is trademark dilution a unicorn duke · is trademark dilution a unicorn? an experimental...

1

Is Trademark Dilution a Unicorn? An Experimental Investigation

Barton Beebe,1 Roy Germano,2 Christopher Jon Sprigman,3 and Joel Steckel4

DRAFT

INTRODUCTION

Trademark dilution is among the most elusive concepts in intellectual property law. In theory, trademark dilution describes situations in which one trademark dilutes the distinctiveness of another similar mark even when consumers are aware that the two marks come from two different sources. If someone opens a bar in Brooklyn called “Apple,” it is highly unlikely that patrons will be misled into believing that the global high-technology company owns the bar and is leveraging its brand into the sale of craft cocktails. But the concept of trademark dilution asserts that there may nevertheless be harm to the famous global brand. When Brooklynites hear the term “apple,” they may now “think for a moment”5 to determine to which enterprise the speaker is referring. If multiple Brooklyn businesses -- artisanal chocolatiers, purveyors of vintage men’s shaving accessories, shops selling Japanese raw denim -- begin to call themselves “Apple,” the link between the term and the Cupertino company could be significantly “blurred.” This blurring would damage the famous brand.

That, in any case, is the theory of trademark dilution. This theory (and a great deal of lobbying by owners of famous brands) prompted Congress to pass the Federal Trademark

1 John M. Desmarais Professor of Intellectual Property Law, New York University School of Law; Co-Director, Engelberg Center on Innovation Law and Policy. 2 Research Scholar, New York University School of Law. 3 Professor, New York University School of Law; Co-Director, Engelberg Center on Innovation Law and Policy. 4 Professor of Marketing and Vice Dean for Doctoral Education, NYU Stern School of Business. The authors thank the Engelberg Center for providing funding that supported this project. [OTHER ACKNOWLEDGMENTS] 5 Richard Posner, When Is Parody Fair Use?, 21 J. LEGAL STUDIES 67, 75 (1992) (“A trademark seeks to economize on information costs by providing a compact, memorable and unambiguous identifier of a product or service. The economy is less when, because the trademark has other associations, a person seeing it must think for a moment before recognizing it as the mark of the product or service.”). See also Ty Inc. v. Perryman, 306 F.3d 509, 511 (7th Cir. 2002) (Posner, J.) (describing dilution by blurring as involving situations where “[c]onsumers will have to think harder—incur as it were a higher imagination cost—” to determine to which entity a trademark is referring).

2

Dilution Act (FTDA) in 1995, which defined and prohibited dilution.6 The law was a failure. Suspicious of the in-gross rights that the antidilution cause of action entailed and unsure in any case of what dilution was, courts applied the law, if they applied it all, in a variety of conflicting ways.7 A Supreme Court ruling in 2003 effectively rendered it a nullity.8 In 2006, Congress tried a second time to regulate dilution with the Trademark Dilution Revision Act (TDRA), which replaced the FTDA in its entirety.9 The TDRA defines “dilution by blurring,” which is the form of dilution that concerns us in this paper, as “association arising from the similarity between a mark or trade name and a famous mark that impairs the distinctiveness of the famous mark.”10 Now in its eleventh year, the TDRA is on course to survive longer than the FTDA. Empirical studies confirm that it has had almost no effect on outcomes in court,11 but owners of famous marks (and some not-so-famous ones) continue to press dilution claims.

In practice, the antidilution cause of action has created evidentiary conflicts. We lack persuasive means of showing whether dilution even occurs, not just in a particular litigation, but whether it ever occurs at all. How would one go about showing that the boîte in Brooklyn has harmed in any way the world-famous brand with which it shares a name? The leading— arguably, the only—theoretically based method of doing so takes the form of response time studies.12 In essence, these studies expose a treatment group to diluting stimuli (such as an

6 Federal Trademark Dilution Act of 1995, Pub. L. No. 104-98, 109 Stat. 985 (1996). 7 See Monica Hof Wallace, Using the Past to Predict the Future: Refocusing the Analysis of a Federal Dilution Claim, 73 U. Cɪɴ. L. Rᴇᴠ. 945, 945 (2005) (discussing courts’ “scattered analyses” of the FTDA); Sara Stadler Nelson, The Wages of Ubiquity in Trademark Law, 88 Iᴏᴡᴀ L. Rᴇᴠ. 731, 771-72 (2003) (discussing Circuit Courts of Appeals’ differing interpretations of the concept of distinctiveness under the FTDA). 8 Moseley v. V Secret Catalogue, Inc., 537 U.S. 418 (2003) (holding that federal antidilution law in effect at the time required plaintiffs to show “actual dilution” to prevail in an antidilution cause of action). 9 Trademark Dilution Act of 2006, Pub. L. No. 109-312, 120 Stat. 1730. 10 15 U.S.C. § 1125(c)(2)(B). An alternative form of dilution, which we do not discuss in this paper, is dilution by tarnishment. The TDRA defines dilution by tarnishment as “association arising from the similarity between a mark or trade name and a famous mark that harms the reputation of the famous mark.” 15 U.S.C. § 1125(c)(2)(C). See also Deere & Co. v. MTD Prods., Inc., 41 F.3d 39, 43 (2d Cir. 1994) (“‘Tarnishment’ generally arises when the plaintiff's trademark is linked to products of shoddy quality, or is portrayed in an unwholesome or unsavory context likely to evoke unflattering thoughts about the owner's product[s].”). 11 See Barton Beebe, The Continuing Debacle of U.S. Antidilution Law: Evidence from One Year of Trademark Dilution Revision Act Case Law, 24 SANTA CLARA COMPUTER & HIGH TECH. L.J. 449 (2008) (reviewing all federal case law applying the TDRA in its first year of operation and showing that the new law had no appreciable effect on case outcomes or on remedies issuing from those outcomes); Barton Beebe, The Still Continuing Debacle of U.S. Antidilution Law: Evidence from Ten Years of Trademark Dilution Revision Act Case Law (working paper) (reviewing all federal court case law applying the TDRA from 2006 through 2016 and noting the very limited number of cases in which relief was granted solely on dilution grounds). 12 The most common approach to measuring dilution in litigation is that used by the plaintiff in Nike, Inc. v. Nikepal Int’l, Inc. (Case No. 2:05-1468-GEB-JFM, United States District Court, Eastern District of California.). Nikepal used its brand on glass syringes and other sorts of laboratory equipment. The study produced in that case measured dilution by inquiring whether, when subjects were presented with the

3

advertisement for HEINEKEN popcorn that explicitly states that it is not made by the beer maker) and then require the treatment and control groups to match brand names and product categories as quickly but as accurately as possible. The studies, which we discuss in more detail below, report that respondents exposed to diluting stimuli take longer to match—in the example used just now, respondents might take longer to match HEINEKEN to its correct product category (beer) or a product attribute (e.g., bitterness) that consumer surveys previously have identified as linked to the brand. This, then, shows that consumers exposed to diluting marks must “think for a moment” whenever they see the diluted mark.

We initiated this project to test the hypothesis that existing response time studies were flawed because they did not place respondents in a purchasing context. Our hunch was that if both the treatment and control groups were prompted to think about purchasing a product in the product area of the famous brand, and not just merely shown the brand, when shown a diluting stimulus, the difference in matching response time would vanish. We developed a series of protocols to test this hypothesis. In doing so, we included a control group that was exposed to diluting stimuli for famous brands other than those that were the main focus of the test. For example, we exposed a control group to an advertisement for NIKE Toothpaste and then tested that group’s response times in linking MERCEDES with cars.

