It also enabled him to have his www.walocaust.com domain name bring viewers to the home page of his CafePress account. In hopes that profit from his CafePress site would cover the costs of his premium fees and domain name, Smith retained CafePress's default "medium" mark-up setting, which set his items' sale price at approximately thirty percent above cost.

The only actions that Smith undertook to promote his CafePress Walocaust account and the designs he had available on it were to tell family and friends about it, to send word to discussion groups whose participants he thought would be sympathetic, to start a Walocaust discussion group, and to accept an unsolicited offer of a link from a website where dissatisfied Wal-Mart employees would go to vent their frustrations with the company. When his Walocaust website became active, Smith also included a link from it to his CafePress account.

On December 28, 2005, and again on February 1, 2006, Wal-Mart wrote to Smith and to CafePress, asserting that Smith's Walocaust CafePress webpage was violating Wal-Mart's trademark rights, and demanding that they cease selling all products imprinted with his various anti-Wal-Mart designs. Wal-Mart also objected to Smith's registration and use of the 1311 domain name www.walocaust.com, demanding that Smith cease using the domain name and transfer ownership of it to Wal-Mart.

In response, CafePress removed all of Smith's Wal-Mart-related merchandise from his online store so that only non-Wal-Mart-related merchandise remained available at www.cafepress.com/walocaust.[5]

On March 6, 2006, Smith filed this action, seeking a declaratory judgment of his right to sell his Walocaust merchandise and demanding costs and attorneys' fees.

Smith also posted additional content on www.walocaust.com explaining his Walocaust theme and how he came up with the word "Walocaust." He added links to other anti-Wal-Mart websites and discussion groups and provided a link to a new page that allowed viewers to download printable copies of the censored Walocaust graphics for free so that they could print their own t-shirts or bumper stickers.

After learning that some courts of appeals had approved disclaimers as a technique for minimizing possible trademark confusion, Smith added one to the top of his Walocaust webpage, stating that the site is unaffiliated with Wal-Mart and containing, the URL for Wal-Mart's official website to help redirect any visitors who may have intended to visit www.wal-mart. com but instead accessed the Walocaust site by mistake. He also updated the site to denounce Wal-Mart's role in forcing this litigation and filing counterclaims, and he posted A link to an entity called "Public Citizen" through which visitors have donated $1040.01 in support of his legal activities.

On or about March 8, 2006, after filing his declaratory judgment complaint, Smith also registered the domain names www. wal-qaeda.com and www.walqaeda.com.[6] "Wal-Qaeda" was another portmanteau word Smith coined, this time combining the name "Wal-Mart" with "Al-Qaeda."[7] Smith intended the word "Wal-Qaeda" as a comment on what he considered to be Wal-Mart's terrorist-like attack on his free speech through threats of litigation.

On a new site that was accessible via both www.wal-qaeda.com and www. walqaeda.com, Smith displayed various graphics incorporating his new word. He also posted other anti-Wal-Mart slogans such as "FREEDOM-HATER-MART STOP Stomping on our free speech!" and "Freedom-Haters ALWAYS," intended to call to mind Wal-Mart's trademark "ALWAYS LOW PRICES. ALWAYS."

Once he became certain that CafePress was open to carrying his new Wal-Qaeda concepts, he created a new Wal-Qaeda CafePress webpage where he again offered various items commenting on Wal-Mart.[8] All of the products offered on 1312 Smith's new Wal-Qaeda CafePress webpage incorporated the word "Wal-Qaeda" except for three concepts: one imprinted with a graphic depicting a shoe hovering over a yellow unhappy face with the legend "FREEDOM-HATER-MART STOP Stomping on our free speech!"; another with the legend "BENTON★VILLEBULIES ALWAYS"; and a third reading simply, "FREEDOM HATER MART."

The site offered two text-only designs that depicted the word "WAL-QAEDA" in a blue block letter font similar to Wal-Mart's: one with the legend "SUPPORT OUR TROOPS[.] BOYCOTT WAL-QADA" and another reading, "WAL-QADA[.] Freedom Haters ALWAYS."

The site also offered products imprinted with five other graphical concepts. One of those concepts was a revision of the Walocaust storefront design, altered to replace "WAL★OCAUST" with "WAL-QAEDA[.] THE DIME STORE FROM HELL"; to replace the Nazi eagle with "FREEDOM HATERS ALWAYS" and "2 days without a k[n]ife fight"; and, to make other small changes. Another concept depicted an American flag in the shape of a United States map with the word "DECEASED" stamped over it. Above the flag was printed "WAL-QAEDA[.] THE DIME STORE FROM HELL," and under the flag appeared the phrase "CAUSE OF DEATH: A Dime Store." In the third concept, the slogan "ATTENTION WAL★QAEDA[.] THESE COLORS DON'T RUN" was imprinted over a modified American flag, and in the last two concepts, Hillary Clinton was named the "WAL-QAEDA Employee of the Year 1986-1992," and Chairman Mao Zedong was awarded the "WAL-QAEDA Human Resource Achievement Award."[9]

Although he hoped to help finance this lawsuit with the proceeds,[10] Smith did not actively market his designs. He did, however, post his new Wal-Qaeda home page, his Wal-Qaeda CafePress account and a link to the Wal-Qaeda home page from his Walocaust website at a time when he knew that reporters were working on stories about this litigation. As a result, news about his new Wal-Qaeda designs was reported in the press and on blogs, and almost all of the sales of Smith's Wal-Qaeda items occurred within a month of the first publicity that followed upon the press and bloggers discovering those designs, The revenues from Smith's CafePress Walocaust and Wal-Qaeda account sales have been less than his costs for the domain names and CafePress account fees.[11]

On April 28, 2006, Wal-Mart filed its answer and counterclaim, asserting various 1313 federal trademark claims and related state law claims against Smith for both the Walocaust and the Wal-Qaeda products. Wal-Mart contends that Smith has engaged in (1) trademark infringement in violation of 15 U.S.C. § 1114(1) and common law; (2) unfair competition in violation of 15 U.S.C. § 1125(a); (3) trademark dilution by tarnishment in violation of 15 U.S.C. § 1125(c); and (4) cybersquatting in violation of 15 U.S.C. § 1125(d). Wal-Mart also brings state law claims for (1) common law trademark infringement; (2) unfair competition in violation of O.C.G.A. § 23-2-55 and common law; (3) deceptive trade practices in violation of O.C.G.A § 10-1-370 et seq.; and (4) trademark dilution and injury to business reputation in violation of O.C.G.A. § 10-1-451(b). Wal-Mart also claims costs and attorneys' fees under 15 U.S.C. § 1117.

II. Analysis

Wal-Mart contends that Smith is a merchant who misappropriated its trademarks and business reputation in pursuit of illegal profit and who disingenuously seeks to cloak those activities under the First Amendment. Smith alleges that Wal-Mart is attempting to misuse trademark laws to censor his criticism of the company. According to Smith, at stake in this case is a person's right to publicly criticize the world's largest retailer — or any other business.

Wal-Mart moves for summary judgment on its claims of federal and state law trademark infringement, federal and state law unfair competition, federal law trademark dilution by tarnishment, state law deceptive trade practices, state law trademark dilution and injury to business reputation, and state law unfair competition. Smith moves the Court for summary judgment on his claim for a declaratory judgment that he has not violated any of Wal-Mart's trademark rights and a dismissal of Wal-Mart's claims with prejudice.

A. Summary Judgment Legal Standard

Summary judgment is proper when no genuine issue as to any material fact is present and the moving party is entitled to judgment as a matter of law. Fed. R.Civ.P. 56(c). The movant carries the initial burden and must show that there is "an absence of evidence to support the nonmoving party's case." Celotex Corp. v. Catrett, 477 U.S. 317, 325, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). "Only when that burden has been met does the burden shift to the non-moving party to demonstrate that there is indeed a material issue of fact that precludes summary judgment." Clark v. Coats & Clark, Inc., 929 F.2d 604, 608 (11th Cir.1991). The nonmovant is then required to "go beyond the pleadings" and present competent evidence in the form of affidavits, depositions, admissions and the like, designating "specific facts showing that there is a genuine issue for trial." Celotex, 477 U.S. at 324, 106 S.Ct. 2548. "The mere existence of a scintilla of evidence" supporting the nonmovant's case is insufficient to defeat a motion for summary judgment. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 252, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). Resolving all doubts in favor of the nonmoving party, the Court must determine "whether a fair-minded jury could return a verdict for the [nonmoving party] on the evidence presented." Id.

B. Threshold Issue: Trademark Ownership

Wal-Mart contends that Smith infringed its registered trademarks, "WALMART," 1314 "WAL-MART," and "WAL★MART," its registered word mark "ALWAYS LOW PRICES. ALWAYS," and the "well-known smiley face mark" to which Wal-Mart contends it has common law trademark rights. Before Wal-Mart may prevail on any of its infringement claims, it must establish that it in fact owns valid trademarks that Smith used in commerce without Wal-Mart's consent. See 15 U.S.C. 1125(c) & (d) (entitling only the owner of a mark to bring a claim for dilution by tarnishment or for cybersquatting); Lone Star Steakhouse & Saloon, Inc. v. Alpha of Virginia, Inc., 43 F.3d 922, 930 (4th Cir.1995) (in order to prevail on a trademark infringement or unfair competition claim, "a complainant must demonstrate that it has a valid, protectible trademark...."); Univ. of Ga. Athletic Ass'n v. Laite, 756 F.2d 1535, 1539 n. 11 (11th Cir.1985) (noting that the test for deceptive trade practice and unfair competition under Georgia law is the same at the test for false designation of origin under the Lanham Act).

It is undisputed that Wal-Mart's registration and marketplace usage of its trademarks "WALMART," "WAL-MART," and "WAL★MART" and its registered word mark "ALWAYS LOW PRICES. ALWAYS." establish both the company's ownership of the marks and its priority over secondary uses. It is similarly undisputed that Smith used the marks in commerce without Wal-Mart's consent.[12]

The parties, however, fervently dispute whether Wal-Mart has established common-law trademark rights to the yellow smiley face.[13]

Trademark protection is available only to "distinctive" marks — those that serve to identify the source of goods or services. Welding Servs., Inc. v. Forman, 509 F.3d 1351, 1357 (11th Cir.2007). A mark that is not inherently distinctive[14] 1315 may acquire distinctiveness or secondary meaning by "becoming associated in the minds of the public with the products or services offered by the proprietor of the mark." Id. A mark has acquired secondary meaning when the primary significance of the term in the minds of the consuming public is the producer. Id. at 1358. The Eleventh Circuit has established that

Whether a [mark] has attained secondary meaning depends on the length and nature of the [mark's] use, the nature and extent of advertising and promotion of the [mark], the efforts of the proprietor to promote a conscious connection between the [mark] and the business, and the degree of actual recognition by the public that the [mark] designates the proprietor's product or service.

Id.

Wal-Mart's only offer of proof supporting its contention that it has infused the smiley face with secondary meaning is a conclusory affidavit from a senior marketing manager who has been with the company since November 2006. The marketing manager states that "Wal-Mart has for many years used its smiley face in a yellow circle in conjunction with its registered marks" and that he counts it among the Wal-Mart marks in which "Wal-Mart has invested substantial resources and years of effort in developing" and which "have become synonymous with Wal-Mart." He also states that the smiley face is used with the other Wal-Mart marks, but he offers no description or depiction of these claimed uses.

