Trademarks, Triggers, and Online Search

Published date01 December 2014
AuthorStefan Bechtold,Catherine Tucker
DOIhttp://doi.org/10.1111/jels.12054
Date01 December 2014
Trademarks, Triggers, and Online Search
Stefan Bechtold and Catherine Tucker*
Internet search engines display advertisements along with search results, providing them
with a major source of revenue. The display of ads is triggered by the use of keywords, which
are found in the searches performed by search engine users. The fact that advertisers can
buy a keyword that contains a trademark they do not own has caused controversy worldwide.
To explore the actual effects of trademark and keyword advertising policies, we exploit a
natural experiment in Europe. Following a decision by the Court of Justice of the European
Union, Google relaxed its AdWords policy in continental Europe in September 2010. After
the policy change, Google allowed advertisers to select a third party’s trademark as a keyword
to trigger the display of ads, with only a limited complaint procedure for trademark owners.
We use click-stream data from European Internet users to explore the effect this policy
change had on browsing behavior. Based on a data set of 5.38 million website visits before
and after the policy change, we find little average change. However, we present evidence that
this lack of average effect stems from an aggregation of two opposing effects. While naviga-
tional searches are less likely to lead to the trademark owner’s website, non-navigational
searches are more likely to lead to the trademark owner’s website after the policy change.
The effect of changing keyword advertising policies varies with the purpose of the consumers
using the trademark, and it is more pronounced for lesser-known trademarks. The article
points to tradeoffs trademark policy is facing beyond consumer confusion. More generally,
the article proposes a novel way of analyzing the effect of different allocations of property
rights in intellectual property law.
I. Introduction
Since the commercialization of the Internet in the early 1990s, electronic communi-
cation networks have led to significant changes in value chains. Traditionally strong
intermediaries—such as record companies, publishers, newspapers, or movie companies—
*Address correspondence to Stefan Bechtold, Professor of Intellectual Property, Center for Law & Economics, ETH
Zurich, IFW E 47.2, 8092 Zurich, Switzerland; email: sbechtold@ethz.ch. Tucker is Associate Professor at MIT Sloan
School of Management, Massachusetts Institute of Technology, Cambridge, MA, and Research Associate at NBER.
The authors thank two anonymous referees, Sofia Amaral Garcia, Oren Bar-Gill, Lionel Bently, Daniel Chen,
Graeme Dinwoodie, Christoph Engel, Amit Gandhi, Eric Goldman, Andrea Günster, Daniel Ho, Annette Kur, Mark
Lemley, Marc McKenna, Amanda Meyer, Abishek Nagaraj, Axel Ockenfels, Lisa Ouellette, Martin Schonger, Chris-
topher Sprigman, Rebecca Tushnet, Achim Wambach, Joel Waldfogel, participants at the American Law and Eco-
nomics Association meeting (2013), the Conference on Empirical Legal Studies (2013), the Annual Conference on
Internet Search and Innovation at Northwestern University (2013), and the ETH Zurich/MPI workshop on the Law
& Economics of Intellectual Property and Competition Law (2013), as well as seminar audiences at Oxford, Zurich,
and the U.S. Patent and Trademark Office for helpful discussions. Damian George, Aurelia Tamò, and Miriam Tinner
provided helpful research assistance.
The copyright line for this article was changed on 30 September 2015 after original online publication.
bs_bs_banner
Journal of Empirical Legal Studies
Volume 11, Issue 4, 718–750, December 2014
© 2014 The Authors. Journal of Empirical Legal Studies published by Cornell Law School and Wiley Periodicals, Inc.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited.
718
have been struggling to define their future roles and find profitable business models
in a radically changed environment of content consumption. At the same time, new
intermediaries—such as search engines, auction sites, or social networks—have emerged.
In some cases, they threaten to displace traditional intermediaries. In other cases, they
complement them or create entirely new business models.
In recent years, the discussion on intellectual property rules for intermediaries has
increasingly focused on Internet search engines. In particular, a vigorous debate focuses
on whether keyword-based advertising violates trademark law. In Google’s version of the
system—called Google AdWords—advertisers can buy advertising links in the “sponsored
links” section of a Google search results page. Thereby, advertisers purchase the possibility
of having their ad displayed with the search results for a particular keyword that is relevant
to their business. When a Google search user enters a search term that contains a keyword
bought by the advertiser, the ad will appear in the upper right-hand corner or on top of the
search results page. In principle, advertisers are free to select any keyword for their ad. This
becomes a legal issue, however, if an advertiser chooses a keyword that has been registered
as a trademark by another company.
Trademark owners on both sides of the Atlantic have argued that such use of a
trademark as a keyword by a third party violates trademark law and that not only the third
party, but also the intermediary search engine, can be held liable based on either primary
or secondary liability doctrines. Google has argued that it should not be liable for trade-
mark infringement in such cases, as either the third-party keyword registration does not
infringe trademark law or, even if it did, Google cannot be held liable for such infringe-
ment. As Google’s business model relies extensively on the ad auction mechanism under-
lying AdWords,1it has a vital interest in not becoming involved in trademark-related
disputes between trademark owners and third-party advertisers.