What we found surprised us. In short, we found that any diluting stimuli would trigger increased response times with respect to nearly any brand. Our data led us to suspect that upon exposure to any diluting stimuli, test respondents become wary and cautious, which is reflected in their increased response times. In other words, exposure to a diluting stimulus causes respondents to “think for a moment” in matching regardless of whether the diluting stimulus carried a brand name or trademark similar to that of the target brand.

On this basis, we conclude that the current leading response time studies that purport to show dilution are fatally flawed. Those studies exposed a treatment group to diluting stimuli and a control group to no diluting stimuli and compared response times. In retrospect, it should not be surprising that these studies found significant differences in response times. Instead, they should have exposed the control group to diluting stimuli, but for brands other than those that were the focus of the test. In essence, these studies used the wrong control.

word Nikepal, Nike came to mind. The study, and the court, took this as evidence of dilution. We would argue that it is less than surprising that consumers would think of “Nike” when presented with the word “Nikepal”. Nike is one of the world’s best known brand names. But the fact that a consumer thinks of a famous mark when he or she sees a word containing that mark doesn’t mean that the famous mark is “diluted”. Many experts have commented that the measure used in Nike v. Nikepal and other dilution cases lacks construct validity. More specifically, the court did not explain how mere association translates into evidence that use of the junior mark impairs the distinctiveness (or harms the reputation) of Nike. See also Starbucks Corp. v. Wolfe’s Borough Coffee, Inc., 736 F.3d 198, 210-11 (2d Cir. 2013) (affirming the district court’s accordance of little weight to a survey similar to that conducted in Nikepal because the survey did not present the marks as they appear in commerce and did not show more than minimal actual association).

4

Our findings show that the only empirical studies claiming to show that dilution occurs are invalid. Congress and the courts have devoted an enormous amount of effort to regulating a harm that remains at best theoretical. While future empirical studies may be able to detect “association” between two similar marks that “impairs the distinctiveness” of one of them, there is currently no proof that trademark dilution even exists.

Part I reviews in more detail the response time studies of trademark dilution. Part II reviews our experimental protocols and reports our results. Part III discusses those results.

I. PREVIOUS SCHOLARSHIP

A. Morrin and Jacoby

The Morrin & Jacoby (“MJ”) study conceptualizes trademark dilution “in terms of the strength of association between brand names and their product categories,” and sets out to test the hypothesis that “[c]onsumers exposed to trademark-diluting advertisements will (a) commit more brand recognition errors and (b) exhibit slower brand recognition reaction times than will consumers not exposed to such advertisements.”13 Their method was straightforward. They exposed 64 subjects each to a total of six “tombstone” text advertisements, some of which were ostensibly diluting of three target brands (HEINEKEN, HYATT, and GODIVA), and then recorded the speed and accuracy with which the subjects attempted to match the target brands with their product categories and characteristics.14

The treatment group consisted of thirty-three subjects who viewed three text-only “filler” (tombstone-style) advertisements for well-known brands that contained true representations of each brand’s product category and attributes (for CHASE MANHATTAN, KODAK, and PERRIER), one diluting advertisement, one tarnishing advertisement,15 and one brand extension advertisement.16

The blurring advertisement took the form of an advertisement either for HEINEKEN popcorn or HYATT legal services. Each blurring advertisement contained a disclaimer stating either that “HEINEKEN popcorn is NOT associated with the makers of Heineken beer” or that “HYATT legal services is NOT associated with Hyatt Hotels, Inc.”17 The tarnishing advertisement was for DOGIVA dog biscuits and contained no disclaimer with respect to GODIVA.18 The brand extension advertisement was for either HEINEKEN popcorn or HYATT legal services—whichever was not used in the blurring advertisement—and apparently contained language announcing the brand extension.

13 Maureen Morrin & Jacob Jacoby, Trademark Dilution: Empirical Measures for an Elusive Concept, 19 J. Pᴜʙ. Pᴏʟ’ʏ & Mᴀʀᴋᴇᴛɪɴɢ 265, 268 (2000). 14 Id. 15 Though Morrin & Jacoby focused their study on dilution by blurring, they also tested for dilution by tarnishment. We discuss dilution by tarnishment infra, note __. 16 Id. 17 Id. at 275. 18 Id.

5

One control group consisted of sixteen subjects who were exposed to the three filler advertisements and to three unrelated advertisements.19 A second control group consisted of fifteen subjects who were exposed to the three filler advertisements and three reinforcing advertisements for HEINEKEN beer, HYATT hotels, and GODIVA chocolate.20

Immediately after viewing the six advertisements, subjects engaged in a computer task in which two words appeared sequentially on a screen. Subjects were instructed to indicate as quickly but as accurately as possible whether the two words “represented a match”21 by hitting one key for “Yes” or another for “No.” Among the word pairs presented were pairs relating to category membership (for example, GODIVA/chocolates, chocolates/GODIVA) and attribute possession (for example, GODIVA/rich taste, rich taste/GODIVA). In explaining this part of their protocol, MJ present examples relating only to GODIVA, but apparently the appearance of pairings such as HEINEKEN/popcorn and HEINEKEN/beer would both represent correct matches for those subjects exposed to the HEINEKEN popcorn advertisement. MJ reported that the mean response times of the treatment group were significantly higher than those of either of the control groups, but only with respect to matches relating to HEINEKEN and GODIVA and not with respect to matches relating to HYATT. For matches relating to HEINEKEN and GODIVA taken together, treatment subjects’ mean response time was 836 milliseconds while it was 672 milliseconds for the control group exposed to reinforcing advertisements and 713 milliseconds for the control group exposed to unrelated advertisements.22 Meanwhile, for matches relating to HYATT, the treatment group’s mean response time was 680 milliseconds, which was exactly the same as the reinforced control group’s mean response time and substantially less than the unrelated control group’s mean response time of 810 milliseconds.23 MJ further reported that the treatment group’s matching accuracy was lower than both control groups and that this difference was statistically significant. Specifically, the treatment group accurately judged 73.1% of the matches, while the unrelated control group and the reinforced control group accurately judged 84.3% and 88.9%, respectively, of the matches.24 As MJ saw it, these results fully supported the hypothesis that consumers exposed to diluting advertisements will commit more brand recognition errors, but only partially supported the hypothesis that such consumers will react more slowly to brand recognition tasks. They speculated that subjects’ increased familiarity with the HYATT brand moderated the effect on response times of the HYATT legal services advertisement.25 MJ conducted a subsequent study that tested consumers’ ability after exposure to diluting stimuli to recall the categories of goods in connection with which certain brand were used. The study tested both well-known and unknown brands. Based on the results of this study, MJ concluded that “it appears that very

19 Id. at 268. 20 Id. 21 Id. at 269. 22 Id. 23 Id. 24 Id. 25 Id.

6

strong brands are immune to dilution because their memory connections are so strong that it is difficult for consumers to alter them or create new ones with the same brand.”26

B. Pullig, Simmons, and Netemeyer

Another response time study similarly purported to measure a diluting effect on both category and attribute association. For Pullig, Simmons, and Netemeyer (“PSN”), blurring is “a weakening of the associations between a brand and its distinctive aspects (e.g., category, attributes, benefits) that define the meaning of the brand in consumers’ minds.”27 More specifically, they define blurring as (1) a lessening of a brand’s “brand name accessibility,” i.e., the likelihood that a brand will come to mind when a brand characteristic is presented, or (2) a lessening of a brand’s “aspect accessibility,” i.e., the likelihood that a particular characteristic of the brand will come to mind when the brand name is presented.28

PSN’s response time studies sought to investigate whether dilution is mitigated when the senior and junior uses of a mark are attached to goods in a similar category or that have similar attributes. Their first study focused on category similarity and held attribute similarity constant. With respect to category similarity, they hypothesized that when category similarity is high, the junior use of the mark will reinforce the senior mark’s brand name accessibility and attribute accessibility, and that when category similarity is low, the junior use will lessen accessibility and thus dilute the mark.