Because the proffered testimony provides no specific facts regarding the length and nature of the smiley face's use, the nature and extent of advertising and promotion of the smiley face, Wal-Mart's efforts to promote a connection between the smiley face and its business, or the degree of actual recognition by the public that the smiley face designates Wal-Mart's products or services, Wal-Mart has failed to establish that the smiley face has acquired secondary meaning or that it is otherwise a protectible trademark. See Celotex, 477 U.S. at 324, 106 S.Ct. 2548 (requiring "specific facts showing that there is a genuine issue for trial"); Anderson, 477 U.S. at 252, 106 S.Ct. 2505 ("The mere existence of a scintilla of evidence" supporting the nonmovant's case is insufficient to defeat a motion for summary judgment.). As a result, Wal-Mart's claim that Smith infringed its trademark rights in the smiley face icon fails as well.

Therefore, the Court GRANTS summary judgment to Smith with regard to all of Wal-Mart's claims that pertain to the smiley face icon. Accordingly, the Court will analyze the remaining claims only as they relate to the registered Wal-Mart marks at issue in this case.

C. Trademark Infringement, Unfair Competition, Cybersquatting and Deceptive Trade Practices Claims

To prove that Smith committed trademark infringement or cybersquatting, or subjected Wal-Mart to unfair competition or deceptive trade practices, Wal-Mart must also show that Smith's use of its trademarks is likely to cause an appreciable number of potential buyers to be confused about the source, affiliation or sponsorship of Smith's products. See 15 U.S.C. § 1125(d)(1)(A) (subjecting to a cybersquatting claim only domain names that are "identical or confusingly similar" to a senior mark); Babbit Elecs., Inc. v. Dynascan Corp., 38 F.3d 1161, 1178 (11th 1316 Cir.1994) (citing the 15 U.S.C. § 1114 trademark infringement statute); Lone Star Steakhouse, 43 F.3d at 930 (applying the same standard to an unfair competition claim); Looney v. M-Squared, Inc., 262 Ga.App. 499, 505, 586 S.E.2d 44, 50 (2003) (citing O.C.G.A. § 10-1-372(a)).

In making this inquiry, courts consider a variety of factors, including the strength of the allegedly infringed mark, whether the designs that incorporate the registered mark are similar, whether the products sold by the parties are similar, whether the retail outlets and purchasers are similar, whether the parties use the same advertising media, whether the defendant intended to usurp the registered trademark, and whether any consumers were actually confused. Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., 675 F.2d 1160, 1164 (11th Cir.1982). The Court must balance the factors according to its own judgment based oh the facts in the case before it. Custom. Mfg. & Eng'g, Inc. v. Midway Servs., Inc., 508 F.3d 641, 649-50 (11th Cir.2007) (explaining that a district court is not simply to compute the percentage of factors that weigh in favor of likelihood of confusion to determine whether such a likelihood exists).

Because Smith's arguments with regard to the Safeway factors depend heavily on whether his designs are successful parodies, the Court must first consider whether the contested designs are in fact parodies of Wal-Mart's registered marks. See Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394, 1405 (9th Cir.1997) (noting that the claim that a secondary use is a parody is not a separate defense to a charge of trademark infringement but is instead is considered within the likelihood of confusion analysis); see also Connick v. Myers, 461 U.S. 138, 148 n. 7, 103 S.Ct. 1684, 75 L.Ed.2d 708 (1983) ("The inquiry into the protected status of speech is one of law, not fact."). For the purposes of trademark analysis, "a parody is defined as a simple form of entertainment conveyed by juxtaposing the irreverent representation of the trademark with the idealized image created by the mark's owner." Louis Vuitton Malletier v. Haute Diggity Dog, LLC, 507 F.3d 252, 260 (4th Cir.2007). To be considered successful, the alleged parody must both call to mind and differentiate itself from the original, and it must "communicate some articulable element of satire, ridicule, joking or amusement." Id.

When applying these criteria to the facts of the case, it is clear that Smith's concepts are parodies of the registered Wal-Mart marks. Smith successfully calls Wal-Mart to mind by using either "WAL" or "MART" as part of the concept; by mimicking its fonts and storefront design; by mentioning Bentonville, the location of Wal-Mart's headquarters; or by including various other icons typically associated with Wal-Mart. As Wal-Mart fervently contends, it is obvious that Smith's concepts use Wal-Mart imagery to evoke the company in the mind of his viewers.

It is equally obvious that Smith's concepts are not the "idealized image" of the registered Wal-Mart marks. "Walocaust," "Wal-Qaeda" and "Freedom-Hater-Mart" are not "Wal-Mart." The imagery on Smith's t-shirts includes portraits of Mao Zedong, a United States map with the word "DECEASED" stamped over it, and the slogan "FREEDOM HATERS ALWAYS."

Finally, the juxtaposition of the similar and dissimilar — the satirical representation and the idealized image of Wal-Mart 1317 — conveys a scathing parody. In the "smiley eagle" Walocaust concept, the reference to the Holocaust and the image of the Nazi eagle clutching a smiley face at once portrays and contradicts the benign image that Wal-Mart portrays to the community. In the "SUPPORT OUR TROOPS" Wal-Qaeda concept, Smith transforms all-American "Wal-Mart" into the terrorist group "Wal-Qaeda" and satirically urges the viewer to support Wal-Qaeda's troops, apparently commenting both on what Smith considers to be Wal-Mart's ruthless business tactics and its detrimental impact on the United States. Other concepts juxtapose Wal-Mart's reputation for low prices with a reference to poor store security and the company's family values imagery with the fact that it offers for sale inexpensive alcohol, tobacco and firearms — products known better for destroying families.

The Court thus concludes that Smith's concepts adequately evoke Wal-Mart while maintaining their differentiation, and they convey Smith's satirical commentary; thus, they are successful parodies. See Louis Vuitton, 507 F.3d at 261.

The finding that Smith's concepts are parodies does not preclude the likelihood of confusion analysis, however; it merely influences the way the likelihood of confusion factors are applied. Id. "[A]n effective parody will actually diminish the likelihood of confusion, while an ineffective parody does not." Id. Because even a parody may constitute trademark infringement if that parody is confusing, the Court will next consider the likelihood of confusion factors. Id.

1. Actual Confusion

Proof of actual confusion is considered the best evidence of likelihood of confusion. Roto-Rooter Corp. v. O'Neal, 513 F.2d 44, 45-46 (5th Cir.1975). A claimant may present anecdotal evidence of marketplace confusion, and surveys, when appropriately and accurately conducted and reported, are also widely and routinely accepted as probative of actual confusion. See, e.g., AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1544 (11th Cir.1986) (considering the proffered survey but giving it little weight); SunAmerica Corp. v. Sun Life Assurance Co. of Canada, 890 F.Supp. 1559, 1576 (N.D.Ga.1994) (viewing the proffered survey as confirmation of consistent anecdotal evidence).

Wal-Mart concedes that it has no marketplace evidence of actual consumer confusion. Instead, it presents two consumer research studies conducted by Dr. Jacob Jacoby that purport to prove that consumer confusion and damage to Wal-Mart's reputation are likely.

a. The Jacoby Report

Jacoby developed two surveys for Wal-Mart that both purported to measure consumer confusion and dilution by tarnishment. Specifically, the stated objectives of the research were (1) "To determine whether (and if so, to what extent), when confronted with merchandise bearing Mr. Smith's designs either in person or via the Internet, prospective consumers would be confused into believing that these items either came from Wal-Mart, came froth a firm affiliated with Wal-Mart, or had been authorized by Wal-Mart," and (2) "To determine whether (and if so, to what extent) exposure to Mr. Smith's designs would generate dilution via tarnishment."

Deeming it impractical to test all of Smith's designs, Jacoby chose instead to test two products as representative of all 1318 of Smith's allegedly infringing products — the white t-shirt with the word "WAL★OCAUST" in blue font over the Nazi eagle clutching a yellow smiley face, and another white t-shirt that depicted the word "WAL-QAEDA" in a blue font as part of the phrase "SUPPORT OUR TROOPS. BOYCOTT WAL-QAEDA."

He also tested consumer reactions to "control" designs, which he compared to consumer responses to the Walocaust and Wal-Qaeda designs. To develop the control for the Walocaust design, Jacoby replaced the star with a hyphen and removed the smiley face from the yellow circle, and for both the Walocaust and Wal-Qaeda controls, he substituted "Z" for "W." These substitutions resulted in control concepts entitled "Zal-ocaust" and "Zal-Qaeda."

Jacoby engaged a market research firm to test each of the t-shirt designs in (1) a "product" study intended to test for postpurchase confusion and tarnishment, and (2) a "website" study intended to test for point-of-sale confusion and tarnishment.[15]

The market research company conducted the studies in a mall-intercept format. The company's researchers would approach people who appeared to be thirteen years old or older and ask a series of screening questions.[16] To qualify for either survey, the respondent was required to be at least thirteen years old[17] and must have in the past year bought, or would in the coming year consider buying, bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them. To qualify for the "website" study, the respondent must also have (1) used the Internet in the past month to search for information about products or services and (2) either (a) in the past year used the Internet to buy or to search for information about bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them, or (b) in the coming year would consider buying over the Internet bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them.[18] If the respondent met the qualifications, he or she was asked to go with the researcher to the mall's enclosed interviewing facility for a five-minute interview.[19]

For the "product" study, the interviewers presented to each respondent one of 1319 the four t-shirts described above and asked the respondent to imagine seeing someone wearing the shirt. The interviewer then asked a series of questions.

The first three sets of questions were designed to test for consumer confusion, The interviewers were directed to ask each of the "likelihood of confusion" questions sequentially unless the respondent answered "Sears," "Wal-Mart," "Youngblood's" or "K-Mart," in which case the interviewer was to record the answer, skip the remaining confusion questions, and go directly to the tarnishment questions.

In the consumer confusion series, the first set of questions tested for confusion as to source. The interviewer would ask "which company or store" the respondent thought "put out" the shirt, and if the respondent named a company or store, the interviewer then asked what about the shirt made the respondent think the shirt was "put out" by that company or store. The second set of questions, which dealt with confusion as to connection or relation ship, asked the respondent whether the company or store that "put out" the shirt had some "business connection or relationship with another company" and if so, with what company. The respondent was then asked why he or she believed the companies had a business connection or relationship. A third set of questions, aimed at testing for confusion as to authorization or sponsorship, asked whether the company that "put out" the shirt needed permission from another company to do so, and if so, which company.

Finally, if the respondent had not yet answered "Sears," "Wal-Mart," "Youngblood's" or "K-Mart" to any of the first three sets of questions, he or she was then asked what the shirt made him or her "think of and then "which company or store" the shirt brought to mind.

The fifth set of questions, which tested for dilution by tarnishment, were asked in reference to any company or store the respondent mentioned in his or her answers to the first four sets of questions. The first question asked whether seeing the shirt made the respondent more or less likely to shop at the store he or she had named, and the second question asked whether the perceived association with the store made the respondent more or less likely to buy the shirt.

The interviews for the website study were much like those for the product study, except that instead of being shown the actual shirts, the respondents were exposed to a simulation of Smith's Walocaust CafePress homepage, his Wal-Qaeda CafePress homepage or the associated control homepage.[20] In each of the simulations, all of the hyperlinks were removed from the homepages except for the one hyperlink associated with the t-shirt that Jacoby had decided to test.

