Google to Settle Click Fraud Suit for $90M

Google has agreed to settle a class action click fraud case brought in an Arkansas state court for up to $90 million.

“We’ve been discussing the case with the plaintiffs for some time and have recently come to an agreement with them which we believe is a good outcome for everyone involved. As a result, Google and the plaintiffs are going to ask the judge to approve the settlement, which would resolve the case,” Google said in a statement on its blog.

News of the settlement was first reported by Search Engine Watch.

Terms of the settlement are confidential until approved by the judge, but Google shared some of the details in its statement. The company will offer refunds on confirmed invalid clicks for advertisers beyond its current 60-day limit, in the form of credits to purchase ads with Google. The total amount of credits, plus attorneys’ fees, will not exceed $90 million.

The agreement will cover all advertisers who claim to have been charged but not reimbursed for invalid clicks dating from 2002 when Google launched its cost-per-click advertising program through the date the settlement is approved by the judge.

The case was brought by lead plaintiffs Lane’s Gifts and Collectibles, an Arkansas-based gift shop, and Caulfield Investigations, a Florida-based private investigator. It was filed last April in the circuit court of Miller County, Arkansas, naming defendants Google, Yahoo, Lycos, AskJeeves,, Buena Vista Internet Group, LookSmart, America Online, Netscape and Time Warner. The case was first moved at Google’s request to a federal court, and then moved back to state court in July.

In May, attorneys for the plaintiffs brought the fight online, launching a Web site,, to try to entice other advertisers to join the suit.

At the crux of the argument is the way Google guards advertisers against click fraud — invalid clicks on ads in its AdWords and AdSense programs. Click fraud can occur when a competitor clicks on an advertiser’s ads to cause the advertiser to spend more money, or when a distribution partner clicks on ads to increase its revenue share.

The suit alleges that Google and other search advertising providers have not done enough to prevent click fraud, and are not responsive to complaints from advertisers who bring such issues to their attention. It seeks to cause the providers to refund advertisers for past invalid clicks, and to force them to change their business practices in the future.

In addressing click fraud at the Search Engine Strategies conference in New York last week, Shuman Ghosemajumder, business product manager at Google said it was an issue that Google is taking very seriously, one that it has assigned some of its best thinkers to address.

“Interesting, challenging mathematics and computer science problems attract smart people,” he said. “This is an area we take very seriously, and one that we’ve had some success in as well.”

Ghosemajumder said Google filters out more invalid clicks proactively, before they are billed to an advertiser, than it receives complaints about after billing. In many cases, the invalid clicks are caused by a change that an advertiser makes to the campaign, such as including a new country or using broad-matching on keywords, instead of any kind of malicious activity, he said.

According to Jessie Stricchiola, president of SEO firm Alchemist Media and a subject-matter expert for the plaintiff on the Lane’s case, that’s less a measure of success than a measure of frustration of advertisers who don’t want to waste time bringing click fraud complaints to the search engines.

“We’ve seen a consistent downward trend in responsiveness to submitted activity,” Stricchiola said last week, noting that about a third of the requests for audits she’s sent in have gotten responses, and only a few of those resulted in refunds.

The problem, according to Stricchiola, is that search engines are making a decision about what constitutes a valid click without having some of the more crucial information required to make that decision, including site visitor behavior, clickstream and conversion data.

Related reading