PSN’s experimental protocol was complex and need not be described in detail here. In essence, they focused on BIG RED, a brand of chewing gum reasonably well-known in the U.S. for having a cinnamon flavor, and used a hypothetical BRAND A as a control. To study the effect of category similarity, they exposed eighty-nine subjects to a total of three advertisements. These advertisements implemented a “2 (category similarity: high [bubble gum] or low [snack bars]) x 2 (new product: junior brand [BIG RED] and unbranded [BRAND A]) x 2 (order of ad presentation) between subjects design.”29 More specifically, one advertisement was for bubble gum (branded as either BIG RED, BRAND A, or BAZOOKA), one was for bite-size candy (branded as SKITTLES), and one for snack bars (branded as either BIG RED, BRAND A, or NUTRI-GRAIN). Each advertisement contained the brand name, a product category, and two claims. To hold constant attribute similarity, certain ads claimed “great cinnamon flavor.” The BIG RED bubble gum advertisement also promised to “blow awesome bubbles,” while for the BIG RED snack bars advertisement, the other claim stated “enriched with vitamins and minerals.”

After viewing the advertisements, subjects engaged in a computer-administered task. The subject was presented with a brand name or aspect and, after pressing the space bar, an aspect or

26 Id. at 274. 27 Chris Pullig, Carolyn J. Simmons, and Richard G. Netemyer, Brand Dilution: When Do New Brands Hurt Existing Brands? 70 J. Mᴀʀᴋᴇᴛɪɴɢ 52, 54 (Apr. 2006). 28 Id. 29 Id. at 56-57.

7

brand name (depending on what was not presented first) appeared. The subject then pressed a key as quickly but as accurately as possible to indicate whether the brand name and aspect matched.

PSN reported that subjects exposed to the use of BIG RED in the similar product category (bubble gum) exhibited significantly faster response times linking BIG RED and chewing gum and linking BIG RED and cinnamon flavor than subjects exposed to BRAND A bubble gum. The difference in speed ranged from 174 milliseconds to 302 milliseconds. These results suggested that junior uses in a similar product category reinforce rather than dilute the senior mark.

By contrast, subjects exposed to the use of BIG RED in the different product category (snack bars) exhibited significantly slower response times than subjects exposed to BRAND A snack bars. Here the difference ranged from 119 milliseconds to 334 milliseconds. These results suggested to the authors that junior uses in a different product category dilute the mark.30

In a second study, PSN manipulated both category similarity and attribute similarity. Their protocol was based on that of the previous study but with an additional dimension varying attribute similarity between high, “cinnamon flavor and breath freshening,” and low, “strawberry flavor and containing immunity-boosting Echinacea.” Their response times results suggested to them that when the junior use is in a similar product category and the product possesses similar product attributes, the senior mark is reinforced. When the junior use is in a different product category and the product possesses different attributes, the senior mark is diluted. However, the response time results were otherwise mixed and suggest that attribute similarity may sometimes offset category dissimilarity and forestall dilution.31

C. Criticisms

Scholars have criticized both the MJ and PSN studies on a number of grounds.32 Most significantly, Rebecca Tushnet has questioned whether increases in consumer response times of a few hundred milliseconds reflect any significant damage to the diluted brand.33 The studies do not show that the delays they describe actually change purchasing decisions. Furthermore, none of the response time studies sought to place their test subjects in a purchasing context. On the contrary, both the MJ and PSN studies were explicitly decontextualized. In MJ’s study, after the test subjects had viewed the six advertisements and before they engaged in the response time

30 Id. at 58. 31 Id. at 60. 32 See, e.g., Daniel Klerman, Trademark Dilution, Search Costs, and Naked Licensing, 74 Fᴏʀᴅʜᴀᴍ L. Rᴇᴠ. 1759, 1765 (2006); Matthew D. Bunker & Kim Bissell, Lost in the Semiotic Maze: Empirical Approaches to Blurring in Trademark Law, 18 Cᴏᴍᴍ. L. & Pᴏʟ’ʏ 375, 389 (Autumn 2013); Alexander Dworkowitz, Ending Dilution Doublespeak: Reviving the Concept of Economic Harm in the Dilution Action, 20 Tᴇx. Iɴᴛᴇʟʟ. Pʀᴏᴘ. L.J. 25, 39 (Fall 2011). 33 Rebecca Tushnet, Gone in Sixty Milliseconds: Trademark Law and Cognitive Science, 86 Tᴇx. L. Rᴇᴠ. 508, 528 (2008).

8

test, the subjects were told that they would now be tested for information about various brands.34 Similarly, PSN told their test subjects that they were interested in the subjects’ reactions to the informational content of various advertisements.35 Each of these response time studies tested subjects on trademarks that appeared in the abstract. As Tushnet emphasizes, no consumer confronts trademarks in this manner.36

We propose a further, more fundamental limitation of the response time studies, one which we believe may be fatal. The brands MJ and PSN chose to study (HEINEKEN, HYATT, GODIVA, BIG RED) were chosen precisely because consumers tend strongly to associate these brands with particular product categories (beer, hotels, chocolate, chewing gum). Each of the brands possessed high “brand typicality.”37 The studies then presented treatment subjects with a use of the well-known brand on a product category different from that with which the brand was conventionally associated. We believe that this alone would have been enough to lead the treatment subjects to be surprised by the treatment ad. This might further lead treatment subjects to be conscious of the motives underlying the test and thus wary of any matching task.

Both MJ and PSN used control groups consisting of subjects that were not exposed to any diluting advertisements and thus were never given cause to be suspicious of the test. It should not be surprising, then, that the control groups responded more quickly to the matching tasks assigned to them. The appropriate control, by contrast, consists of test subjects exposed to diluting advertisements directed not at the brand of interest but at other brands. In this design, both the treatment group and the control group would experience the surprise of being exposed to a well-known brand associated with an unexpected product category. Treatment subjects would be exposed to a diluting advertisement directed at the targeted brand. Control subjects would be exposed to a diluting advertisement directed at another brand. The difference of interest is the difference in how each group responds to a matching task involving the targeted brand.

II. Protocol and Results

A. Brand Strength

When we began our research, we were interested in two primary questions:

(1) Do diluting stimuli cause increases in subjects’ response time in matching a brand with a product category?

(2) Does context that relates to the product category tend to vitiate a diluting effect on response times?

34 Morrin & Jacoby at 269. 35 Pullig et al at 57. In a different study not involving response times, Pullig, Simmons, and Netemeyer did place subjects in a “simulated shopping experience.” See id. at 62-64. 36 Tushnet at 530-32. 37 See also Alexander F. Simonson, How and When Do Trademarks Dilute: A Behavioral Framework to Judge “Likelihood of Dilution”, 83 Tʀᴀᴅᴇᴍᴀʀᴋ Rᴇᴘ. 149, 152-54 (1993).

9

We began our study with the assumption that response time experiments are a good way to measure dilution. Our objective was to test whether context could mitigate the effects of dilution on response time. We hypothesized that by providing some context relevant to a brand’s product category, we could reduce or eliminate delays in response time caused by diluting stimuli.

We tested our hypothesis on two car brands, one very strong, and the other relatively less strong. To measure brand strength, we first showed 500 people the names of up to twenty popular car brands. The brand names were shown one by one in random order. For each brand name, respondents were asked whether or not they were familiar with the brand. Familiarity with the brand was defined as knowing something about it other than that it exists. Respondents responded “yes” or “no” to indicate familiarity.