Jacoby directed the interviewers to begin each website interview by providing a URL to the respondent and asking the respondent to imagine that the URL was a search term the respondent had heard or seen somewhere and wanted to look up on the Internet. The interviewer would then have the respondent sit at a computer and type the URL into the browser. The URL would take the respondent to the simulated home page for testing.

1320 The interviewer would then direct the respondent to look at the screen and scroll down the page "as [he or she] normally would" and click through to the first t-shirt on the screen. The respondent was then directed to click on the "view larger" box and look at the shirt as though he or she "found it interesting and [was] considering whether or not to order it...." The interviewer would then ask the respondent exactly the same series of questions posed in the product study, including the same skip pattern to be applied in the event that the respondent mentioned Sears, Wal-Mart, Youngblood's or K-Mart in response to any of the consumer confusion questions.

In order to be tallied as "confused," the respondent had to meet two tests. First, the respondent had to indicate either that the shirt came from Wal-Mart (first confusion series), came from a company that had some business connection or relationship with Wal-Mart (second confusion series), or came from a source that required or obtained permission from Wal-Mart (third confusion series). Second, the respondent had to indicate that his or her reason for that understanding was either because of the prefix "Wal," the name (or equivalent), the smiley face, or the star after the prefix "Wal." Thus, a respondent who believed that there was a connection between Wal-Mart and the t-shirt that he or she was shown but who did not mention the prefix "Wal," the name (or equivalent), the smiley face, or the star, would not be counted as "confused."

Any respondent who perceived an association between Wal-Mart and the t-shirt that he or she was shown and reported that the perceived association either made the respondent less likely to shop at Wal-Mart or more likely to buy that t-shirt was deemed to satisfy the requirement for dilution.

The field interviewers returned 322 completed interviews for the product study and 335 for the website study. Three responses were eliminated from the sample after the research company conducted a review to ensure that each respondent was qualified to participate in the study and that the questionnaires had been completed properly. The research company then sent the name and phone number of each of the interview respondents to an independent telephone interviewing service for validation, which consisted of calling each mall-intercept respondent to ensure that the respondent had actually participated in the study and that his or her answers were accurately recorded.

In the product study, 181 respondents (fifty-six percent of the usable sample) were positively validated, and sixteen respondents (about five percent) reported either different answers to the survey questions or claimed not to have participated in the study. The remainder either could not be reached during the twenty days Jacoby allocated for the validation or refused to respond to the validation survey.

Jacoby reported the results of those respondents who were positively validated plus the results from the respondents who could not be reached or would not respond to the validation survey, and he eliminated the results of the respondents who provided non-affirming answers during the validation process. This resulted in 305 reported responses to the product study: seventy-three for the Wal★ocaust concept, seventy-six for the Wal-Qaeda concept, seventy-nine for the Zal-ocaust concept, and seventy-seven for the Zal-Qaeda concept.

1321 In the website study, 169 respondents (fifty-one percent of the usable sample) were positively validated, and forty-six respondents (about fourteen percent) reported either different answers to the survey questions or claimed not to have participated in the study. The remainder either could not be reached during the twenty days Jacoby allocated for the validation or refused to respond to the validation survey.

As he did in the product study, Jacoby reported the results of those respondents who were positively validated plus the results from the respondents who could not be reached or would not respond to the validation survey, and he eliminated the results of the respondents who provided non-affirming answers during the validation process. This resulted in 287 reported responses to the product study: seventy for the Wal★ocaust concept, seventy-eight for the Wal-Qaeda concept, sixty-nine for the Zal-ocaust concept, and seventy for the Zal-Qaeda concept.

Jacoby reported that the survey reflected high levels of consumer confusion and dilution by tarnishment. He claimed that the post-purchase confusion "product study" indicated a likelihood of confusion in nearly forty-eight percent of the respondents and that the point-of-sale confusion "website" study indicated a likelihood of confusion in almost forty-one percent of the respondents.[21] Jacoby also claimed that the "dilution" study indicated that almost twelve percent of the respondents were less likely to shop at Wal-Mart after seeing Smith's designs.

b. Evidentiary Objections

Smith moves to exclude Wal-Mart's expert report [78]. He claims that Jacoby did not have the requisite Internet expertise to conduct the web-based "point-of-sale" portion of this particular study and that several aspects of Jacoby's methodology affecting both portions of the study were faulty; thus, he contends, Jacoby's study is "too deeply flawed to be considered...."

Wal-Mart argues that the Jacoby test was performed by a competent expert according to industry standards and therefore is valid. Wal-Mart further contends that the expert witnesses Smith presents in rebuttal are not experts in the area of consumer-goods "likelihood of confusion" trademark studies, and therefore their testimony is irrelevant and should be excluded [81, 82].

Whether a given survey constitutes acceptable evidence depends on the survey's ability to satisfy the demands of Federal Rule of Evidence 703, which requires consideration of the "validity of the techniques employed." 233-34 FED. JUD. CTR., REFERENCE MANUAL ON SCI. EVIDENCE (2d ed.2002) (explaining that in the context of surveys for litigation purposes, "[t]he inquiry under Rule 703[, which] focuses on whether facts or data are `of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject' ... becomes, `Was the survey conducted in accordance with generally accepted survey principles, and were the results used in a statistically correct way?'"). See also BFI Waste Sys. of N. Am v. Dekalb County, 303 F.Supp.2d 1335,1346 (N.D.Ga.2004) (noting that the opposing party could have challenged an expert witness's reference to a recent survey by questioning whether the survey methodology satisfied Rule 703).

1322 The Eleventh Circuit has held that alleged technical deficiencies in a survey presented in a Lanham Act action affect the weight to be accorded to the survey and not its admissibility. Jellibeans, Inc. v. Skating Clubs of Ga., Inc., 716 F.2d 833, 844 (11th Cir.1983). Other courts have held that a significantly flawed survey may be excludable as evidence under either Rule 403 (the rule barring evidence that is more prejudicial than probative) or Rule 702 (the rule barring unreliable expert testimony). Citizens Fin. Group, Inc. v. Citizens Nat'l Bank, 383 F.3d 110, 188-21 (3d Cir.2004) (finding that the district court properly excluded survey evidence under Rules 702 and 403 where the survey contained flaws that were not merely technical, but were so damaging to the reliability of the results as to be "fatal": the survey relied on an improper universe and its questions were imprecise); Malletier v. Dooney & Bourke, Inc., 525 F.Supp.2d 558, 562-63 (S.D.N.Y.2007). Even when a party presents an admissible survey purporting to show consumer confusion, however, the survey "does not itself create a triable issue of fact." Mattel, Inc. v. MCA Records, Inc., 28 F.Supp.2d 1120, 1133 (C.D.Cal.1998) (citing Universal City Studios, Inc. v. Nintendo Co., 746 F.2d 112, 118 (2d Cir.1984), which found a survey "so badly flawed that it cannot be used to demonstrate the existence of a question of fact of the likelihood of consumer confusion"). Accord Leelanau Wine Cellars, Ltd. v. Black & Red, Inc., 502 F.3d 504, 518 (6th Cir.2007); Scott Fetzer Co. v. House of Vacuums, Inc., 381 F.3d 477, 488 (5th Cir.2004) (holding that a court may disregard survey evidence if the, survey contains such serious flaws that any reliance on its results would be unreasonable).

To ground a survey as trustworthy, its proponent must establish foundation evidence showing that

(1) the `universe' was properly defined, (2) a representative sample of that universe was selected, (3) the questions to be asked of interviewees were framed in a clear, precise and non-leading manner, (4) sound interview procedures were followed by competent interviewers who had no knowledge of the litigation or the purpose for which the survey was conducted, (5) the data gathered was accurately reported, (6) the data was analyzed in accordance with accepted statistical principles and (7) objectivity of the entire process was assured.

Toys R Us, Inc. v. Canarsie Kiddie Shop, 559 F.Supp. 1189, 1205 (D.C.N.Y. 1983) (citing MANUAL FOR COMPLEX LITIG., 116 (5th ed.1981), 4 LOUISELL & MUELLER, FED. EVIDENCE § 472 (1979), and J. THOMAS MCCARTHY, TRADEMARKS & UNFAIR COMPETITION § 32:53 (1973)); accord Rush Indus., Inc. v. Gamier LLC, 496 F.Supp.2d 220, 227 (E.D.N.Y.2007). Failure to satisfy any of the listed criteria may seriously compromise the survey's impact on a court's likelihood of confusion evaluation. Id.

Smith cites several grounds for excluding the Jacoby survey. He argues that the survey is inadmissible because it (1) failed to identify the relevant consumer universe or used a consumer universe that was substantially overbroad; (2) failed to replicate shopping conditions as consumers would encounter them in the marketplace; (3) was improperly leading; (4) violated the survey structure protocol necessary to comply with double-blind standards; and (5) failed to establish a relevant factual basis for Wal-Mart's dilution by tarnishment claims. Smith further argues that even if the Court admits the survey, its consideration should be limited to only the two tested designs, despite Jacoby's claim that they are representative of all the designs Wal-Mart seeks to enjoin.

1323 As an initial matter, the Court observes that Smith does not take issue with Jacoby's qualifications to design and conduct a consumer confusion survey and to analyze its results. It is undisputed that Jacoby is a nationally renowned trademark survey expert who has testified hundreds of times. Smith contends, however, that Jacoby was unqualified to conduct this particular survey because he "lacks knowledge, experience, [and] sophistication" with regard to products marketed exclusively over the Internet and that as a result Jacoby's survey protocol contained significant flaws.

Based upon its own review of Jacoby's education and experience, the Court concludes that Jacoby is qualified to design and conduct a consumer survey and to testify about its results. To the extent that Jacoby's purported lack of experience with surveys concerning goods sold exclusively online may have led him to test the wrong universe or to fail to replicate the shopping experience, as Smith has alleged, these factors will be examined when the Court evaluates the trustworthiness of the survey.

i. Web-Related Challenges

In undertaking to demonstrate likelihood of confusion in a trademark infringement case by use of survey evidence, the "appropriate universe should include a fair sampling of those purchasers most likely to partake of the alleged infringer's goods or services." Amstar Corp. v. Domino's Pizza, Inc., 615 F.2d 252, 264 (5th Cir.1980). Selection of the proper universe is one of the most important factors in assessing the validity of a survey and the weight that it should receive because "the persons interviewed must adequately represent the opinions which are relevant to the litigation." Id. "Selection of a proper universe is so critical that `even if the proper questions are asked in a proper manner, if the wrong persons are asked, the results are likely to be irrelevant.'" Wells Fargo & Co. v. WhenU.com, Inc., 293 F.Supp.2d 734, 767 (E.D.Mich.2003) (quoting 5 MCCARTHY, § 32:159). "A survey must use respondents from the appropriate universe because `there may be systemic differences in the responses given... by persons [with a particular] characteristic or preference and the responses given to those same questions ... by persons who do not have that ... characteristic or preference.'" Id. (quoting FED. EVIDENCE PRACTICE GUIDE (Matthew Bender 2003) § [4][6][i]).

