Antecedentes de la propuesta

(CONSIDERING RESPONDENTS’ “REASONS WHY”)

TEST Control

1 Control2 Control3 AVERAGES T+C2 C1+C3

Name typed in to

initiate search GEICO GEICO GEICO NIKE

Base = (101) (50) (47) (55) (148) (105)

% % % % % %

Q8. Do you think [link] . . . has been authorized by GEICO to offer GEICO

insurance?

GEICO ALLSTATE GEICO GEICO

Free Insurance Quotes

www.Autoand Home.com)

17.8 8.0 14.9 12.7 16.9 10.5

Shop for Auto Insurance (www.insureCom. com) 21.8 4.0 12.8 14.5 18.9 9.5 Insurance Finder—Save (www.insurance finder.com) 20.8 8.0 17.0 9.1 19.6 8.6 YAHOO!—GEICO Company Profile (biz.yahoo.com/ic/ 10/10616.html) NA NA 34.0 NA NA Average of the averages 18.5 9.5

As can be seen from Table 6, when consumers who think a particular Sponsored Link has been authorized by GEICO to offer GEICO insurance are asked to indicate the reasons why they think so, few (less than 10%) indicate that their confusion is due to the fact that either they had typed in “GEICO,” or they thought they were on a GEICO website/home page, or they thought the link was affiliated or connected with GEICO. This compares to approximately 34% who mistakenly thought that a non-sponsored

organic link mentioning GEICO’s name had been authorized by GEICO to offer GEICO insurance (Control Group 2).

E. Jacoby Google Survey Conclusions

Based upon the foregoing survey data and analyses, co-author Jacoby concluded as follows:

1. The sponsored links at issue in GEICO v. Google did not lead to hijacking. People who type “GEICO” on a Google start page and then generate a search results page that includes organic search results associated with the name GEICO and Sponsored Links to other auto insurance sites that do not use the name GEICO in their heading or text were not confused on how to get to GEICO’s website or home page. Contrary to GEICO’s complaint allegations, these people were not misled or hijacked by these Sponsored Links into going to other insurance sites when they intended to go to GEICO.

2. The Sponsored Links at issue in GEICO v. Google did not create initial interest confusion. Sponsored Links that did not use the name GEICO generate negligible (non-actionable) levels of confusion as to their affiliation with, authorization by, or offering products from GEICO.

Based upon the data gathered in this survey, it can be concluded that a negligible percent of consumers using Google in the marketplace were likely to be misled or confused in a manner being alleged by GEICO in its complaint.

IV. CONCLUSION

A sound foundation of the facts relating to search service practices, and consumer understandings of them, will ultimately point the way to the best resolution of trademark keyword-based search service advertisements and featured listings. The first few waves of cases have, unfortunately, not fully developed these necessary facts, and accordingly they left practitioners with incomplete conclusions and guidance for future cases. By highlighting the factual voids so far, and the areas that need fuller factual inquiry, and by providing an example of a viable full survey of consumer understandings of keyword advertising, it is hoped that this article will point the way for litigants, courts, and survey experts to build the much-needed factual bases for resolving keyword-based trademark disputes.

Sound factual development is especially important in litigation involving new technologies. In many areas of Internet and technology law, it takes time for the courts and the established legal doctrines to adjust to new technology. New developments are often seen as radical and violative of legal rights at first, but

attitudes often change once they become more prevalent and are better understood. Initially, for example, any use of a competitor’s trademark on a website, particularly in “hidden metatags,” was viewed by courts as constituting illegal trademark infringement.170

As the law developed, courts became more open to allowing uses of metatags.171 _{In the same fashion, while one may take issue with}

the analysis of the United Kingdom’s Court of Appeal in the Reed Employment case,172 _{we are likely to see more of that court’s}

approach, by viewing Internet users as accustomed to “fuzzy results,” including “much rubbish thrown in,” and hence relatively unlikely to expect close connections between the use of a search term and everything that is posted on a search service in response.

The development of Internet law in the last decade has shown that as familiarity with technology increases, courts generally become less hostile to the use of new techniques and the new technologies, and more open-minded about the possibilities of their allowable uses. The time is ripe for a full exploration of how Internet advertising techniques such as keyword-based advertising work, and of how consumers use and understand them.

170. E.g., Playboy Enters., Inc. v. AsiaFocus Int’l Inc., 1998 WL 724000, 1998 U.S. Dist. Lexis 10359 (E.D. Va. 1998) (suggesting that use of trademarks in metatags is inherently deceptive).

171. E.g., Playboy Enters., Inc. v. Welles, 279 F.3d 796 (9th Cir. 2002) (recognizing allowable nominative use of trademarks in metatags).

172. Reed Executive plc v. Reed Business Info. Ltd., 2004 EWCA (Civ) 159 (March 3, 2004).