Once a respondent identified five familiar brands, we stopped showing that respondent brands, and permitted the respondent to move on to the next phase of the survey. In this second phase, respondents were asked to name the top five things that came to mind when they think of each of the brands they identified as familiar; i.e., we asked for each individual’s top five brand associations. Our assumption was that very strong brands would bring to mind a relatively concentrated set of associations, most of which would be positive. Conversely, weaker brands would bring to mind a larger set of more diffuse, and sometimes negative, associations.

To determine how concentrated a brand’s associations were, we collected all of the associations that respondents provided for each brand. We then grouped like associations and calculated the percentage of subjects identifying each association. We then added up the percentages for the five most common associations mentioned for each brand. This gave us a concentration score. The mean concentration score was 39 percent. We considered a brand “strong” if its concentration score was at least one standard deviation above the mean. This left us with MERCEDES (57.41 percent), BMW (55.92 percent), and INFINITI (48.34 percent). Of our three strongest brands, we first selected MERCEDES because it had the highest concentration score. For our second test brand, we chose INFINITI. INFINITI had a relatively high concentration of (mostly positive) associations, but the brand was far less familiar to respondents in our sample versus MERCEDES, BMW, and many other brands. INFINITI, in fact, was among the least recognized brands out of the twenty we asked about. We suspected that although INFINITI has the makings of a strong brand among those who are familiar with it, it may be more easily diluted since it is not as well known.

B. Response Time

In our first of three experiments, we measured the time it took subjects (N = 1800) to identify whether the terms MERCEDES-CARS and INFINITI-CARS are a match. We recruited our subjects via Amazon’s Mechanical Turk. Once a subject selected the task we listed on

10

Mechanical Turk, he or she was directed to the Qualtrics online social science research platform to complete the protocol.38

1) Pilot #1

In this initial stage, we built an instrument in Qualtrics consistent with those used by Morrin and Jacoby and Pullig et al.39 We piloted this instrument on a small sample of subjects (N = 100).

Subjects in our first pilot were randomly assigned to a brand—either MERCEDES or INFINITI—then to one of three conditions. Subjects in the control group saw three “filler” text-only “tombstone” advertisements for well-known brands (PERRIER sparkling water, CHASE bank, UNITED VAN LINES) that contained true representations of each brand’s product category and attributes. Subjects in our two treatment groups saw the three filler ads plus an ad for a fictitious diluting brand. Subjects assigned to the MERCEDES group saw an ad for a diluting brand called MERCEDES Toothpaste and subjects assigned to the INFINITI group saw an ad for a diluting brand called INFINITI Toothpaste.

38 Online crowd-sourcing services like Amazon Mechanical Turk (AMT) offer researchers the advantage of being able to run experimental protocols on large samples at a fraction of the cost of lab studies. They also allow researchers to pilot various instruments so that hypotheses and empirical tools can be tested, refined, and retested. Online reaction time experiments are possible because JavaScript timers can record clicks down to the millisecond, which we confirmed with representatives from Qualtrics. There are, however, some potential disadvantages to conducting a reaction time experiment online that we investigated and addressed. Namely, subjects in an online experiment use their own hardware and software, in contrast to a laboratory setting where subjects use the same hardware and software.

While there are surely differences in the connection speeds, browser speeds, processor speeds, and keyboard and mouse clicking speeds of subjects recruited through AMT, we follow Crump and colleagues in assuming that these differences are small and randomly distributed across subjects. Crump and colleagues confirm this assumption in a widely cited study that used subjects recruited through AMT to replicate outcomes in four types of common laboratory reaction time studies. See Matthew J. C. Crump, John V. McDonnell, Todd M. Gureckis, Evaluating Amazon’s Mechanical Turk as a Tool for Experimental Behavioral Research, PLoS ONE 8(3) (2013).

Another potential disadvantage is that subjects recruited online cannot be monitored to ensure that they are following instructions or remaining attentive to the experiment. One instruction we feared might be broken was the rule that subjects use a laptop or desktop computer with a mouse, not a touchscreen device like a tablet or smartphone. To ensure that we were only measuring reaction times of people using a mouse, we applied a setting in Qualtrics that prohibited people with touchscreen mobile operating systems from participating in the experiment. To make sure subjects were paying attention to the experiment, we examined the data closely for invalid responses. We omitted a small number of subjects during the data cleaning stage, either because they typed gibberish on open-ended questions in our distractor task or because their response times were greater than 30 seconds, which signaled distraction. It is worth noting that these excessively large response times constituted less than 0.02 percent of all observations in our final data set. Furthermore, 99 percent of subjects responded to the matching questions in 3.5 seconds or less, which indicates that the vast majority of subjects were focused and engaged. 39 CITES

11

Because our objective initially was to test for the effects of context, after looking at the ads, subjects in one of the treatment groups participated in a brief car-buying role-play task—i.e., a task related to the targeted brands and intended to contextualize the matching task. These subjects were asked to write about the steps they would take in the process of purchasing a car, to think about how many dealerships they would visit, and to rank various features such as speed, roominess, and affordability. Subjects in the control group and the other treatment group performed a similar task after seeing the ads, but instead of being asked to imagine the process of buying a car, they were asked to imagine the process of searching for a job—a process unrelated to the targeted brands. In addition to allowing us the opportunity to expose some subjects to purchasing context, these exercises served as distractors between the ads and the response time portion of the study.

In the response time portion of the protocol, all subjects were informed that on the following screens they would see a brand name and a word. They were told to click a “Yes” button if the brand and the word matched and click a “No” button if they did not. Subjects were asked to click as quickly as possible without sacrificing accuracy.

Subjects were then shown thirty brand-word pairs. Twenty-five of these were “filler” pairs for brands such as MCDONALD’S, CHASE, COCA COLA, PERRIER, MICROSOFT, and UNITED. Mixed in with the filler pairs were pairs for one of our target brands. Subjects in the MERCEDES group, for instance, saw MERCEDES-CARS, MERCEDES-WEALTH, MERCEDES-LUXURY, MERCEDES-PRINTERS, and MERCEDES-CHEAP mixed in with the twenty-five filler pairs. Subjects in the INFINITI group saw INFINITI-CARS, INFINITI-FAST, INFINITI-LUXURY, INFINITI-PRINTERS, and INFINITI-CHEAP mixed in with the twenty-five filler pairs. Through Qualtrics, we were able to measure the time it took each subject to click “Yes” or “No,” down to one one-thousandth of a second.

When we compared average response times, our results for the MERCEDES group matched what others have found before: subjects who saw a diluting MERCEDES Toothpaste ad were slightly slower (by about two-tenths of a second) to click “Yes” when the pair MERCEDES-CARS flashed onto the screen. That delay was statistically significant (p < 0.05).

But we also noticed something that we did not expect. In addition to a statistically significant difference in average response time between treatment and control on the MERCEDES-CARS pair, we noticed statistically significant differences between control and treatment groups on many of the filler pairs. Subjects in the treatment groups, for example, were slightly slower to click “Yes” or “No” when pairs like MCDONALD’S-HAMBURGERS, MICROSOFT-RELIABLE, MICROSOFT-SOFTWARE, COCA COLA-EXPENSIVE, CHASE-MONEY, and eleven other pairs appeared on screen.

The fact that our MERCEDES Toothpaste ad was associated with slower response times not only on the MERCEDES-CARS pair, but in a total of sixteen categories unrelated to the MERCEDES brand, gave us pause and caused us to question what these response time experiments are actually measuring. Were we measuring dilution of the MERCEDES and

12

INFINITI brands? Or was the diluting ad so strange to subjects that it surprised them and caused made them suspect that experimenters were trying to trick them?

2) Pilot #2

Based on what we saw in the pilot data, we hypothesized that a surprising stimulus could cause subjects in the treatment groups to proceed through the experiment with caution, resulting in slower response times in general. If this were the case, differences in response time between treatment and control groups would not be a sign of dilution, but rather an artifact of an experimental design that fails to properly control for the effects of being surprised. Accordingly, we designed another pilot to try to identify this potential source of methodological error.

In a second pilot of 100 subjects, we used a revised instrument that attempted to control for surprise. The treatment groups were the same as in the first pilot. Now, however, our control group subjects would see an ad for a fictitious brand called NIKE Toothpaste in addition to the three filler ads. Our logic was that if our MERCEDES/INFINITI Toothpaste ads were surprising subjects and causing them to slow down on the matching exercise, that effect could be offset by including an ad that was just as surprising in the control group. Statistically significant differences between treatment groups and the control group, in other words, should go away if a surprise effect is at work.

When we compared average response times for the treatment and control groups in Pilot #2, we found that this is exactly what happened. In our second pilot there were virtually no statistically significant differences in response times for the MERCEDES pairs, the INFINITI pairs, or any of the filler pairs.

The results of our two pilots caused us to question the validity of using traditional response time experiments to measure dilution. Whereas before we assumed that response time experiments were a valid way to measure dilution, we now hypothesized that statistically significant results in the prior MJ and PSN experiments could have been due to failure to control for the surprise effect caused by the supposedly diluting ad. We then designed an experiment to more formally test the extent to which response time experiments measure dilution versus surprise.

3) Redesigned Experiment

In our redesigned experiment (N = 1800), we randomly assigned subjects to a MERCEDES group or an INFINITI group. For each target brand, subjects were randomly assigned to one of three groups: a treatment group (T) or one of two control groups (C1 and C2). Subjects in group T saw the three filler ads plus a MERCEDES or INFINITI diluting ad. Subjects in group C1 saw only the three filler ads and subjects in group C2 saw the three filler ads plus a NIKE Toothpaste ad. After viewing the ads, subjects participated in the job search role play task, as a distractor. They then went on to answer the same thirty matching questions we presented in our pilot studies.

13

Our analyses were guided by two possible theories. Theory 1 states that response time experiments are successful in measuring dilution when the mean response time of group T is greater than the mean response time of both group C1 and group C2 for the MERCEDES and INFINITI pairs and for those pairs only, realizing that there may be one or two false positives for other pairs due to chance. In other words, the average treatment effect (ATE), measured as the treatment response time minus the control response time for a given pair, should be positive and statistically significant even when we control for surprise and only for the target brand-product category pairs.

Theory 2 states that response time experiments measure a surprise effect rather than dilution if we find that group T’s mean response time is greater than group C1’s mean response time and not statistically different from group C2’s mean response time. The ATE, in other words, should be positive and statistically significant when comparing T and C1 and not statistically different from zero when comparing T and C2. The ATE for the T and C1 comparison should furthermore be positive and statistically significant in the case of multiple brand pairs—not only the MERCEDES or INFINITI pairs.

Tables 1A and 1B present average treatment effects for subjects assigned to receive MERCEDES questions. Table 1A presents ATEs in a study where the treatment group saw the three filler ads plus the MERCEDES Toothpaste ad and the control group only saw the three filler ads. In the first row, we see that there is no statistically significant difference between treatment and control groups on the product category pair MERCEDES-CARS, which indicates that the diluting stimulus was not causing subjects to disassociate the brand MERCEDES from its true product category. We do see, however, what could be interpreted as a form of brand attribute dilution on the MERCEDES-LUXURY pair. On average, subjects who saw the MERCEDES Toothpaste ad took 0.075 seconds longer than subjects in the control group to determine whether MERCEDES and LUXURY are a pair (p < 0.05). Although this difference is statistically significant, it is very small. Furthermore, no other differences for the target MERCEDES brand are statistically significant from zero.

14

Table 1A. Response Time Experiments (Target Brand: Mercedes). No Control for Surprise. Average Treatment Effects.

Treatment Group: Mercedes Toothpaste Ad Control Group: No 4thAd

Target Brand Mercedes-

Luxury Mercedes-

Printers Mercedes-

Cars Mercedes-

Cheap Mercedes-

Wealth 0.075** 0.023 -0.046 0.092 0.063 (0.041) (0.053) (0.105) (0.079) (0.066)

Filler Brands Chase-Money

Chase-Restaurants

Coca Cola-Expensive

Coca Cola-Soft Drinks

Coca Cola-Sour

0.025 0.037 -0.005 0.007 0.098** (0.036) (0.036) (0.059) (0.034) (0.059)

Coca Cola-Toothpaste

McDonald's-Affordable

McDonald's-Fries

McDonald's-Hamburgers

McDonald's-Healthy

0.064 -0.008 0.067 0.060 0.042 (0.045) (0.041) (0.052) (0.059) (0.040)

McDonald's-

Sunset Microsoft-Bicycles

Microsoft-Computers

Microsoft-Reliable

Microsoft-Software

0.020 -0.038 0.164** -0.012 0.024 (0.064) (0.057) (0.077) (0.098) (0.061)

Microsoft-

Luxury Perrier-Bottle Perrier-

Clothing Perrier-Pure Perrier-Soft

-0.021 -0.121 -0.071 0.041 -0.039 (0.058) (0.105) (0.067) (0.088) (0.077)

Perrier-Water United-Boxes United-Horses United-Small United-Trucks

0.019 -0.010 0.111 -0.077 0.080 (0.068) (0.070) (0.075) (0.087) (0.082)

Standard errors in parentheses. One tailed test: * p < 0.10, ** p < 0.05, *** p < 0.01

15

We also see in Table 1A that the diluting stimulus is associated with slower response times on two unrelated brands. Treatment group participants, for example, were slower by 0.098 seconds to react to the COCA COLA-SOUR pair and by 0.164 to the MICROSOFT-COMPUTERS pair. This signals that the MERCEDES Toothpaste ad may have been surprising to subjects and slowed them down generally.

This intuition is supported by the results presented in Table 1B. Table 1B presents average treatment effects in a study where the treatment group saw three filler ads plus a MERCEDES Toothpaste ad and the control group saw three filler ads plus a NIKE Toothpaste ad. Here we can see that after controlling for surprise, the statistically significant difference we saw in the MERCEDES-LUXURY pair goes away and that there are no statistically significant response time differences for the MERCEDES-CAR pair or any other target-brand pairs. Furthermore, statistically significant differences in response time on the COCA COLA-SOUR and MICROSOFT-COMPUTERS pair go away as well. Only one statistically significant response time difference remains (MCDONALD’S-HAMBURGERS), which could be due to chance.

16

Table 1B. Response Time Experiments (Target Brand: Mercedes). Controlling for Surprise. Average Treatment Effects.

Treatment Group: Mercedes Toothpaste Ad Control Group: Nike Toothpaste Ad

Target Brand Mercedes-

Luxury Mercedes-

Printers Mercedes-

Cars Mercedes-

Cheap Mercedes-

Wealth -0.026 -0.096 0.022 0.047 0.083 (0.096) (0.129) (0.037) (0.103) (0.065)

Filler Brands Chase-Money Chase-

Restaurants Coca Cola-Expensive


Coca Cola-Sour

-0.014 0.029 -0.098 -0.066 0.032 (0.048) (0.036) (0.073) (0.040) (0.062)



McDonald's-Fries


McDonald's-Healthy

0.060 0.006 0.015 0.117** -0.017 (0.042) (0.036) (0.073) (0.050) (0.042)

McDonald's-


Microsoft-Computers

Microsoft-Reliable

Microsoft-Software

0.049 -0.079 0.104 -0.010 0.036 (0.056) (0.091) (0.079) (0.060) (0.050)

Microsoft-

Luxury Perrier-Bottle Perrier-Clothing Perrier-Pure Perrier-Soft

-0.083 0.025 0.017 0.120 -0.033 (0.058) (0.050) (0.046) (0.077) (0.069)


0.091 0.023 0.029 -0.012 0.021 (0.065) (0.062) (0.080) (0.066) (0.089)


17

Taken together, Tables 1A and 1B support the theory that response time experiments, as they have been conducted previously, are measuring surprise as opposed to dilution. When we did not control for surprise, we see a number of statistically significant differences between the treatment and control groups. When we control for surprise, these differences largely go away.