Similarly, "[a] survey that fails to adequately replicate market conditions is entitled to little weight, if any." Leelanau Wine Cellars, Ltd. v. Black & Red, Inc., 452 F.Supp.2d 772, 783 (W.D.Mich.2006), aff'd, 502 F.3d 504 (6th Cir.2007) (quoting Wells Fargo & Co., 293 F.Supp.2d at 766). Although "[n]o survey model is suitable for every case ... a survey to test likelihood of confusion must attempt to replicate the thought processes of consumers encountering the disputed mark or marks as they would in the marketplace." Simon Prop. Group L.P. v. mySimon, Inc., 104 F.Supp.2d 1033, 1038 (S.D.Ind.2000) (citing MCCARTHY ON TRADEMARKS § 32:163 (4th ed.1999) for the principle that "the closer the survey methods mirror the situation in which the ordinary person would encounter the trademark, the greater the evidentiary weight of the survey results").

Smith hired Dr. Alan Jay Rosenblatt as a rebuttal witness to point out Internet-related deficiencies in Jacoby's survey methodology — particularly deficiencies in universe selection and replication of marketplace conditions — that he claims resulted 1324 from Jacoby's erroneous assumptions about how people reach and interact with websites. Smith uses Rosenblatt's expertise on Internet user experience and navigation to support his Daubert argument that because Jacoby surveyed an improperly broad universe and his survey design did not approximate the actual consumer marketplace experience, the Jacoby studies are legally insufficient to prove consumer confusion or trademark dilution. Thus, Smith argues, the studies should be afforded little, if any, evidentiary value.

Coming from an academic background in political science and survey methodology — subjects he taught at the university level for ten years — Rosenblatt is a professional in the area of Internet advocacy (the use of online tools to promote a cause). His experience includes helping organizations bring people to their websites, induce the visitors to read the portion of the website that contains the call to action, and encourage the visitors to take the suggested action. He also helps the organizations track visitor behavior in order to increase website effectiveness.

Wal-Mart moves the Court to exclude Rosenblatt's testimony [82], contending that because he is not an expert in the area of consumer-goods "likelihood of confusion" trademark studies, he is unqualified to comment on Jacoby's studies and therefore his testimony is irrelevant. It also argues that the portion of Rosenblatt's report that is devoted to a general analysis of people's Internet browsing habits rather than specific criticisms of Jacoby's report is improper rebuttal and must be excluded. Wal-Mart also criticizes Rosenblatt for basing his evaluation of Jacoby's studies on "conventional wisdom in the industry" and Rosenblatt's personal experience on other website projects rather than on tests or experiments it suggests Rosenblatt should have conducted on Smith's actual websites.

Smith contends that Rosenblatt is offered solely as a rebuttal expert whose function is not to testify regarding consumer-goods "likelihood of confusion" trademark studies but rather to question a discrete underlying issue on which he is well-qualified to testify: Jacoby's assumptions about how Internet users interact with websites and how they search for content online.

Rule 702 of the Federal Rules of Evidence provides that a "witness qualified as an expert by knowledge, skill, experience, training, or education" may testify when specialized knowledge will help the factfinder determine a fact in issue. Fed.R.Evid. 702. Expert testimony concerning specialized knowledge is admissible to assist the trier of fact if "(1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case." Fed.R.Evid. 702.

It is true that Rosenblatt has no experience evaluating the merits of trademark infringement or dilution claims and that only one of the surveys he has designed involved a consumer product. The Court finds, however, that his extensive experience studying Internet user behavior and designing social science surveys qualifies him to provide testimony about (1) how Internet users interact with websites and how they search for content online, (2) whether Jacoby's survey methodology comported with those tendencies, and (3) how Jacoby's assumptions about Internet user behavior impacted the accuracy of the surveyed universe and the survey's replication of the online shopping experience. 1325 The Court finds Rosenblatt's testimony evaluating Jacoby's survey protocol to be both relevant and, because it is based on Rosenblatt's undisputed area of expertise, reliable.[22] Therefore, to the extent that Rosenblatt's testimony focuses on those issues, Wal-Mart's motion to exclude it [82] is DENIED.[23]

(a) Survey Universe

The appropriate universe in this case is the consumers most likely to purchase Smith's Walocaust and Wal-Qaeda merchandise. See Amstar Corp., 615 F.2d at 264. To qualify for either of Jacoby's surveys, the respondent had to be over thirteen years of age and had to have in the past year bought, or would in the coming year consider buying, bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them. To qualify for the web-based point-of-purchase study, the respondent must also have (1) used the Internet in the past month to search for information about products or services and (2) either (a) in the past year used the Internet to buy or search for information about bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them, or (b) in the coming year would consider buying over the Internet bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them. Wal-Mart maintains that Jacoby's universe selection was proper. Smith counters that it was overly broad.

Although the universe Jacoby selected would include purchasers of Smith's Walocaust or Wal-Qaeda merchandise, the Court finds that it is significantly overbroad. Because Smith's merchandise was available only through his CafePress webstores and the links to his CafePress webstores from his Walocaust and Wal-Qaeda websites, it is likely that only a small percentage of the consumers in the universe selected by Jacoby would be potential purchasers of Smith's products. A survey respondent who purchases bumper stickers, t-shirts or coffee mugs with words, symbols or designs on them may buy such merchandise because the imprint represents his or her school, company, favorite sports team, cartoon character, social group, or any of hundreds of other interests or affiliations; he or she may have no interest at all in purchasing merchandise containing messages about Wal-Mart, pro or con. The respondent may buy from brick-and-mortar stores or well-known retailers with Internet storefronts Without being aware of Smith's website or CafePress, or may have little interest in buying such merchandise over the Internet at all. Therefore, a respondent who clearly falls within Jacoby's survey universe 1326 may nevertheless have no potential to purchase Smith's imprinted products.[24] See Leelanau Wine Cellars, 452 F.Supp.2d at 782.

Other courts have similarly criticized surveys — including surveys Jacoby conducted in other trademark infringement cases — that failed to properly screen the universe to ensure that it was limited to respondents who were potential purchasers of the alleged infringer's product.

For example, in Weight Watchers Int'l, Inc. v. Stouffer Corp., 744 F.Supp. 1259 (S.D.N.Y.1990), Weight Watchers sued Stouffer for trademark infringement after Stouffer launched an advertising campaign that suggested that new exchange listings on Stouffer's Lean Cuisine packages would allow adherents to the Weight Watchers program to use Lean Cuisine entrees in their diets. Id. at 1262. Stouffer's likelihood of confusion survey, also conducted by Jacoby, identified the universe as "women between the ages of 18 and 55 who have purchased frozen food entrees in the past six months and who have tried to lose weight through diet and/or exercise in the past year." Id. at 1272. The court found that the universe was overbroad because the screener had not limited it to dieters, but also had included respondents who may have tried to lose weight by exercise only. The court concluded that as a result the survey likely included respondents who were not potential consumers, and because "[r]espondents who are not potential consumers may well be less likely to be aware of and to make relevant distinctions when reading ads than those who are potential consumers," that portion of the survey universe may have failed to make "crucial" distinctions in the likelihood of confusion testing. Id. at 1273.

Similarly, in Leelanau Wine Cellars, 452 F.Supp.2d 772, the court found that the universe in a survey designed to show a likelihood of confusion between a wine producer's wines and a competitor's wines was overbroad. The junior mark user's product, like Smith's, was distributed through limited channels; the challenged wines were sold only through the junior user's tasting room and website, while the senior mark holder sold its wines through mass retail channels. The survey expert defined 1327 the universe as Michigan consumers over twenty-one years of age who had either purchased a bottle of wine in the five-to-fourteen dollar price range in the last three months or who expected to purchase a bottle of wine in that price range in the next three months. The court held that a purchaser of a wine in that price range would, in general, be a potential consumer of the competitor's wine only if the purchaser planned to buy from some winery's tasting room or website and that the survey universe therefore was overbroad and entitled to little weight.

(b) Shopping Experience

To be valid for the purposes of demonstrating actual confusion in a trademark infringement suit, it is necessary for a survey's protocol to take into account marketplace conditions and typical consumer behavior so that the survey may as accurately as possible measure the relevant "thought processes of consumers encountering the disputed mark' ... as they would in the marketplace." Simon Prop. Group, 104 F.Supp.2d at 1038; accord WE Media, Inc. v. Gen. Elec. Co., 218 F.Supp.2d 463, 474 (S.D.N.Y.2002).

Smith contends that Jacoby's point-of-purchase' study, which purported to measure consumer confusion over merchandise that Smith sold exclusively online, was improperly designed because it failed to take into account typical consumer Internet behavior. Wal-Mart does not contradict the expert testimony Smith proffers regarding consumer Internet behavior but instead maintains that it is irrelevant.

Jacoby's point-of-purchase survey called for interviewers to provide each respondent with specific "search terms" that would take the respondent to a simulation of one of Smith's websites. The respondent was asked to pretend that the resulting web page was of interest and to act accordingly (looking at the page and scrolling through it as the respondent would "normally" do), and then was directed to scroll down the page, below the first screen, and click on a specific t-shirt link. The respondent was not asked what message he or she took from the website or whether the website was in fact of interest. The survey protocol also gave the respondent no choice but to scroll down to the next screen and click on the t-shirt link, the only live link in the simulation.

In presenting Smith's website and directing the survey respondents to click on one specific t-shirt link, Jacoby's survey design presumed that all consumers who might be interested in a printed t-shirt, mug or bumper sticker would be equally likely to happen across Smith's designs, regardless of the respondent's level of interest in the messages on Smith's webpage.

Although, as Wal-Mart points out, it is possible that `some consumers may view web pages randomly and may scroll through and clink on links on pages that are not of interest to them, the Court finds that the survey protocol did not sufficiently reflect actual marketplace conditions or typical consumer shopping behavior and therefore was unlikely to have elicited a shopping mindset that would have allowed Jacoby to accurately gauge actual consumer confusion.

Because Smith's merchandise was available only through his CafePress webstores and the links to his CafePress webstores from his Walocaust and Wal-Qaeda websites, it is unlikely that many consumers randomly happen across Smith's products. According to Rosenblatt's uncontroverted 1328 testimony, people do not come to websites randomly, and they do not move within websites randomly. A great majority of Internet users arrive at a particular website after searching specific terms via an Internet search engine or by following links from another website. The user makes a judgment based on contextual cues — what is shown about a prospective website from the text of a search result or what is said about a prospective website in the hyperlinked words and surrounding text of the website currently being viewed — in determining where to surf next. He moves from website to website, he moves within websites, and he performs actions such as signing a petition — or buying a product — by making choices based on what he sees and whether what he sees leads him to believe that going to the next page or following a link to another website will bring him to something he is interested in seeing, doing or buying.

In the marketplace, the visitor would be presented with a screen full of Smith's anti-Wal-Mart messages. Consumers who were interested in the messages on Smith's web pages would be motivated to choose the links that would eventually lead to his products, while those who were uninterested in Smith's messages would simply leave the page. Because the survey protocol directed the respondents to "pretend" to be interested in Smith's anti-Wal-Mart homepages and then directed them to click on a specific link, there is no assurance that the respondent actually read the homepage or would have been interested enough in it to be motivated to click on the t-shirt link. See Gen. Motors Corp. v. Cadillac Marine & Boat Co., 226 F.Supp. 716, 737 (D.C.Mich.1964) (observing that because survey respondents had little interest the allegedly infringing product, it followed that their inspection of the advertisement shown to them as part of the survey protocol was "casual, cursory and careless" and therefore of little probative value).

