Deceptive clicks on internet advertising links distributed by Google Inc., Yahoo Inc. and other online marketing vehicles are probably occurring far more frequently than the network operators acknowledge, according to a study by fraud detection specialist Fair Isaac Corp.

The chicanery involves automated computer programs or scam artists who repeatedly click on ad links with no intention of buying anything. The short ad links, which appear alongside search results and other content at thousands of websites, typically trigger a commission with each click — a financial formula ripe for mischief, Minneapolis-based Fair Isaac found.

The study's preliminary conclusions, scheduled to be discussed Friday during a Fair Isaac conference in San Francisco, threaten to revive suspicions among advertisers that they have been overcharged as part of a ruse known as "click fraud."

After reviewing a handful of websites since last August, Fair Isaac believes 10 to 15 per cent of the advertising traffic is "pathological," indicating a likelihood of click fraud, said Joseph Milana, the company's chief scientist of research and development.

"It's still an early result," Milana said. "The question remains about how broad the problem is in the entire marketplace."

Filters aimed at fraud artists

The culprits behind click fraud typically are either trying to make more money from the ads appearing on their own websites or maliciously trying to drain the marketing budgets of a competitor.

Google, which runs the internet's largest ad network, maintains that its engineers and filters identify all but 0.02 per cent of the click fraud on its network. The Mountain View-based company says it doesn't bill advertisers for any of the flagged click fraud.

Yahoo, which runs the second-largest ad network, also maintains its preventive measures weed out all but a small portion of click fraud.

Fair Isaac's initial estimates fall in the same range as those made by Click Forensics, a San Antonio-based consulting service that compiles a quarterly index tracking click fraud rates. Other studies have estimated click fraud rates as high as 30 per cent, a figure implying advertisers have paid billions of dollars for bogus sales referrals during the past few years.

Google and Yahoo have consistently ridiculed double-digit click fraud estimates as the handiwork of search engine consultants trying to drum up more demand for their services by alarming advertisers. Critics point out that Google and Yahoo have a powerful incentive to debunk the click fraud claims to preserve confidence in a system that generates most of their profits. Last year alone, Google and Yahoo sold a combined $16 billion US in internet ads.

Ads on other websites accounted for $4.16 billion US, or 39 per cent, of Google's revenue last year. Google shared $3.31 billion of that revenue with its advertising partners. Yahoo doesn't break out how much of its revenue comes from ads on other websites.

Fair Isaac enters the debate with a track record for ferreting out fraudulent conduct in other industries. Best known for a scoring system that rates the creditworthiness of consumers, Fair Isaac also has helped banks fight credit card fraud for 15 years. More recently, the company has sold anti-fraud tools to health-care providers and telecommunications companies.

Now, Fair Isaac is trying to determine whether click fraud is a big enough problem to justify the company developing a potential solution that could help boost its own profits. "This is a problem that fits well in our sweet spot," Milana said.

Click fraud doesn't appear to be a major problem when the ads appear on Google's and Yahoo's respective websites, Milana said. The trouble starts cropping up once Google and Yahoo deliver the ads to other websites that are part of their vast marketing networks.

"They just don't know what happens beyond their own firewalls," Milana said of Google and Yahoo.