Tables 2A and 2B provide further support for the hypothesis that response time experiments measure surprise rather than dilution. Table 2A shows average treatment effects for subjects assigned to receive INFINITI questions. Subjects in the treatment group saw the three filler ads plus an INFINITI Toothpaste ad. Subjects in the control group saw only the three filler ads. In this study we see a difference of 0.131 seconds in the mean response time of the treatment group and the control group on the INFINITI-CHEAP pair. People who saw the INFINITI Toothpaste ad, in other words, were slightly slower to respond to this question compared to people who only saw the three filler ads. This could again be interpreted as product attribute dilution. As we were in the case of the MERCEDES questions, however, we are skeptical because we do not observe a statistically significant difference between treatment and control on the product category pair, INFINITI-CARS. At the same time, we do observe statistically significant differences in mean response times for a number of other pairs unrelated to INFINITI. Treatment group subjects were also slightly slower to respond when shown the CHASE-RESTAURANTS, MCDONALD’S-HAMBURGERS, MCDONALD’S-SUNSET, PERRIER-CLOTHING, and PERRIER-WATER pairs.

18

Table 2A. Response Time Experiments (Target Brand: Infiniti). No Control for Surprise. Average Treatment Effects.

Treatment Group: Infiniti Toothpaste Ad Control Group: No 4thAd

Target Brand Infiniti-Luxury Infiniti-Printers Infiniti-Cars Infiniti-Fast Infiniti-Cheap

0.029 -0.078 0.125 0.108 0.131** (0.080) (0.078) (0.096) (0.110) (0.053)

Filler Brands Chase-Money Chase-

Restaurants Coca Cola-Expensive


Coca Cola-Sour

-0.022 0.121** 0.043 -0.023 0.024 (0.033) (0.066) (0.045) (0.045) (0.035)



McDonald's-Fries


McDonald's-Healthy

0.080 -0.038 -0.003 0.072** 0.038 (0.117) (0.043) (0.038) (0.036) (0.038)

McDonald's-


Microsoft-Computers

Microsoft-Reliable

Microsoft-Software

0.120** -0.013 -0.012 -0.099 0.031 (0.053) (0.061) (0.040) (0.084) (0.039)

Microsoft-


0.065 0.039 0.231** -0.013 -0.024 (0.056) (0.061) (0.137) (0.067) (0.098)


0.083** 0.226 0.045 -0.073 0.043 (0.035) (0.161) (0.075) (0.073) (0.061)


19

Table 2B shows what happens when we control for surprise with respect to the INFINITI mark: differences in mean response times between treatment and control groups evaporate.

Table 2B. Response Time Experiments (Target Brand: Infiniti). Controlling for Surprise. Average Treatment Effects.

Treatment Group: Infiniti Toothpaste Ad Control Group: Nike Toothpaste Ad

Target Brand Infiniti-Luxury Infiniti-Printers Infiniti-Cars Infiniti-Fast Infiniti-Cheap

0.021 -0.014 -0.043 0.079 -0.043 (0.054) (0.055) (0.141) (0.101) (0.107)

Filler Brands Chase-Money

Chase-Restaurants

Coca Cola-Expensive


Coca Cola-Sour

-0.046 0.034 -0.068 -0.071 -0.045 (0.033) (0.081) (0.047) (0.051) (0.038)



McDonald's-Fries


McDonald's-Healthy

0.023 -0.080 -0.020 0.037 -0.045 (0.107) (0.044) (0.036) (0.037) (0.043)

McDonald's-


Microsoft-Computers

Microsoft-Reliable

Microsoft-Software

-0.045 -0.158 -0.047 -0.063 -0.060 (0.066) (0.119) (0.036) (0.052) (0.044)

Microsoft-


-0.032 0.002 0.216 -0.100 -0.094 (0.064) (0.063) (0.137) (0.076) (0.089)


-0.018 0.081 -0.034 -0.110 -0.201 (0.041) (0.172) (0.076) (0.082) (0.121)


20

On the basis of our two pilots and subsequent large-sample study, we conclude that response time studies, at least as they have traditionally been conducted in the literature, are ill suited to measure the concept of trademark dilution. We find that once we properly control for the surprise effect of seeing a strange ad, differences in response time between treatment and control group subjects are essentially eliminated.

C. Category Recall Tests

Next we explored whether the MERCEDES Toothpaste and INFINITI Toothpaste ads had any impact on subjects’ ability to recall the true product category of the MERCEDES or INFINITI brand. If the MERCEDES Toothpaste or INFINITI Toothpaste ads are causing dilution, we would expect fewer subjects in the treatment groups to identify MERCEDES-CARS or INFINITI-CARS as a match compared to subjects in the control groups.

Table 3 shows average treatment effects for the category recall tests. The negative signs on the coefficients indicate that treatment group subjects were less likely to state that MERCEDES-CARS or INFINITI-CARS is a match. These differences, however, were generally quite small and statistically significant in just half of cases. Panel A shows comparisons between treatment and control subjects when the control group saw no fourth ad. The difference between treatment and control subjects on the MERCEDES-CARS exercise was tiny and not statistically different from zero. In fact, 97.8 percent of subjects in the control group said these words were a match compared to 97.2 percent of subjects in the treatment group for a difference of 0.06 percent, or 0.006. The difference between treatment and control subjects on the INFINITI-CARS exercise is slightly larger and marginally statistically significant (p=0.055). Control group subjects identified these words as a match in 93.3 percent of cases compared to treatment subjects, who identified the words as a match in 89 percent of cases. Turning to Panel B, however, we see that the differences in the INFINITI-CARS column do not hold up when the control group saw a NIKE Toothpaste Ad. Control group subjects said INFINITI-CARS is a match in 90.6 percent of cases compared to 89 percent of cases among treatment group subjects, a difference that is not statistically different from zero. In the MERCEDES-CARS column in Panel B, on the other hand, we now find that there is statistically significant difference between treatment and control groups when the control group sees a NIKE Toothpaste ad. Control group subjects said these words were a match in 99.5 percent of cases compared to 97.2 percent of cases in the treatment group.

21

Table 3. Category Recall Experiment. Average Treatment Effects. A.

Treatment Group: Mercedes/Infiniti Toothpaste Ad

Control Group: No 4th Ad

Mercedes-Cars Infiniti-Cars

-0.006 -0.043*

(0.015) (0.027)

B. Treatment Group: Mercedes/Infiniti Toothpaste

Ad Control Group: Nike Ad

Mercedes-Cars Infiniti-Cars

-0.023** -0.016

(0.012) (0.028)


Taken together, the results presented in Table 3 show that subjects who saw a MERCEDES Toothpaste or INFINITI Toothpaste ad were always less accurate in tests that asked them to recall the primary product category of the MERCEDES or INFINITI brand. At the very least, this test indicates that seeing an ad for a familiar brand name in an unexpected product category can cause a small number of people to dissociate a brand name from its primary product category. This test cannot tell us, however, whether what we are observing is dilution or the fact that a small number of subjects became confused after seeing the toothpaste ads—that is, they thought that the association we had actually presented to them (MERCEDES/Toothpaste; INFINITI/Toothpaste) was the product category association we were asking about. If the dissociation we are observing is a sign of dilution, rather than arising from confusion that is itself an artifact of our experimental design, we should also observe that the toothpaste ads have damaging effects on the mark. The remaining experiments explore this possibility by testing the effects of the toothpaste ads on the selling power of the MERCEDES and INFINITI marks.

