More Evidence That Competitive Keyword Advertising Benefits Trademark Owners (Forbes Cross-Post)

By Eric Goldman

Many trademark owners hate that rivals can bid on Google AdWords ads triggered by their trademarks, a process I call “competitive keyword advertising.” Trademark owners (and many judges) often assume that a Google search on a trademark means the searcher must be looking for the trademark owner’s goods or services; in which case any competitive advertising on that search term is trying to “steal” the trademark owner’s customer. However, as I’ve previously blogged, this assumption is false because searchers routinely use trademarks as a “proxy” for its class of goods. In this post, I’ll discuss a new empirical study that provides more evidence that competitive keyword advertising doesn’t harm trademark owners and may, in fact, benefit them.


Since 2004, Google’s United States trademark policy lets competitors bid on the trademark of third parties, though trademark owners can prevent their trademark from appearing in ad copy in some cases. Google’s United States trademark policy is controversial, spurring over 20 lawsuits against Google in the past decade. Although a few meritless cases are still pending, Google has had extraordinary success defending its U.S. trademark policy in court.

Google’s trademark posture in Europe is more complicated. Court rulings have been less supportive of its practices, and Google has given trademark owners more power to block competitive keyword ads. Things changed in 2010 when the European Court of Justice issued a ruling suggesting Google could legally permit competitive keyword advertising. Soon after this ruling, Google liberalized its European trademark policy to generally allow competitive keyword advertising, thus bringing its European policy closer to its U.S. policy.

The New Empirical Study

How did search behavior change following Google’s 2010 European trademark policy change? Did unleashing competitive keyword advertising hurt trademark owners? Two researchers, Stefan Bechtold (ETH Zurich) and Catherine Tucker (MIT Sloan School of Management) set to find out. Using a huge database of consumer clickstream data from Nielsen Europe, they could track what searchers did after submitting search queries to Google, and they could see how the searchers’ behavior differed before and after Google’s 2010 policy change.

Their top-line assessment of Google’s trademark policy change:

there was no large measurable average effect that can be traced in our data to the change in Google’s policy

shutterstock_4236616.jpgIn other words, Google’s policy change was a non-event for searchers. Nothing happened. This alone raises serious doubts about trademark owners’ concerns over competitive keyword advertising.

While the net effect is negligible, the researchers argue the story is more complicated. They think Google’s policy change produced two offsetting changes in searcher behavior:

after the policy change, there was a 9.2 percent decrease in consumers visiting the trademark owners’ website who used a search phrase that exactly matched the trademark. However, consumers who were searching using the trademark alongside other words were more likely to reach the trademark owners’ website in 14.7% of all browsing sessions.

About 20% of the dataset consisted of searches using the trademark alone or otherwise looked like they wanted to find the trademark owner’s site; the remaining 80% of the queries were the trademark plus other keywords. So out of the small minority of searches in the first group, trademark owners lost some traffic to their website following Google’s policy change; but for the larger remainder of search queries, Google’s policy change produced more traffic for the trademark owner (and, the study adds, those users spent more time on the trademark owner’s website once they got there).

This traffic doesn’t necessarily translate into greater sales for the trademark owner, but it raises serious questions about the trademark owners’ fears that competitive keyword advertising means that rivals can hijack customers. If anything, the dataset suggests that, on average, Google’s trademark policy liberalization produces a win-win-win situation: trademark owners get more traffic, consumers get exposed to more competitive options, and Google makes more money.

If competitive keyword advertising is truly good news for all players, it’s a reminder that policy-makers should tread cautiously when regulating consumer search processes. If they act on intuition or overly cater to trademark owners’ requests–instead of basing their policy decisions on empirical evidence about consumer behavior–trademark regulators can easily lock in socially suboptimal policy results.

For more in-depth reading on this topic, see my 2005 article, Deregulating Relevancy in Internet Trademark Law (anticipating some of the results validated by the study), and my 2009 article, Brand Spillovers.

Citation: Stefan Bechtold & Catherine Tucker, Trademarks, Triggers and Online Searches (July 23, 2013 version).

[Photo credit: Funny vintage bronze sign “On this site in 1897 nothing happened” // ShutterStock]