There are two main types of click fraud: One is when a competitive firm clicks on another's ads to drive up its costs, deplete its budget and hopefully have that company drop out of competition. These frauds typically occur on search engines. The second type of click fraud is perpetrated by publishers or networks on contextually linked ads in order to boost their own pay-per-click revenues.
In both cases, when we say "clicks," it often is not done manually, but instead by using automated scripts or programs, typically called "bots."
Among search agencies, competitive click fraud is ubiquitous, according to a study commissioned by the Search Engine Marketing Professionals Organization (SEMPO), conducted by Intellisurvey and
Radar Research. All of the 86 responding organizations said they had been conned at some point. In the advertisers' experience, click fraud on contextual links was more common than the competitive kind.
Fair Isaac has been examining ad network data to find exactly how many clicks are fraudulent, what the nature of the fraud is and (potentially) how to fight it.
In a May 2007 interview, Fair Isaac chief scientist Joe Milana told eMarketer that although the results are still preliminary, initial findings indicate that click fraud on ad networks may be as high as 15%-20%.
Clicks are especially important for ad networks because measuring click-through rates is one of the primary metrics agencies and advertisers use to judge ROI, according to the "2007 Ad Network Study" conducted in April by Collective Media.
Previous estimates by Yankee Group calculated that fraud affected one out of every 10 clicks on text ads, which is not far from Google's rough estimates. The research firm estimated that the fraud equated to a $500 million-a-year problem.
Not surprisingly, Google has disputed high click fraud claims before.
Regardless of the theoretical dollars lost, some marketers may consider click fraud as just part of the cost of doing business. A MarketingSherpa survey conducted in September 2006 found that a majority of marketers thought that like e-mail spam, click fraud would just continue to cost time and money.
Notably, 18% of those respondents who monitored click fraud said it was becoming a non-issue. That was a higher percentage than those who were not monitoring the problem.