Other courts have similarly criticized surveys that failed to adequately replicate the shopping experience. In Gen. Motors Corp., 226 F.Supp. at 737, the court criticized the proffered survey because it did not take into account typical consumer behavior:

Actual purchasers of a boat would not hastily read an advertisement, nor would a potential purchaser read it carelessly. A reasonable man, anticipating the purchase of a boat, would peruse the material at least well enough to note the manufacturer as being "Cadillac Marine & Boat Company, 406 Seventh Street, Cadillac, Michigan." Also, most buyers would want to see the boat itself before making a purchase.

Although the purchase of a t-shirt obviously does not involve the same level of financial consideration a consumer typically makes when buying a boat, a consumer is likely to consider the meaning of an imprinted t-shirt such as Smith's before wearing it in public. A reasonable person who was considering buying a t-shirt that references Al-Qaeda or the Holocaust would likely read the associated webpage at least well enough to see the harsh criticism of Wal-Mart and the prominent disclaimer dispelling any notion of a possible association with the company.

(c) Impact of Internet-Related Flaws on Survey's Evidentiary Value

For all of these reasons, the survey Jacoby conducted for. Wal-Mart is of dubious value as proof of consumer confusion both because its survey universe was over-inclusive 1329 and because its design failed to approximate real-world marketplace conditions. Jacoby's survey is subject to the same criticisms as his Weight Watchers survey and the survey in Leelanau Wine Cellars: Jacoby failed to screen the respondents to ensure that they would likely be aware of and make relevant distinctions concerning the specific product. See Weight Watchers, 744 F.Supp. at 1273; Leelanau Wine Cellars, 452 F.Supp.2d at 783. By failing to approximate actual market conditions, Jacoby further ensured that the survey would not "replicate the thought processes of [likely] consumers [of the junior user's merchandise] encountering the disputed mark ... as they would in the marketplace." See Simon Prop. Group, 104 F.Supp.2d at 1038; accord Gen. Motors Corp., 226 F.Supp. at 737. Therefore, the Court must consider these flaws in determining whether the survey is admissible and, if so, what evidentiary weight to afford it.

ii. Structural Flaws

Smith further alleges that the Jacoby study suffers from several structural flaws that diminish the trustworthiness of the results of both the web-based point-of-sale portion and the post-purchase t-shirt portion of the survey. He contends that (1) both the structure of the survey and the wording of several questions suggested the answers Wal-Mart wanted, and (2) the survey results should not be presumed to represent consumer reaction to any of the challenged merchandise that was not actually tested.

Smith hired Dr. Richard Teach as a rebuttal witness to point out deficiencies in Jacoby's website study survey methodology. Teach is an emeritus marketing professor and former dean at the Georgia Tech School of Business who has designed and conducted over one hundred surveys, including about fifty buyer surveys, and has taught survey methodology, statistics and related courses. Teach testifies that he agrees with Rosenblatt's testimony and also offers criticisms of his own. Smith uses Teach's survey expertise to support his Daubert argument that because the survey protocol contains multiple technical flaws, the results are unreliable and hence should be afforded very light evidentiary value if not completely excluded from evidence.

Wal-Mart moves to exclude Teach's testimony [81], supporting its motion with arguments much like those it used in its motion to exclude Rosenblatt's testimony. It contends that because Teach is not an expert in the area of consumer-goods "likelihood of confusion" trademark studies, he is unqualified to comment on Jacoby's study, and therefore his testimony is irrelevant.[25]

Smith contends that Teach is offered solely as a rebuttal expert whose function is not to testify regarding consumer-goods "likelihood of confusion" trademark studies 1330 but rather to testify as an expert on surveys generally and the extent to which Jacoby's study and report comported with generally accepted techniques, subjects upon which Teach is well qualified to testify.

As it did for Rosenblatt, the Court must evaluate the admissibility of Teach's testimony under the standards provided by Rule 702 of the Federal Rules of Evidence. It is true that Teach has no experience evaluating the merits of trademark infringement or dilution claims. The Court finds, however, that his extensive experience designing and evaluating surveys qualifies him to provide testimony about technical flaws in the design of Jacoby's study and the impact of those flaws on the trustworthiness of Jacoby's reported results.[26]

Indeed, the basic standards for survey research that Jacoby himself cites as authoritative are provided by the Federal Judicial Center's Reference Guide on Survey Research. See generally REFERENCE MANUAL at 229-76. The guide is not specifically devoted to apparel surveys or even trademark surveys; instead, it sets forth considerations by which a court may assess the validity and reliability of surveys generally.

A survey expert can look at Jacoby's procedures and compare them with general standards for survey protocol and explain whether those standards have been met. Similarly, a general survey expert can look at the questions being asked and compare them with the stated objectives that the survey is supposed to be testing and opine as to whether there is a good fit between question and objective, or whether the question may be biased or misleading. Finally, Teach's background in teaching and writing about statistics in the area of marketing qualifies him to discuss whether the testing methodology would lead to results that could be projected to the general population.

Therefore, to the extent that Teach's testimony focuses on general survey methodology, whether Jacoby's survey protocol deviated from standard methodology, and what impact any deviations may have had on the trustworthiness of Jacoby's reported results, Wal-Mart's motion to exclude it [81] is DENIED.

(a) Leading Survey Structure and Questions

Smith argues that both the structure of the survey and the wording of several questions suggested the answers Wal-Mart wanted. Wal-Mart, of course, contends that Jacoby's survey presented no such risk.

(i) Double-Blind Survey Design

To ensure objectivity in the administration of the survey, it is standard practice to conduct survey interviews in such a way as to ensure that "both the interviewer and the respondent are blind to the sponsor of the survey and its purpose." REFERENCE MANUAL at 266. The parties agree that double-blind conditions are essential because if the respondents know what the interviewer wants, they may try to please the interviewer by giving the desired answer, and if the interviewer knows what his employer wants, he may 1331 consciously or unconsciously bias the survey through variations in the wording or the tone of his questions. See id.

Smith argues that the skip pattern included in Jacoby's survey hinted to the interviewers that Wal-Mart was the survey's sponsor. The survey protocol directed the interviewers to skip to the final tarnishment question, question five, if the respondent gave any one of four specific store names — Sears, Wal-Mart, K-Mart or Youngblood's — to any of the first three questions. Similarly, if the respondent did not give any of those four names in response to the first three questions, the interviewer was directed to ask "what other companies or stores" the stimulus t-shirt brought to mind, and only if the respondent answered with one of the four names was the interviewer to ask question five, the dilution question. The text on both of the tested t-shirts began with the prefix "Wal," and Wal-Mart was the only one of the four listed names that began with that prefix.

Smith argues that this series of questions combined with the t-shirt stimulus subtly informed the interviewers not only that a store name was desired, but also that a particular store name — Wal-Mart — was sought. Thus, Smith contends, because the survey failed to meet the double-blind requirement, it was not conducted in an objective manner and must be excluded for what must therefore be biased results; See REFERENCE MANUAL at 248 (noting that poorly formed questions may lead to distorted responses and increased error and therefore may be the basis for rejecting a survey)"

Wal-Mart argues that the skip patterns followed proper protocol and that even if the interviewers guessed that Wal-Mart was involved, there could be no risk of bias because (1) interviewers are professionally trained and adhere to extremely high ethical standards, and (2) it was impossible to determine from the design of the study who sponsored the study and for which aide of a dispute the survey evidence was to be proffered.

Based on the facts that (1) both of the tested t-shirts include the prefix "Wal" and (2) the only store on the specified list of four that included that same prefix was Wal-Mart, it is safe to surmise that the interviewers at least suspected that Wal-Mart was involved in the survey in some manner. Aside from a common sense assumption that the party with deep pockets and reason to be insulted by the tested concepts was likely to have sponsored the research, however, the interviewers had no way to know who was the proponent of the research and who was the opponent. Thus, although the survey design may have breached generally accepted double-blind protocol to some degree, because the breach offered little risk of bias toward one party or the other the Court finds this issue to be of little import in its trustworthiness determination.

(ii) Leading Questions

Smith also argues that the wording of Jacoby's confusion questions was improperly leading. Although the challenged t-shirts were created and offered for sale by Charles Smith, an individual, via his CafePress webstore, the survey asked about sponsorship only in the context of companies or stores, such as in the survey's lead question, which asked, "[W]hich company or store do you think 1332 puts out this shirt?"[27] Smith contends that this wording suggested to the respondent that the interviewer was looking for the name of a company or store, which would lead the respondent away from the answer that the shirt was put out by an individual who was criticizing a company. Wal-Mart counters that because Smith's merchandise was sold through his Cafe-Press webstores, the questions were accurately worded and thus not misleading.

The Court agrees with Smith that the disputed questions improperly led respondents to limit their answers to companies or stores. Though Smith did offer his merchandise through his CafePress webstore, as Wal-Mart argues, the Court finds this characterization disingenuous; the party Wal-Mart sued for offering the Walocaust and Wal-Qaeda merchandise for sale is not a company or a store, but instead Charles Smith, an individual. Furthermore, Wal-Mart has failed to point to any authority supporting the use of the "company or store" language in a consumer "likelihood of confusion" apparel survey or any such surveys previously conducted by Jacoby. Thus, the Court must consider this weakness in determining the admissibility or evidentiary weight to be accorded the survey.

(b) Representativeness

(i) Testing Stimuli

Smith also argues that the Jacoby survey results should not be presumed to represent consumer reaction to any of the challenged merchandise that was not actually tested. Jacoby limited his surveys to testing two specific t-shirts (the Wal★ocaust smiley eagle shirt and the "SUPPORT OUR TROOPS" Wal-Qaeda shirt), and the conclusions stated in his report were narrowly drawn to refer to the tested t-shirts. At his deposition, however, he stated that because the tested shirts were "reasonably representative" of all the shirts that included the prefix "Wal" and the star, as in Wal★ocaust, or the prefix "Wal" and a hyphen, as in Wal-Qaeda, his results could be extrapolated from the tested t-shirts to all of the challenged t-shirts that shared those features.

Jacoby's own deposition testimony supplies a fitting framework for analyzing this issue. When declining to offer an opinion about whether consumers would also be confused over the sponsorship of Smith's Walocaust website, Jacoby stated that consumers respond differently to a given stimulus depending on the context in which is it presented, and because his survey tested only Smith's CafePress webstores, his survey provided him with no data upon which to answer the question about consumer confusion regarding Smith's website.

Applying the same reasoning, the Court finds that test results from one Walocaust or Wal-Qaeda t-shirt provide no data upon which to estimate consumer confusion regarding another Walocaust or Wal-Qaeda t-shirt. A consumer confused about the sponsorship of a shirt that says "SUPPORT OUR TROOPS[.] BOYCOTT WAL-QAEDA" may easily grasp the commentary in the more straightforwardly derogatory "WAL-QAEDA[.] Freedom Haters ALWAYS" concept. Similarly, a consumer confused over the sponsorship of a "Walocaust" shirt paired with an eagle and a smiley face might 1333 have a crystal clear understanding of the word's meaning when it is superimposed over a drawing of a Wal-Mart-like building paired with a sign that advertises family values and discounted alcohol, firearms, and tobacco or when it is presented along with the additional text "The World is Our Labor Camp. Walmart Sucks." As a result, this weakness will also impact the Court's assessment of the survey's evidentiary value.