One important dimension to whether third parties should be allowed to register
trademarks as keywords is the effect such use has on consumer behavior. On the one hand,
it could be that consumers become confused by ads based on third-party keyword registra-
tions because they assume that such ads originate from or are sponsored by the trademark
owner. One of the policy justifications for trademark protection is to overcome information
asymmetries between product manufacturers or service providers and their customers
(Landes & Posner 1987). On the other hand, it could be that consumers realize that an ad
based on a third-party keyword registration is not linked to the trademark owner, and that
they appreciate the increased information and competition resulting from such keyword
use. From a policy perspective, allowing third parties to register trademarked keywords
could increase information availability and transparency in the marketplace and could
therefore be desirable (Goldman 2009).
To shed light on this tradeoff and determine the effect of third-party keyword
advertising on consumer behavior requires an empirical investigation. Traditionally,
1Keywords are sold through auctions where advertisers bid competitively against each other for the position of the ad
on the search results page. The mechanisms underlying keyword auctions have been studied extensively (Edelman
et al. 2007; Varian 2007), but are not the focus of this article. Of Google’s $55.5 billion revenues in 2013, $50.5 billion
came from advertising (Google 2014).
Trademarks, Triggers, and Online Search 719
evidence on consumer behavior and confusion has been presented in trademark litigation
in various forms: consumer surveys, evidence of actual confusion, expert witnesses, or direct
comparison of trademarks (McCarthy 2014:§23:2:50; Bird & Steckel 2012; Sarel &
Marmorstein 2009). Such evidence reflects the traditionally passive role of the consumer:
that she uses the trademark to distinguish between products or services of different origins.
Internet search engines have substantially expanded the role trademarks play in
consumers’ search processes. First, consumers are now actively using trademarks as they
choose a string of words to query a search engine. They decide whether to use a trademark
alone to search with, or whether to combine it with other words to make the meaning and
use of the trademark more precise. Second, unlike in traditional settings, it is now easy for
firms to monitor the use of trademarks by consumers (Goldfarb 2014).
In this environment of active trademark use by both firms and consumers, more
direct evidence of consumer behavior is of particular interest. Although consumer surveys
have been used to measure consumer behavior in the context of keyword advertising
(Franklyn & Hyman 2013),2and courts have recently begun to explore advanced digital
data sources,3we propose a novel, more direct, and richer way to analyze consumer
behavior. We use micro-level click-stream data on web browsing to directly observe con-
sumer browsing behavior. Such data have been used before to explore consumer behavior
in the context of software licenses (Bakos et al. 2014); to analyze the effect of illegal music
file sharing on legal music consumption (Aguiar & Martens 2013); to document how often
search engine results are triggered by competitors’ trademarked keywords (Rosso & Jansen
2010); and aggregate summaries of such data have been used to study the implications of
a trademarked search term in the actual text of search engine ads (Chiou & Tucker 2012).
To our knowledge, our study is the first using this kind of data to explore the relationship
between keyword advertising and trademark law.
After the Court of Justice of the European Union held in March 2010 that Google’s
AdWords system does not violate European trademark law,4Google changed its Adwords
policy in various European countries in September 2010. Following this change, Google
allowed third parties to register keywords without the approval of the trademark owner,
with only a limited complaint procedure for trademark owners. We use this exogenous
change in the AdWords policy as a natural experiment to explore the relationship between
keyword advertising and consumer behavior.
We are interested in whether we can observe any visible change in consumer behavior
in click-stream browsing data by comparing web-browsing patterns before and after the
2On the limited admissibility of survey evidence in keyword advertising cases under U.K. law, see Court of Appeal,
Interflora Inc. v. Marks & Spencer plc., [2012] EWCA Civ 1501, [2013] EWCA Civ 319 (Eng.); High Court of Justice,
[2013] EWHC 1291, para. 210, 223 (Ch.) (Eng.); under U.S. law, see 1-800 Contacts, Inc. v. Lens.com, Inc., 722 F.3d
1229, 1246–47 (10th Cir. 2013).
3Whereas the U.S. Court of Appeals for the Tenth Circuit relied on ad impression and click-through data in 1-800
Contacts, Inc. v. Lens.com, Inc., 722 F.3d 1229, 1244 (10th Cir. 2013), the British High Court of Justice has also
utilized eye-tracking studies and aggregate summaries of browsing data, [2013] EWHC 1291 (Ch.) (Eng.).
4Court of Justice of the EU, Google France v. Louis Vuitton Malletier, Mar. 23, 2010, Joint Cases C-236/08 to
C-238/08, ECR 2010, I-02417. For a more detailed discussion, see Section II.A.
720 Bechtold and Tucker

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT