SAN FRANCISCO - Google Inc. appears to be doing reasonably well protecting advertisers from scam artists preying upon the Internet's largest marketing network, but it remains unclear how much the system is being bilked under a ruse known as click fraud, according to an independent report filed in court Friday. Alexander Tuzhilin, a professor of information systems at New York University, compiled the 47-page analysis of Google's fraud-fighting defenses as part of a proposed $90 million class-action settlement that will be considered for final approval in a two-day hearing scheduled to begin Monday in an Arkansas state court. At least 51 objections have been filed against the settlement, which would pay up to $30 million to lawyers and divvy up the remainder to thousands of advertisers worldwide. The advertisers paid for each click on short text ads primarily tied to requests entered into Google's Internet-leading search engine during the past 4 1/2 years. Google also distributes the ads to thousands of other Web sites that share in the commissions. Those austere ads have generated $15.7 billion in revenue for Google and its partners since 2001, turning the Mountain View, Calif.-based company into one of the world's most prized businesses. Lane's Gifts and Collectibles filed the lawsuit last year, alleging Google has been improperly billing advertisers for bogus sales leads triggered by swindlers and mischief makers who repeatedly click on the commercial messages with no intention of buying anything. Clicking on the ads, typically displayed at the top and sides of Web pages, triggers sales commissions even if the activity doesn't lead to a sale. Click fraud cropped up several years ago as a way to drain advertising budgets or funnel illicit revenue to Web sites that belong to Google's network. Google has long maintained that its computer programs and engineers are highly effective at weeding out click fraud — a thesis Tuzhilin generally supported, although he also quibbled with some of the defense that the company has adopted since it began to tackle the problem in 2003. "These efforts were not perfect since Google missed certain amounts of invalid clicks over these years...," Tuzhilin wrote. "However, click fraud is a very difficult problem to solve, Google put a significant effort to solve it, and I find their efforts to be reasonable." Tuzhilin wrote that Google's fraud-fighting efforts have improved significantly since March 2005 when the company stopped charging for two rapid clicks on the same ad link. Other studies by outsiders have asserted click fraud is far more costly for advertisers than Google and Yahoo Inc. (Nasdaq:YHOO - news), the owner of the Web's second largest marketing network, are letting on. A report released this week by Click Forensics Inc. estimated 14.1 percent of the ad clicking during the second quarter was fraudulent, based on information provided by 1,300 advertisers. Other industry consultants believe the click fraud rate is much higher than that. In his analysis, Tuzhilin dismissed the findings of other click fraud studies because they "would not stand hard scientific scrutiny." At the same time, Tuzhilin conceded that he didn't gather enough hard data in his study of Google's fraud-fighting techniques "to arrive at any definitive conclusions beyond any reasonable doubt."
"In his analysis, Tuzhilin dismissed the findings of other click fraud studies because they "would not stand hard scientific scrutiny." Their job is to make Google and Yahoo look bad. They are probably inflating the click fraud figures.
Google were very proactive in detecting and refunding fraud against my AdWords ads. I reported every instance, just so I had record of the activity and my contact. I was elated to find that Google had already detected and refunded 90% of the fraudulent clicks; all coming from the same IP address. They even reaudited my account per my complaints, found other repetitive clicks from other visitors, and notified me that they had refunded my account even further. I think the best way to keep them alert, besides their regular method of monitoring clicks, is to report any suspicious activity; whether the activity was purposefully perpetrated or not. I am happy with the results, and Google's follow-through. This is just my experience, though.