(ii) Sample Size and Selection

Smith also challenges the survey's small sample size; the Court additionally notes that Jacoby's study employed mall-intercept methodology, which necessarily results in a non-random survey sample.

It is true that the majority of surveys presented for litigation purposes do, in fact, include small and non-random samples that are not projectible to the general population or susceptible to evaluations of statistical significance. 6 MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 32:165 (4th ed.2006). Courts have found that "nonprobability `mall intercept' surveys are sufficiently reliable to be admitted into evidence," reasoning that because "nonprobability surveys are of a type often relied upon by marketing experts and social scientists in forming opinions on customer attitudes and perceptions," they may be admitted into evidence under Federal Rule of Evidence 703 as being "of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." Id.

However, probability surveys are preferred to non-probability surveys. Id. (citing Jacob Jacoby, Survey & Field Experimental Evidence, in SAUL KASSIN & LAWRENCE S. WRIGHTSMAN, JR., 185-86 THE PSYCHOLOGY OF EVIDENCE AND TRIAL PROCEDURE (1985)). Jacoby himself has written that "behavioral science treatises on research methodology are in general agreement that, all other things being equal, probability sampling is preferred to nonprobability sampling." Jacob Jacoby & Amy H. Handlin, Non-Probability Sampling Designs for Litig. Surveys, 81 TRADEMARK REP. 169, 170 (Mar.-Apr.1991) (citing Ku B. RAI AND JOHN C. BLYDENBURGH, POL. SCI. STATS. 99 (Holbrook Press Inc.1973) and quoting its comment that "nonprobability samples do not represent the population truly, and the inapplicability of probability models as well as the impossibility of measuring or controlling random sampling error makes them even less attractive for scientific studies."). Jacoby has similarly noted that although the vast majority of in-person surveys conducted for marketing purposes employ non-probability design, marketers more typically use telephone interviews, a "sizable proportion" of which employ probability designs. Jacoby & Handlin, 81 TRADEMARK REP. at 172 & Table 1 (estimating that sixty-nine percent of commercial marketing and advertising research is conducted by telephone).

Although courts typically admit nonprobability surveys into evidence, many recognize that "the results of a nonprobability survey cannot be statistically extrapolated to the entire universe," and they consequently discount the evidentiary weight accorded to them. Id.; accord Am. Home Prods. Corp. v. Barr Labs., Inc., 656 F.Supp. 1058, 1070 (D.N.J.1987) (criticizing a Jacoby survey and noting, "While non-probability survey results may be admissible, they are weak evidence of behavior patterns in the test universe.") Similarly, "[c]onducting a survey with a number of respondents too small to justify 1334 a reasonable extrapolation to the target group at large will lessen the weight of the survey." 6 MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 32:171.

This Court finds troubling the Jacoby survey's implicit assumption that a study protocol insufficient for many marketing purposes and heavily criticized for behavioral science purposes is nevertheless sufficient to aid a factfinder in a legal action challenging free speech. Therefore, this factor will also affect the Court's assessment of the survey's evidentiary value.

c. Admissibility

Having identified numerous substantial flaws in Jacoby's survey, the Court must now determine whether the flaws limit the survey's evidentiary weight or are so substantial as to render the survey irrelevant or unreliable and therefore inadmissible under Federal Rule of Evidence 403, 702, or 703. See Starter Corp. v. Converse, Inc., 170 F.3d 286, 297 (2d Cir.1999) (excluding a survey under Rule 403 because the probative value of the survey was outweighed by potential prejudice and further noting that "a survey may be kept from the jury's attention entirely by the trial judge if it is irrelevant to the issues") (citing C.A. May Marine Supply Co. v. Brunswick Corp., 649 F.2d 1049 (5th Cir. 1981)); accord Ramdass v. Angelone, 530 U.S. 156, 173, 120 S.Ct. 2113, 147 L.Ed.2d 125 (2000) (listing numerous cases in which courts have excluded or minimized survey evidence as unreliable).

Courts in the Eleventh Circuit typically decline to exclude likelihood of confusion surveys and instead consider a survey's technical flaws when determining the amount of evidentiary weight to accord the survey. See, e.g., Jellibeans, 716 F.2d at 845; Nightlight Sys., Inc. v. Nitelites Franchise Sys., Inc., 2007 WL 4563873 at *5 (N.D.Ga. Jul.17, 2007). Consequently, although this is a close case, the Court concludes that the better option is to admit the survey evidence and to consider the survey's flaws in determining the evidentiary weight to assign the survey in the likelihood of confusion analysis.

The Court finds, however, that because the survey tested only the "SUPPORT OUR TROOPS[.] BOYCOTT WAL-QAEDA" t-shirt and the Walocaust eagle t-shirt, it has no relevance to any of Smith's other Wal-Mart-related concepts. The Court agrees with Jacoby that context matters — a lot — and therefore will not consider Jacoby's survey as evidence of likelihood of confusion with regard to the words "Walocaust" and "Wal-Qaeda" in general; the study is admissible only as to the two concepts that Jacoby actually tested. See Fed.R.Evid. 702 (limiting expert testimony to that "based upon sufficient facts or data").

Even with regard to the tested concepts, the Court finds that the survey was so flawed that it does not create a genuine issue of material fact. See Spraying Sys. Co. v. Delavan, Inc., 975 F.2d 387, 394 (7th Cir.1992) (recognizing that if a proffered survey is severely and materially flawed, it may not be sufficient to establish a genuine issue of material fact even if it purports to show evidence of actual confusion). Jacoby surveyed an overbroad universe, failed to adequately replicate the shopping experience, and asked leading questions. He also surveyed a non-random sample that in any case was too small to allow the results to be projected upon the general market. Thus, the Court finds that the Jacoby survey is so flawed that it does not establish a genuine issue 1335 of material fact with regard to actual confusion, much less prove actual confusion.

Lack of survey evidence showing consumer confusion is not dispositive, however; the Eleventh Circuit has moved away from relying on survey evidence. Frehling Enters. v. Int'l Select Group, Inc., 192 F.3d 1330, 1341 n. 5 (11th Cir. 1999). In fact, a court may find a likelihood of confusion in the absence of any evidence of actual confusion, even though actual confusion is the best evidence of likelihood of confusion. E. Remy Martin & Co. v. Shaw-Ross Int'l Imps., Inc., 756 F.2d 1525, 1529 (11th Cir.1985). Accordingly, the Court will now consider the remaining likelihood of confusion factors.

2. Strength of the Senior Mark

The "strength" of a mark is the measure of its distinctiveness. Safeway Stores, 675 F.2d at 1164. Smith has conceded that the registered trademarks at issue in this case are all very strong. In general, the more the public recognizes a mark as an indication of the origin of certain products or services, the greater the protection that it is afforded. Frehling Enters., 192 F.3d at 1335.

In cases of parody, however, courts have held that the strength of the mark may actually cut against likelihood of confusion because consumers are more likely to recognize that a very famous mark "is being used as part of a jest." Hormel Foods Corp. v. Jim Henson Prods., Inc., 73 F.3d 497, 503 (2d Cir.1996) (finding consumers of Muppet merchandise unlikely to confuse a character named "Spa'am" with the wellknown meat product brand). Accord Louis Vuitton Malletier v. Haute Diggity Dog, LLC, 464 F.Supp.2d 495, 499 (E.D.Va.2006) (noting that "[i]n cases of parody, a strong mark's fame and popularity is precisely the mechanism by which likelihood of confusion is avoided."); World Wrestling Fed'n Entm't Inc. v. Big Dog Holdings, Inc., 280 F.Supp.2d 413, 435-36 (W.D.Pa.2003). This is because a parody depends on lack of confusion to make its point; the parodist relies on the viewer's familiarity with the distinct and idealized image created by the primary mark's owner as a foil for the parodist's own irreverent representation of the mark. Hormel Foods Corp., 73 F.3d at 503 ("A parody must convey two simultaneous — and contradictory — messages: that it is the original, but also that it is not the original and is instead a parody.") (citation omitted) (emphasis in original). If the primary mark is not distinct in the viewer's mind, it will be impossible for that viewer to understand that the parodist's representation does not simply co-opt the primary mark but instead stands in juxtaposition to it. Thus, when the questioned use is a parody, the strength of the primary mark serves to reduce likelihood of confusion.

Although it is undisputed that Wal-Mart possesses strong and widely recognized marks, the Court is persuaded that Smith's use of the marks is unlikely to cause confusion. The terms "Walocaust" and "Wal-Qaeda" are clearly a play on the famous Wal-Mart name. The fact that the real Wal-Mart name and marks are strong and recognizable makes it unlikely that a parody — particularly one that calls to mind the genocide of millions of people, another that evokes the name of a notorious terrorist organization, or even one that simply refers to "Freedom Haters" — will be confused with Wal-Mart's real products. A distinctive mark will not favor the senior mark holder in such circumstances. See Louis Vuitton, 464 F.Supp.2d at 499.

3. Similarity of the Marks at Issue

The parties agree that a parodist must use at least some of the mark he criticizes. 1336 Indeed, for the alleged infringer's work to be a "parody" in the legal sense, the senior user's protected work must be "at least in part the target" of the alleged infringer's satire. Dr. Seuss Enters., 109 F.3d at 1400-01. The point of contention on this factor is whether Smith's designs contained so much Wal-Mart indicia that they became ineffective as parodies and instead appeared to be actually associated with or sponsored by Wal-Mart.

Similarity of the marks in question is based on the "overall impression that the marks create, including the sound, appearance, and manner in which they are used." Custom Mfg. & Eng'g, 508 F.3d at 648. Although Smith repeatedly admits that he intentionally used recognizable portions of Wal-Mart's trademarks and other Wal-Mart-associated indicia[28] for the purpose of evoking Wal-Mart in the minds of consumers, there are significant dissimilarities between the elements he uses and Wal-Mart's marks.

Smith never uses an unmodified Wal-Mart trademark or symbol in its entirety. In the Walocaust and Wal-Qaeda concepts, he uses only the "WAL" portion of the "WAL-MART" mark. In "FREEDOM HATER MART," he uses only the "MART" portion of the mark. The words he combines with the portions of the marks he does use depart significantly both in appearance and sound from Wal-Mart's registered marks: neither "WALCAUST," "WAL-QAEDA" nor "FREEDOM HATER MART" looks or sounds like "WAL-MART."

Smith and Wal-Mart also use their icons in very distinct manners. Smith couples portions of Wal-Mart's registered trademarks with unflattering words, images and portions of words that no rational consumer would expect Wal-Mart to associate with its own marks[29] He arranges to have these concepts, which comment unfavorably on Wal-Mart, printed on t-shirts and other products. Wal-Mart, in contrast, uses its registered marks to identify its buildings, advertising and community support programs and in connection with the services it offers.

"In order that there be infringement of a trademark, the offending mark must so closely approximate the original mark that there is likely to be `palming off of one product as the other." See B.H. Bunn Co. v. AAA Replacement Parts Co., 451 F.2d 1254, 1261 (5th Cir.1971). Smith's uses, viewed in their entirety, depart significantly from Wal-Mart's. Therefore, this factor also weighs against a finding of likelihood of confusion.

4. Similarity of the Products the Marks Represent

Smith arranges to have his concepts printed on t-shirts, beer steins, boxer 1337 shorts, jumpers, hoodies, camisoles, teddy bears and bibs, and he sells them through his Walocaust and Wal-Qaeda webstores. Wal-Mart sells the same categories of merchandise, also in part via its Wal-Mart webstore.

Smith argues that his politically charged designs render his products clearly distinct from any of Wal-Mart's like products and furthermore that there is no evidence that a single Wal-Mart product is sold to the public with a Wal-Mart trademark imprinted on it. Therefore, according to Smith, his products are not similar to Wal-Mart's.

The Eleventh Circuit has held that "[t]his factor requires a determination as to whether the products are the kind that the public attributes to a single source, not whether or not the purchasing public can readily distinguish between the products of the respective parties." Frehling Enters., Inc. v. Int'l Select Group, Inc., 192 F.3d 1330, 1338 (11th Cir.1999). In Frehling, the plaintiff company produced custom, high-end decorative furniture, the alleged infringer sold inexpensive ready-to-assemble furniture, and the companies' furniture was "somewhat dissimilar in composition, function, and design." Id. The district court found that the products were different in function, design, style and price, and accordingly noted that this factor significantly favored the defendant. On appeal, the Eleventh Circuit found that "the attribution of such weight to this factor in [the defendant's favor] was clearly erroneous." According to the court, the products were somewhat similar because they fell within the home furnishings product category, and thus "a reasonable consumer could possibly attribute the products ... to the same source." Id. Thus, the factor was neutral.

Applying this logic, though Smith and Wal-Mart's merchandise is distinguishable in its composition, function and design, these differences are insufficient to cause the "similarity of products" factor to weigh heavily in Smith's favor. Because Smith and Wal-Mart's products are drawn from like categories, a reasonable consumer could possibly attribute a Wal-Mart t-shirt and a Wal-Qaeda t-shirt or a Wal-Mart bib and a Walocaust bib to the same source. Consequently, the Court finds this factor to be neutral.

5. Similarity of Sales Methods

"Dissimilarities between the retail outlets for and the predominant customers of plaintiffs and defendant's goods lessen the possibility of, confusion, mistake, or deception." Id. at 1339 (citation omitted). "This factor takes into consideration where, how, and to whom the parties' products are sold." Id. Even if the outlets and customer bases are not identical, this factor will not weigh heavily in the alleged infringer's favor as long as there is some overlap.

Wal-Mart sells its merchandise through its chain of nearly 6500 brick-and-mortar stores and its Internet site, www.wal-mart. com. Smith's merchandise is available only via his CafePress Walocaust and Wal-Qaeda webstores and is not offered for sale in any physical stores. Though these retail channels are obviously distinct from one another, Smith has produced no evidence that the same types of customers do not shop at both CafePress and the ubiquitous Wal-Mart. As a result, this factor weighs only somewhat in favor of Smith,

6. Similarity of Advertising Methods

This factor looks to whether the parties use similar advertising media. Id. Wal-Mart 1338 has a massive advertising budget, including newspapers, television, radio, a website, paid banners on other websites, keyword advertising through Google and other search engines, and a public relations department. Smith has used no paid advertising, instead promoting his designs via communications to selected liberal groups. He also created websites at walocaust.com and walqaeda.com that explain why he created his designs, discuss this lawsuit, and display his designs with hyperlinks to his CafePress Walocaust and Wal-Qaeda webstores, where products displaying the designs may be purchased.

The difference between the two parties' advertising and promotional efforts is obviously vast, and accordingly this factor weighs heavily in Smith's favor.

7. The Alleged Infringer's Intent

Proof that the secondary user intended to confuse the public is unnecessary to a finding of likelihood of confusion. Elvis Presley Enters. v. Capece, 141 F.3d 188 (5th Cir.1998); see also Commc'ns Satellite Corp. v. Comcet, Inc., 429 F.2d 1245, 1249 (4th Cir.1970) ("While evil intent may evidence unfair competition and deception, lack of guile is immaterial."). The fact that an alleged infringer adopted a mark with the intent to cause consumer confusion alone may be sufficient to support a likelihood of confusion inference. Amstar Corp., 615 F.2d at 263. An intent to parody, however, is not an intent to confuse the public. Jordache Enters. v. Hogg Wyld, Ltd., 828 F.2d 1482, 1486 (10th Cir.1987).

The Court has already found that Smith's designs are parodies of Wal-Mart's. The undisputed facts further evidence Smith's lack of intent to confuse. Smith placed direct criticism of Wal-Mart on his CafePress pages before he received Wal-Mart's cease and desist demand.[30] Similarly, on the CafePress Wal-Qaeda homepage, created after the filing of this lawsuit, the viewer first sees a disclaimer of affiliation with Wal-Mart and a hyperlink to Wal-Mart's own website. Words denouncing Wal-Mart are placed next to some of the designs. Smith's Walocaust and Wal-Qaeda homepages also disclaim affiliation with Wal-Mart and criticize the company in the page's first view, before the hyperlinks to Smith's CafePress stores appear. The criticism and disclaimers show Smith's intent to avoid consumer confusion.

Because Smith used Wal-Mart's marks in parodies, and because the evidence overwhelmingly shows that Smith actively intended to avoid consumer confusion, the Court finds that Smith acted with goodfaith intent. Consequently, this factor is immaterial to the likelihood of confusion analysis.

8. Likelihood of Confusion Summary and Conclusion

Evaluating the overall balance of the seven likelihood of confusion factors, the Court finds that Wal-Mart has failed to demonstrate a likelihood that its trademarks "WALMART," "WAL-MART," and "WAL★MART" and its word mark "ALWAYS LOW PRICES. ALWAYS." would be confused with Smith's "WALOCAUST," "WAL-QAEDA," "FREEDOM HATER MART," or "BENTON★VILLEBULLIES 1339 ALWAYS" concepts. In so finding, the Court concludes that factors three (similarity of the marks), five (similarity of sales methods) and six (similarity of advertising methods), weigh in Smith's favor, with particular emphasis on how different the appearance and usage of the marks were and how vastly the parties' advertising methods differed. The Court concludes that factors one (actual confusion), two (strength of the mark), four (similarity of product) and seven (Smith's intent) favor neither party.

In sum, the Court is convinced that no fair-minded jury could find that a reasonable consumer is likely to be confused by the challenged marks. See Anderson, 477 U.S. at 252, 106 S.Ct. 2505. As a result, the Court GRANTS summary judgment to Smith on Wal-Mart's claims of trademark infringement, unfair business competition, cybersquatting and deceptive trade practices.

D. Trademark Dilution by Tarnishment

Wal-Mart contends that Smith's Walocaust and Wal-Qaeda concepts, by associating Wal-Mart with "the perpetrators of such atrocities as the Holocaust and the attacks of September 11, 2001, unquestionably tarnish the Wal-Mart marks." Dilution by tarnishment recognizes an injury when a "trademark is ... portrayed in an unwholesome or unsavory context likely to evoke unflattering thoughts about the owner's product." Deere & Co. v. MTD Prods., Inc., 41 F.3d 39, 43 (2d Cir.1994); accord Original Appalachian Artworks, Inc. v. Topps Chewing Gum, Inc., 642 F.Supp. 1031, 1039 (N.D.Ga.1986).

"However, tarnishment caused merely by an editorial or artistic parody which satirizes [the complainant's] product or its image is not actionable under an anti-dilution statute because of the free speech protections of the First Amendment." Mattel, Inc. v. Walking Mountain Prods., 353 F.3d 792, 812 (9th Cir.2003) (citation and internal, punctuation omitted). "Parody is a form of noncommercial expression if it does more than propose a commercial transaction." Bolger v. Youngs Drug Prods. Corp., 463 U.S. 60, 66-67, 103 S.Ct. 2875, 77 L.Ed.2d 469 (1983).

A claim of dilution applies only to purely commercial speech. Mattel, 353 F.3d at 812. See also Bolger, 463 U.S. at 66-67, 103 S.Ct. 2875 (finding that materials do not become "commercial speech" simply because the author had economic motivation to create them). "The question whether an economic motive existed is more than a question whether there was an economic incentive for the speaker to make the speech; the Bolger test also requires that the speaker acted substantially out of economic motivation." Procter & Gamble Co. v. Amway Corp., 242 F.3d 539, 552-53 (5th Cir.2001) (emphasis supplied). "Thus, for example, speech that is principally based on religious or political convictions, but which may also benefit the speaker economically, would fall short of the requirement that the speech was economically motivated" and therefore would be considered noncommercial. Id.

At least one court of appeals has specifically addressed whether a social advocate selling t-shirts that carried the group's social message was engaging in noncommercial speech, despite the fact that the group sold the t-shirts to the public for profit. See Ayres v. City of Chicago, 125 F.3d 1010 (7th Cir.1997). In Ayres, the court distinguished limitations on "the sale 1340 of goods that are not themselves forms of protected speech," noting that precedent allows more restriction on sales of nonexpressive goods than it does on goods that are forms of protected speech. Id. at 1015. The court likened t-shirts carrying messages of social advocacy to "the sandwich boards that union pickets sometimes wear." Id. at 1014. As such, the t-shirts were "a medium of expression prima facie protected by the free-speech clause of the First Amendment, and they do not lose their protection by being sold rather than given away." Id. (citing Heffron v. Int'l Soc'y for Krishna Consciousness, Inc., 452 U.S. 640, 647, 101 S.Ct. 2559, 69 L.Ed.2d 298 (1981)).

The Court is convinced that a reasonable juror could only find that Smith primarily intended to express himself with his Walocaust and Wal-Qaeda concepts and that commercial success was a secondary motive at most. Smith has strongly adverse opinions about Wal-Mart; he believes that it has a destructive effect on communities, treats workers badly and has a damaging influence on the United States as a whole. He invented the term "Walocaust" to encapsulate his feelings about Wal-Mart, and he created his Walocaust designs with the intent of calling attention to his beliefs and his cause. He never expected to have any exclusive rights to the word. He created the term "Wal-Qaeda" and designs incorporating it with similar expressive intent. The Court has found those designs to be successful parodies.

Thus, Smith's parodic work is considered noncommercial speech and therefore not subject to Wal-Mart's trademark dilution claims, despite the fact that Smith sold the designs to the public on t-shirts and other novelty merchandise. Consequently, Smith's motion for summary judgment on Wal-Mart's trademark dilution claims is hereby GRANTED.

III. Conclusion

For the foregoing reasons, and subject to the limitations discussed above, the Court DENIES Smith's motion in limine to exclude Wal-Mart's expert witness evidence [78], and Wal-Mart's motions in limine to exclude evidence from Smith's two rebuttal witnesses [81, 82]. Smith's motion for summary judgment [76] is hereby GRANTED, and Wal-Mart's motion for summary judgment [77] is DENIED.

The Court hereby issues a declaratory judgment that Smith's activities have not violated any of Wal-Mart's trademark rights. Smith may maintain his domain names and websites. He may also resume offering for sale via his Walocaust and Wal-Qaeda CafePress webstores his parodic WALOCAUST, WAL-QAEDA, FREEDOM HATER MART, and BENTON ★VILLEBULLIES ALWAYS concepts printed on novelty merchandise; on any webpage or other channel offering such merchandise for sale, Smith must continue to include prominent disclaimers of affiliation with Wal-Mart.

The only possible remaining claim is Smith's claim for attorneys' fees. Although Smith prayed for the recovery of "costs and attorney fees" in the ad damain clause of his complaint, his complaint includes no specific claim therefor, nor did his motion for summary judgment mention any such claim.

1341

1342

FOOTNOTES

[1] Wal-Mart's fiscal year 2008 ended on January 31, 2008; domestic sales figures include approximately, $239 billion attributable to Wal-Mart Stores and approximately $44 billion attributable to Sam's Club. Wal-Mart Investor Relations, Wal-Mart Reports Record Fourth Quarter Sales and Earnings 2 (Feb. 19, 2008), available at http://www.walmartfacts. com/articles/5675.aspx.

[2] Smith uses several variations of this word interchangeably. In this Order, the Court will refer to concepts that include the words "Walocaust," "Wal-ocaust" or "Wal★ocaust" as "Walocaust" concepts or designs unless analysis would be aided by a more precise description.

[3] See Appendix A: Challenged Walocaust Images.

[4] CafePress is an online retailer that sells t-shirts and other items imprinted with designs that individuals create. CafePress prints the items when they are ordered by customers, thus allowing the people who sell through www.CafePress.com to avoid the need to build up an inventory of goods for sale. Cafe-Press.com allows visitors to search by category, department and subtopic, and it allows visitors to view products sorted by newest arrivals or bestsellers. CafePress had been Smith's sole retail outlet.

Smith offered both his Walocaust designs (the eagle, knife fights, and family values concepts) and other designs that did not incorporate the word, such as a design that read, "CHOOSE Minding Your Own Business," a design that depicted an image of shackles along with the legend, "Ownership Society," and another design that depicted a headshot of President George W. Bush along with the phrase "I use [sic] to be all messed up on drugs until I found the lord ... now I'm all messed up on the lord."

[5] For example, Smith continued to offer products imprinted with an image of Vice President. Dick Cheney and the caption "Number 2."

[6] Smith has not asked Wal-Mart to pay him for any of his domain names, nor has he otherwise tried to sell them. He also provided accurate contact information when he registered his domain names.

[7] Smith uses two variations of this word interchangeably. In this Order, the Court will refer to concepts that include the words "Wal-Qaeda" or "Walqaeda" as "Wal-Qaeda" concepts or designs unless analysis would be aided by a more precise description.

[8] See Appendix B: Challenged Wal-Qaeda Images.

[9] In all of the concepts, "WAL-QAEDA" was depicted in a block font reminiscent of Wal-Mart's font. On the storefront concept, the font was white, and on all other concepts the font was blue. In the "WAL-QAEDA[.] Freedom Haters ALWAYS" concept, "ALWAYS" was depicted in red at approximately a forty-five degree angle to the remainder of the text, mimicking Wal-Mart's registered mark, "ALWAYS LOW PRICES. ALWAYS."

[10] The Wal-Qaeda CafePress account, like the Walocaust CafePress account, was also a premium account with a "medium" markup.

[11] Aside from the fifteen shirts bearing the "SUPPORT OUR TROOPS[.] BOYCOTT WAL-QAEDA" design that were ordered by and sold to one of the law firms representing Wal-Mart in this matter, Smith sold twenty-two shirts bearing his "Wal-Qaeda mural" design, sixteen shirts bearing his "CAUSE OF DEATH: A Dime Store" design, five shirts bearing his "WAL-QAEDAN Freedom Haters ALWAYS" design, three shirts bearing his "SUPPORT OUR TROOPS[.] BOYCOTT WAL-QAEDA" design, and one shirt bearing his "WAL-QAEDA Human Resource Achievement Award (with a bust of Chairman Mao)" design.

[12] In the context of a trademark infringement action, a trademark has been "used in commerce" if it has been "placed in any manner on the goods ... or the displays associated therewith ... and the goods are sold or transported in commerce." 15 U.S.C. § 1127; accord Optimum Techs., Inc. v. Henkel Consumer Adhesives, Inc., 496 F.3d 1231, 1242 (11th Cir.2007). It is undisputed that Smith used portions of Wal-Mart's registered marks on merchandise that he displayed and sold through his Walocaust and Wal-Qaeda CafePress webstores. As a result, the "use in commerce" element of the trademark infringement action has been satisfied.

[13] The Court rejects Wal-Mart's assertion that its February 21, 2007 Order established that Wal-Mart has enforceable trademark rights in the yellow smiley face. In briefs submitted in connection with that Order, Smith admitted that the registered, whole trademarks for which Wal-Mart seeks protection are all very strong, but he vigorously contended that third-party usage diminished both the strength of certain portions of the registered marks, such as "WAL" without the use of "MART" and "MART" without the use of, "WAL," and the strength of the unregistered smiley face. See Smith v. Wal-Mart Stores, Inc., 475 F.Supp.2d 1318, 1321-22 (N.D.Ga.2007).

While it is true, as the Court noted in the February 21 Order, that use of a trademark in a parody for the purpose of evoking a company in the minds of consumers is generally considered to be evidence of a strong mark, such a use does not establish that an icon or symbol is, in fact, a protectible trademark or the property of the company it references. Therefore, although Smith admittedly used the smiley face to evoke Wal-Mart in the minds of consumers, this fact alone does not establish that the smiley face is a defendable trademark or that Wal-Mart owns trademark rights in it.

[14] It is undisputed that the smiley face is not inherently distinctive.

[15] This resulted in eight test cells: ---------------------------------------------------------------------------------------------------------- Test cells Control cells ---------------------------------------------------------------------------------------------------------- Post-purchase Wal★ocaust Wal-Qaeda Zal-ocaust Zal-Qaeda confusion/tarnishment t-shirt t-shirt t-shirt t-shirt Point-of-sale Wal★ocaust Wal-Qaeda Zal-ocaust Zal-Qaeda confusion/tarnishment website website website website ----------------------------------------------------------------------------------------------------------

[16] The research company conducted the surveys in malls in Trumbull, Connecticut; Philadelphia, Pennsylvania; Youngstown, Ohio; Chicago Ridge, Illinois; Louisville, Kentucky; San Antonio, Texas; Colorado Springs, Colorado; and Northridge, California. The website survey was also conducted in Portland, Oregon.

[17] Because CafePress allowed only consumers over the age of thirteen to purchase from its site, Jacoby similarly limited his universe of respondents.

[18] Respondents who worked at an advertising agency, a market research firm or a business located in the mall (or had an immediate family member who did) were excluded, as were people who normally wore eyeglasses or contact lenses but were not wearing them at the time of the screening.

[19] The screening questionnaire provided to the Court indicates that the respondents who then participated in the surveys were given a monetary reward. Neither Jacoby's report nor any of the supporting survey documents disclosed the amount of the reward.

[20] The simulations were reproduced on a compact disc; the respondents did not view Smith's actual web pages on the Internet.

[21] Jacoby arrived at these numbers by averaging the net survey results for the Walocaust and Wal-Qaeda t-shirts.

[22] Wal-Mart presents no authority supporting its argument that Rosenblatt was required to conduct his own study of Smith's websites, and the Court sees no reason why a specific study of Smith's websites would be necessary to make relevant or reliable Rosenblatt's testimony criticizing Jacoby's assumptions about how consumers generally navigate the Internet. See Hill's Pet Nutrition, Inc. v. Nutro Prods., Inc., 258 F.Supp.2d 1197, 1210 (D.Kan.2003) (rejecting a survey criticized by Jacoby even though Jacoby had not performed his own survey).

[23] The basis for this testimony, however, shall be limited to Rosenblatt's personal observations and published industry knowledge disclosed to Wal-Mart in discovery. Testimony quoting industry "conventional wisdom" statistics unsupported by published research, such as the "50 percent loss-per-click theory" is inadmissible regardless of whether "everybody in the business cites this as a reality." Such a statistic is too specific to rest exclusively on conventional wisdom, which is often wrong.

[24] The case law Wal-Mart cites to support its argument that surveying only CafePress shoppers would have resulted in too narrow a universe is inapposite. In Scott Fetzer Co. v. House of Vacuums, Inc., 381 F.3d 477 (5th Cir.2004) the court did in fact hold that an unduly narrow universe can undermine a study's value, but it rejected the study because the survey universe was limited to purchasers of the markholder's product. Id. at 487-88. The court found that the universe was uniquely familiar with the markholder's marketing and distribution techniques and would therefore be unlikely to accurately represent the views of "purchasers most likely to partake of the alleged infringer's goods or services." Id. (citation and internal punctuation omitted). Wal-Mart has offered no proof that it even sells merchandise imprinted with designs incorporating the prefix "WAL," much less that Rosenblatt ever suggested that Jacoby should have limited his survey universe to purchasers of such Wal-Mart products.

The Court also rejects Jacoby's theory that limiting the universe to CafePress shoppers would have made the universe so small as to make testing unfeasible and would have led to biased results because CafePress shoppers were likely to have seen Smith's t-shirts and may have come to the survey with preconceived notions. In his deposition, Jacoby testified that he had no idea how many customers shopped at CafePress, and Wal-Mart has provided no factual support for the theory that CafePress customers would be familiar with Smith's t-shirts and therefore biased by preconceived notions. Therefore, both Wal-Mart's "narrow universe" argument and its "bias" argument fail.

[25] Wal-Mart also argues that because Teach's report contained numerous typographical errors and his analysis improperly applied certain statistical methodology to Jacoby's research results, the Court should also disregard it as irrelevant. Smith concedes that Teach's attempt to apply statistical significance testing to Jacoby's non-random sample was inappropriate, and thus the Court will not consider that portion of Teach's testimony. Typographical errors do not necessarily reduce the evidentiary value of this general testimony and therefore do not require its exclusion. Therefore, the Court will consider only Teach's first report and not the revised report in which he corrected most of his typographical errors but which Wal-Mart challenges as untimely.

[26] Because the Court finds nothing in Teach's testimony or credentials that even hints that Teach is an expert on consumer Internet behavior, the Court will disregard Teach's indication that he agrees with Rosenblatt's testimony.

[27] Question 2a asked about the business connections of "the company or store that puts out this shirt." Question 4c asked whether the shirt made the respondent "think of any particular companies or stores."

[28] Such as the smiley face, star, colors and fonts Wal-Mart often uses.

[29] Indeed, as Wal-Mart has argued, by evoking the Holocaust and Al-Qaeda, Smith's concepts are not merely unflattering, but veer toward the outrageous and offensive. This actually falls in Smith's favor, as courts have held that "the more distasteful and bizarre the parody, the less likely the public is to mistakenly think that the trademark owner has sponsored or approved it." Burnett v. Twentieth Century Fox Film Corp., 491 F.Supp.2d 962, 972 (C.D.Cal.2007). Smith further distinguishes his Walocaust and Wal-Qaeda CafePress homepages from Wal-Mart's website by including a disclaimer of affiliation with Wal-Mart. See Faegre & Benson, LLP v. Purdy, 367 F.Supp.2d 1238, 1244 (D.Minn. 2005) (noting that although the Lanham Act does not require that a parody carry a disclaimer, a clear disclaimer "should alert most consumers that the item is a parody.") (quoting Cliffs Notes, Inc. v. Bantam Doubleday Dell Publ'g Group, 886 F.2d 490, 496 (2d Cir.1989)) (internal punctuation omitted).

[30] For example, he placed several biting lines about Wal-Mart's labor practices at the top of his CafePress Walocaust homepage, and the phrase "Walocaust: The World is Our Labor Camp. Wal-Mart Sucks" was displayed directly below the "eagle" design the first time it appeared on the page.