NEW YORK—As search-based advertising grows, so too does the concern over fraudulent clicks running up advertisers bills and shifting the rankings of paid search listings.
Measuring the extent of so-called “click fraud” has proven difficult, but the anecdotal evidence presented this week during a panel discussion at the Search Engine Strategies 2005 Conference & Expo here left little doubt that advertisers are worried about losing thousands, if not hundreds of thousands, of dollars a year to fake clicks.
Jessie Stricchiola, president of Alchemist Media Inc., a search-engine optimization company that also offers click-fraud auditing services, described a common scenario for pay-per-click advertisers. She read through an e-mail sent from Google Inc. to an advertiser, which explained that the search company was refunding the advertisers account for fraudulent clicks.
While some advertisers welcome the refunds, others become skeptical when they receive such notices, Stricchiola said. They wonder exactly how Google determined the amount of the refund and why the fraud was not prevented.
“This is often where the [click fraud] paranoia comes from, and its coming from the facts,” Stricchiola said.
Click fraud is putting a spotlight on the practices within the pay-per-click advertising field. In the ad model, advertisers bid on keywords for triggering the listing of sponsored links on search and content pages. They then pay a per-click rate to ad programs such as Googles AdWords and Yahoo Inc.s Overture Services.
For search engines, the ads are the main way they make money. Google alone relies on advertising for about 98 percent of its revenues.
Panelists laid out four typical types of click fraud. Among the more common is an advertiser attempting to improve its paid-search ranking by purposefully inflating a competitors clicks to force the competitor out of the auction, said Lori Weiman, director of KeywordMax, a division of Direct Response Technologies Inc., which provides click-fraud auditing services.
“Were not talking about a competitor clicking on an ad out of curiosity but about successive behavior over time to get you to stop bidding on keywords,” Weiman said.
Another approach involves publishers clicking links on contextual ads, most commonly those from Googles AdSense program where pay-per-click ads are distributed across content sites. More clicks mean a larger revenue-sharing check for the publisher.
Late last year, Google sued an AdSense participant, accusing ad partner Auctions Expert International LLC of purposefully and fraudulently inflating clicks.
Click fraud also could be used as a tool of revenge and for blackmail, panelists said.
In all of the approaches, fraudsters can use a combination of automated software robots, called clickbots, to initiate clicks and people who physically hit clicks.
While Google and Overture do proactively inform advertisers of the click fraud they detect, the majority of the burden for finding and proving click fraud has fallen to advertisers, panelists said.
“Search engines are profiting and have not published rules that speak to an investigation or give those harmed the information to get to the bottom of problem,” said Ben Edelman, an Internet privacy research and Harvard University student.
Advertisers can analyze click patterns, visitor logs and other information, but Edelman said the search engines themselves should be able to access more detailed information that could help advertisers track down click fraud and even prove it in court.
So far, click fraud has yet to be tested in court, but panelists and attendees largely agreed that they expect cases to emerge both against alleged fraudsters and possibly against the search engines.
“While we know this is happening, there are no reported cases on this yet,” said Peter Raymond, a partner at law firm Reed Smith LLP, in New York. “Obviously theres a real problem in gathering this proof to bring this kind of case.”
Check out eWEEK.coms for the latest news, views and analysis on enterprise search technology.