Google will provide more data and tools to help its advertisers assess and combat click fraud, a controversial practice that is the biggest enemy of the otherwise highly popular and profitable online advertising model known as pay per click.
With most of its escalating revenue coming from pay per click ads, Google has a vested interest in addressing click fraud, which happens when someone clicks on these ads with malicious purposes.
Because in this format advertisers pay every time someone clicks on their ads, companies sometimes click on competitors' ads to drive up their ad spending. Another common click-fraud practice is for Web publishers to click on their sites' ads to increase their commissions.
"We're trying to provide advertisers as much transparency, understanding and control around this issue as we can," said Shuman Ghosemajumder, Google's business product manager for trust and safety.
Thus, starting in March, Google plans to allow its advertisers to "blacklist" certain IP addresses for whatever reason, such as suspicion of click fraud or simply because their clicks never lead to a sale, he said.
"IP filtering is going to allow advertisers to say: 'If I believe a given IP address isn't producing productive traffic for me, then I don't want to ever show my ads to that IP address,'" Ghosemajumder said.
Also this month, Google intends to launch a Web site "resource center" devoted to click fraud, where the company will post information and tutorials to educate its advertisers about this issue, which Google prefers to call invalid clicks.
Since July, Google has reported to advertisers the number of invalid clicks to their campaigns, as well as what percentage they comprise of all clicks. Later this month, Google plans to beef up these reports with another figure: the amount of money Google didn't bill the advertiser by detecting and discarding invalid clicks, he said.
Finally, in the second quarter, Google will provide advertisers with a standardized interface for reporting click fraud complaints that lead to an investigation, he said.
Estimates about the incidence of click fraud vary widely, with some people saying it affects a negligible amount of ads and others suggesting that upwards of 40 percent of all clicks are malicious.
Less than 10 percent of clicks on Google ads are deemed invalid, Ghosemajumder said, declining to be more specific. Most of those are proactively detected by Google, and only 0.02 percent are declared invalid as a result of advertisers' complaints, he said.
Google prefers to use the term invalid clicks because it maintains that not all incidents of click fraud are malicious in nature, but can also come about from innocent user behavior, such as some people's tendency to always click twice on a link.
ClickForensics, which provides technology and services to combat click fraud, estimates that in last year's fourth quarter 14.2 percent of clicks were fraudulent.
Advertisers have taken Google, Yahoo and other providers of pay per click ads to court over the issue of click fraud, alleging that these companies haven't done enough to curb the practice.
Google reached a landmark settlement of a click-fraud class action lawsuit last year which many described as a big victory for the company. A loss could have likely cost the company hundreds of millions of dollars, but instead Google managed to settle the case for US$90 million.
As part of the deal, all but several hundred Google advertisers forfeited their right to sue the company over click-fraud instances dating back to 2002.
The settlement also allowed Google, which admitted to no wrongdoing or liability, to only pay a third of the settlement in cash -- all for plaintiffs' attorney fees -- with the rest taking the form of credits for advertisers. Of the US$60 million allotted for ad credits, many estimate that only a fraction, possibly no more than US$12 million, will be claimed.