22

D. Preference Ordering Study

Next we ran an additional experiment to see if the potentially diluting ads had any effect on consumer preferences (N = 1008). If a trademark is diluted, the weakening of desirable associations would likely cause consumers to shift their preferences among brands in the same or a similar product category. Our preference ordering protocol was designed to test this possibility.

Subjects were first shown up to twenty car brands in random order and asked to state whether or not the brand was familiar. When five familiar brands were selected, respondents moved to the next stage of the experiment. Half of subjects who said they were familiar with either MERCEDES or INFINITI were shown a MERCEDES Toothpaste or INFINITI Toothpaste ad in addition to the three filler ads (treatment group). The other half were shown a NIKE Toothpaste ad in addition to the three filler ads (control group).

After viewing the ads, subjects were asked to rank the five brands they had identified as familiar in order of preference, from 1 (favorite brand) to 5 (least favorite). Overall, the mean rank for MERCEDES was 2.31 and mean rank for INFINITI was 2.89.

Table 4 shows the differences between the mean preference ranking for treatment group subjects and mean preference ranking for control group subjects for each brand. In both cases, the differences are not statistically different from zero. We are unable to find any evidence from this protocol that the diluting ad causes subjects to rank MERCEDES or INFINITI lower.

Table 4. Buying Preference Experiment. Average Treatment Effects. Treatment Group: Mercedes/Infiniti Toothpaste

Ad Control Group: Nike Ad

Mercedes Infiniti

0.182 -0.038

(0.169)

(0.242)


E. Likert Association Strength Study

In a final experiment, we tested whether diluting advertisements affected the strength of the linkage between the MERCEDES and INFINITI brands and both their product category (cars) and the top associations (wealth, luxury) we previously recorded for each brand. Marketers distinguish their brands through a set of brand associations; i.e. qualities that a consumer associates with a specific brand. If a trademark is diluted, the weakening of product category and product attribute associations would diminish the selling power of the mark.

23

In this protocol, treatment group subjects saw the three filler ads plus a MERCEDES or INFINITI Toothpaste ad and control group subjects saw only the three filler ads. Subjects were then shown twenty-three brand-word pairs. Eighteen of these pair were filler pairs for brands like UNITED, CHASE, NIKE, and COCA COLA. Randomly distributed among the filler pairs were five brand-word pairs that involved either MERCEDES or INFINITI. The word that was paired with the brand came from the brand association study described at the outset of this paper.

At the outset of the exercise, subjects were told to state the degree to which they associate the brand name and the word on a five-point scale, where a score of 1 means they associate the brand and the word “a great deal” and a score of 5 means they associate the brand and the word “not at all.” If the MERCEDES and INFINITI Toothpaste ads are causing dilution, we would expect that subjects in the treatment groups would report weaker associations between the car brands and the product category “cars”, as well as words that our pretests indicated to be strongly associated with MERCEDES and INFINITI, such as “luxury” and “wealth.”

Results presented in Table 5 indicate, however, that the MERCEDES Toothpaste and INFINITI Toothpaste ads do not erode the strength of associations between the MERCEDES and INFINITI brands and words like luxury, wealth, and cars. Subjects who saw the MERCEDES Toothpaste and INFINITI Toothpaste ads were furthermore no more likely than control group subjects to associate these brands with the word “cheap.” Treatment and control subjects were different on one pair, however: Subjects who saw a MERCEDES Toothpaste or INFINITI Toothpaste ad were somewhat more likely to associate these car brands with the product category “toothpaste.” The difference is 0.369 of a point on a 5-point scale in the case of MERCEDES and 0.73 of a point on a 5-point scale in the case of INFINITI. These differences remained after we controlled for subjects’ stated level of familiarity with the MERCEDES and INFINITI brands, which we asked about at the end of the survey.

What do these differences mean? In practice, almost all control group subjects (94 percent and 95.6 percent) said they do not at all associate the INFINITI or MERCEDES brand with the word toothpaste. The result was much different for subjects in the treatment groups. In the INFINITI treatment group, 54.8 percent of subjects said they do not at all associate INFINITI with the word toothpaste, whereas 23.5 percent said they associate these words “a little,” 9.2 percent “a moderate amount,” 6.2 percent “a lot,” and 6.2 percent “a great deal.” In the MERCEDES treatment group, 73.3 percent of subjects said they do not at all associate the words MERCEDES and toothpaste, whereas 15.8 percent associate them “a little,” 5 percent “a moderate amount,” 2.8 percent “a lot,” and 3.2 percent “a great deal.” These differences are obviously due to the fact that treatment group subjects saw an ad moments earlier that associates these brands with the product category toothpaste. That exposure to the MERCEDES Toothpaste or INFINITI Toothpaste ad did not cause subjects to dissociate these brands from words like cars, wealth, and luxury suggests that the ads do not cause dilution as measured through Likert scales; however, the fact that these ads cause subjects to associate the brands with toothpaste indicates that the ads are contributing to the formation of a new association between the targeted brands and a different product category. But we see no evidence from this protocol that the

24

creation of this new association is accompanied by the dilution or weakening of pre-existing associations, either with respect to product category or product attribute.

Table 5. Likert Association Strength Tests. Average Treatment Effects.

Treatment Group: Mercedes/Infiniti Toothpaste Ad Control Group: No 4th Ad

Mercedes-Luxury

Mercedes-Wealth

Mercedes-Cars

Mercedes-Toothpaste

Mercedes-Cheap

0.052 0.019 0.040 -0.369*** -0.016 (0.050) (0.059) (0.040) (0.048) (0.032)

Infiniti-Luxury

Infiniti-Wealth Infiniti-Cars

Infiniti-Toothpaste

Infiniti-Cheap

-0.009 0.124 0.055 -0.730*** -0.029 (0.078) (0.081) (0.075) (0.058) (0.043)

Standard Errors in parentheses. Two tailed test: * p < 0.10, ** p < 0.05, *** p < 0.01

III. DISCUSSION

Trademark dilution involves situations in which an entity other than a trademark’s original owner uses that trademark (e.g., a brand name or logo) in a way that weakens the strength and uniqueness of the brand associations that consumers hold to the original brand. The trademark law proposes two ways in which dilution can happen. It can occur either because use of a similar mark either “impaired the distinctiveness” of a famous and senior mark, or because such junior use “harmed the reputation” of the famous and senior mark.40 These are commonly referred to, respectively, as “dilution by blurring” and “dilution by tarnishment.”41 Our interest in this Article has been in the former.

The best evidence the literature has found on the existence of dilution by blurring lies in a set of studies that show that the speed with which individuals in an experimental setting make brand association (i.e. category membership and attribute possession) matches slows after a respondent is exposed to an advertisement for a different product using the same or a similar brand name.42 Jacoby has singled out response time as being a particularly significant indicator

40 15 U.S.C. § 1125(c). 41 Id. 42Morrin and Jacoby 2000; Pullig et al. 2006. Morrin and Jacoby (2000) present some results on recognition accuracy in addition to reaction time. However, they do not detail the associations that they examine apart from brand category (e.g. GODIVA chocolate). As such, it is difficult to know what

25

of dilution by blurring. He refers to the reduction in the speed at which people access, recognize, and verify brand associations as “the essence of a weakening of associations” for dilution.43 Armed with the belief that differences in response times is a reliable and valid measure of trademark dilution along with the knowledge that external validity (i.e. generalizability to a marketplace context) is critical for relevance in litigation,44 we set out to investigate whether contextualization of consumer judgments would mitigate the differences in response times found in prior studies. In other words, if we put respondents “in a buying mood”,45 would the apparent dilution seen in prior studies vanish? In the first pilot studies we ran, we did find an effect that appeared to be dilution. More specifically, people who saw a MERCEDES-diluting stimulus were slower to say MERCEDES and Cars are a match and slower to say MERCEDES and Luxury are a match. There were no such differences when people saw an INFINITI-diluting stimulus. This stands in contrast to the MJ’s finding that familiar brands suffered less than unfamiliar brands after exposure to a potentially diluting stimulus. The interesting thing though is that is not all we found. The diluting ads purported to be for MERCEDES Toothpaste and INFINITI Toothpaste. However, we found more significant increases in response time than would have been expected by chance alone for distractor matching tasks that could not have possibly been caused by a “dilution” effect of the ads. Examples included MCDONALD’S-Hamburgers, MCDONALD’S-Sunset, PERRIER-Pure, PERRIER-Water, CHASE-Restaurants, and others that could not possibly have been the result of dilution of the MERCEDES and/or INFINITI trademarks by our ads.

