1	Fraudulent Clicks and Selling Trademarks as Keywords: Current Legal Issues Facing Search Engines and Their Advertisers Benton J. Gaffney
2	Who Are the Main Search Engines?
3	How Big Is Search Engine Advertising? Google’s Annual Revenues 2002: $439,508,000 2003: $1,465,934,000 2004: $3,189,223,000 Yahoo’s Annual Revenues 2002: $953,067,000 2003: $1,625,934,000 2004: $3,574,517,000
4	Search Results 101.
5	Ranking of Webpages is Monetarily Significant. Where a webpage appears in a list of search results has a huge impact on how many Internet users visit that page. Most end users rarely look beyond the second page of search results. In the late 1990’s, websites began using Search Engine Optimization (“SEO”) strategies designed to manipulate search engines into ranking their sites higher in response to certain searches.
6	How Have Search Algorithms Evolved to Avoid Being Manipulated? Early search algorithms were easily manipulated because they only looked at the content on webpage, such as metatags. Many search algorithms now take into account the webpages most clicked on in response to the same or similar searches. Algorithms also evaluate how many other websites link to a given webpage, and rank the page higher if more links exist.
7	How Do Search Engines Generate Revenue? Advertisers pay to have their websites displayed on results pages in response to searches for certain keywords. Most search engines use a bidding system for purchasing popular keywords and phrases. Advertisers also bid to receive a higher ranking (usually somewhere in the top five results) or bid to obtain a particular placement on the results page, such as above or to the right of the search results.
8	What Are Common Search Engine Pricing Models? “Paid Inclusion” is where advertisers pay a fixed amount each time their website is listed on a results page. “Pay-Per-Click,” the most common model used today, is where advertisers pay each time a user clicks through to the advertiser’s website from the results page. Google’s minimum price for a paid click is $.05, while the most sought-after keywords can cost $100 per click. Some search engines are now using pricing models where advertisers pay based on the “conversion rate,” or the frequency that users purchase something from the advertiser’s website (each being a “conversion”) after clicking through to the website from a results page.
9	How Does Search Engine Syndication Work? Search engines, especially Google and Yahoo, increase their revenues by syndicating their search results and search functionality to other websites or software. For example, all searches done through AOL utilize Google’s search engine. So a search made from www.aol.com will display a private-label version of Google’s search results that matches AOL’s website. Search engines then share the advertising revenue generated from such Internet traffic with the syndicated website or software maker.
10	Trademark Infringement Issues for Search Engine Advertising Primary Issue: Should advertisers be able to purchase keywords that are the same or similar to their competitors’ trademarks? Secondary Issue: Should search engines have a duty to monitor and refuse to sell or stop selling keywords that might also be trademarks?
11	What type of trademark infringement analysis applies to keyword advertising? Traditional likelihood of confusion analysis is not applicable to keyword advertising because search engine users typically are not confused as to the source of goods or services listed in the search results. Specifically, when presented with a list of search results, users are not confused that a website belonging to a competitor of the trademark owner is that of the trademark owner. Instead, “initial interest confusion” applies.
12	What is Initial Interest Confusion? Courts have applied the “initial interest confusion” analysis from the Brookfield Communications case in finding that search engines’ and advertisers’ use of keywords may infringe. Playboy Enters., Inc. v. Netscape Communications Corp., 354 F.3d 1020 (9^th Cir. 2004). Initial interest confusion occurs when a consumer initially is seeking the trademark owner’s goods or services but is somehow diverted to the infringer’s goods and services prior to reaching the trademark owner. Courts may still find trademark infringement under this theory even though the consumer is not confused that the infringer’s goods and services are different from those of the trademark owner.
13	How Does Initial Interest Confusion Apply to Search Engines? In Playboy Enterprises, the court found that Netscape’s placement of adult-oriented banner ads on search results pages in response to searches for “playboy” could result in initial interest confusion. By failing to sufficiently distinguish the banner ads from the result listing for Playboy’s website, the 9^th Circuit found that consumers might be diverted to such other websites by improperly using the goodwill associated with the PLAYBOY mark. From a policy standpoint, initial interest confusion is more about protecting trademark owners, not consumers, from the effect of the infringing activity.
14	But Are Search Engines and Keyword Advertisers “Using” Trademarks? Search engines have argued that displaying ads in response to trademarked keywords does not constitute a “use in commerce” of such trade-marked terms actionable under the Lanham Act. The rationale is that, when “used” as a keyword, the trademarked term is not being used as a source identifier. That is, it is not being used to deceive the user into believing the advertiser’s goods or services are those of the trademark owner. The courts are currently split on this issue.
15	What Are Fraudulent Clicks? Fraudulent clicks are a phenomenon that has grown out of pay-per-click advertising. A “fraudulent click” occurs when someone (an individual or, more likely, a software program) clicks on an advertiser’s link for an some motive other than a bona fide interest in the advertiser’s site or products. The problem is that usually the advertiser still has to pay the search engine for the fraudulent click even though it received no benefit from the person or software clicking through to its website.
16	Why does click fraud occur? Click fraud may result from someone deliberately trying to increase a competitor’s advertising fees by clicking on their paid links repeatedly. A more serious source of paid clicks come from syndication partners of search engines having automated software run searches and click on results to increase the advertising fees generated by their websites. Click fraud is a serious problem for the search engine industry, as some estimate that click fraud may account for as many as 20% of all clicks on search-based ads.
17	Have there been lawsuits over click-fraud? The first class action lawsuit over click-fraud was filed in February against the major search engines in Texarkana, Arkansas. The lawsuit named most major search engines, including Yahoo, Google, AOL, Ask Jeeves, Lycos, Looksmart and Findwhat.com. The plaintiffs’ main allegation was that the search engines knowingly overcharged for advertisements resulting from click fraud.