While no one had contemplated a surprise effect for such incongruent ads in the trademark literature, such an effect is consistent with prior work on response times and matching tasks. The psychology literature is replete with studies of perceptual and cognitive matching tasks.46 Such tasks generally involve either simultaneous or sequential presentation of stimuli (e.g. letters, numbers, words) that subjects judge to be the same or different according to criteria set by the experimenter. Consistent with the view emerging from the trademark literature,47 the dependent variable most commonly measured is the time to respond to each stimulus pair. One branch of this very extensive literature involves priming. In these studies a priming stimulus, to which the subject generally does not have to react, is presented prior to another stimulus to which the subject generally does have to react. A neutral prime condition serves as a baseline against which a particular type of prime produces facilitation or inhibition in processing

“accuracy” means. Finally, the numbers presented in their Table 1 (top panel) are inconsistent. Unlike the bottom panel of the table, the mean row is not equal to the mean of the rows. As such, it is difficult to interpret or accept the evidence they allege. 43 Jacob Jacoby, The Psychological Foundations of Trademark Law: Secondary Meaning, Genericism, Fame, Confusion, and Dilution, 91 Tʀᴀᴅᴇᴍᴀʀᴋ Rᴇᴘ., 1013 (Sept.-Oct. 2001). 44 6 McCarthy § 32:163, Steckel 2017 [need cite still]. 45 6 McCarthy § 32:163. 46 Robert W. Proctor, A Unified Theory for Matching-Task Phenomena, 88 Pꜱʏᴄʜᴏʟ. Rᴇᴠ., (July 1981. 47 cf. Jacoby 2001

26

a test stimulus. The studies reported in this paper as well as in the prior trademark dilution literature reviewed earlier all fall into this category. Delays in response time are often evident when the prime is unrelated to the subsequent stimulus, a concept termed prime validity.48 The general consensus in this literature is the delay can be attributed to both automatic and attentional components. In other words, some effects occur automatically as a result of processing the prime, and others involve the subject’s conscious expectations.49 For example, in one category matching study Neely (1977) asked subjects to match a target word to a second possibly related word. Subjects were primed to expect an exemplar of one of the categories prior to each word matching trial. For example, if the prime were BIRD, the target might be “robin”. The prime categories were BIRD, BODY, BUILDING, or XXX (i.e. a nonword). The target word was presented with a second possibly related word. Subjects’ task was to indicate whether or not the target word was related to the second word. On most trials (i.e., two-thirds of them), the target word corresponded to the category the prime was designed to lead the subject to expect. However, on a number of trials (i.e., one-third of them), the target word was not from the category the prime was designed to lead the subject to expect, thereby creating two conditions, expected and unexpected. XXX served as a control prime (i.e. there were no expectations so there could be no surprise.). Results showed that response times were lower in the expected condition than in the XXX control and greater in the unexpected condition than in the XXX control. Our study, as well as those of MJ and PSN, fit right into this paradigm and the results are quite consistent. The degree to which a target conforms to a subject’s expectations induced by the prime impacts the response time of the subject in a subsequent matching task. Or, for our purposes, surprise slows response times. So where does that leave us? The research reported here casts serious doubt on the validity of the best evidence we have that trademark dilution actually exists. What has been reported as (and assumed to be) evidence of trademark dilution may actually be merely the result of nothing more than the subjects being surprised by seeing ads such as those for MERCEDES and INFINITI toothpaste employed in our studies. We are forced to reconsider the question in the title of this paper. Is dilution the unicorn? Like the unicorn, the concept is plausible and one can easily envision it existing. The U.S. Congress along with thousands of brand managers believes in its existence. So we should not let this research signal the end of our hunt for it, no matter how difficult it may be to find.

48 J.H. Neeley, Semantic Priming and Retrieval from Lexical Memory: Roles in Inhibitionless Spreading Activation and Limited Capacity Attention, 106 Jᴏᴜʀɴᴀʟ ᴏꜰ Exᴘᴇʀɪᴍᴇɴᴛᴀʟ Pꜱʏᴄʜᴏʟ.: Gᴇɴ., 226 (1977); M.I. Posner, M.I. & C.R.R. Snyder, Facilitation and Inhibition in the Processing of Signals, in ATTENTION AND PERFORMANCE V, (P.M.A. Rabbitt & S. Dornic eds., 1977); D.A. Taylor, D.A, Time Course of Context Effects, 106 J. EXPERIMENTAL PSYCH.: Ge. 404 (1977). 49 I. Fischler, Associative Facilitation Without Expectancy in a Lexical Decision Task, 3 J. Exᴘᴇʀɪᴍᴇɴᴛᴀʟ Pꜱʏᴄʜ.: Hᴜᴍ. Pᴇʀᴄᴇᴘᴛɪᴏɴ ᴀɴᴅ Pᴇʀꜰᴏʀᴍᴀɴᴄᴇ 18 (1977); Proctor (1981) op. cit.

27

One thing for sure is that current response time methods do not provide an appropriate microscope through which lens we can investigate the existence of our unicorn. Such a development should direct us back to the starting line and reconsider what a trademark is supposed to do. In a landmark textbook, Keller describes the building blocks of a brand’s power, brand awareness and brand associations.50 Professor Keller characterizes the associations that contribute to building positive brand equity as strong, favorable, and unique. Strength relates to the clarity of a brand’s image both in a specific consumer’s mind and across consumers in the marketplace. To benefit the firm, the fact that strong brands need to be based on favorable associations is self-evident. Marketing professionals work diligently using a variety of tools and programs (e.g., product development, pricing, advertising, naming, signing, etc.) to create favorable associations to distinguish their brands. These associations are in turn embodied by and signaled to consumers through trademarks. With that in mind, true harm to a brand as a result of trademark dilution would have to result through a favorable association that becomes weaker or an unfavorable association that has emerged.

It is our view that such strength cannot be measured by response times as prior research has attempted. We believe that Likert scales that measure how much (and not how quickly) an attribute or characteristic is associated with a trademark is a better starting point. Of course, preference or choice is important as well. The ultimate test of trademark dilution is whether it costs a firm any sales. Our work was not able to produce changes in consumer perceptions of critical associations or choice, despite demonstrable differences in response times. Let the hunt for the unicorn continue.

50 Kevin Keller, Sᴛʀᴀᴛᴇɢɪᴄ Bʀᴀɴᴅ Mᴀɴᴀɢᴇᴍᴇɴᴛ: Bᴜɪʟᴅɪɴɢ, Mᴇᴀꜱᴜʀɪɴɢ, ᴀɴᴅ Mᴀɴᴀɢɪɴɢ Bʀᴀɴᴅ Eǫᴜɪᴛʏ 49 (4th ed., 2010).

is trademark dilution a unicorn duke · is trademark dilution a unicorn? an experimental...

Documents