As part of the Lane’s Gifts v Google settlement, an independent expert reviewed Google’s click fraud processes and practices.
Dr. Alexander Tuzhilin, Professor of Information Systems at NYU, submitted his 47-page report to the Arkansas court overseeing the click fraud settlement between Google and Lane’s Gifts and Collectibles. Google’s Shuman Ghosemajumder, Business Product Manager, Trust and Safety, revealed on the Google Blog that Tuzhilin concluded “Google’s efforts to combat click fraud are reasonable.”
Although Google did not completely agree with some parts of the independent review, they were pleased with the overall tone. Ghosemajumder wrote that the review was “a validation of what we have said for some time about our work against invalid clicks.”
Tuzhilin provided an overview on the final page of the report on how Google works to combat click fraud by finding invalid clicks:
“Google has built the following four ‘lines of defense’ for detecting invalid clicks: pre-filtering, online filtering, automated offline detection and manual offline detection, in that order. Google deploys different detection methods in each of these stages: the rule-based and anomaly-based approaches in the pre-filtering and the filtering stages, the combination of all the three approaches in the automated offline detection stage, and the anomaly-based approach in the offline manual inspection stage. This deployment of different methods in different stages gives Google an opportunity to detect invalid clicks using alternative techniques and thus increases their chances of detecting more invalid clicks in one of these stages, preferably proactively in the early stages.”
As part of the information-gathering process, Tuzhilin reviewed a number of internal Google documents discussing the invalid click detection process. Those included descriptions of filtering methods, investigative processes, and an attack simulation system, among other topics.
Part of the problem with detecting click fraud involves incomplete information being available to Google, advertisers, and third-party vendors. Google can collect certain information that they do not make available outside Google. Advertisers and third-parties can choose to provide clickstream data to Google, like post-click through behavior, but many choose not to do so.
Tuzihilin called the long-standing issue of Google not providing information on why a particular click was or was not judged invalid by Google the Fundamental Problem. Advertisers have a right to know this information, he conceded, but also pointed out that such revelations would leave Google open to “more sophisticated fraudulent activities undetectable by Google’s methods.”
Then there is the whole problem of human behavior that cannot be assessed with one hundred percent accuracy, such as in the event of someone clicking an ad, reading the landing page, then clicking back through the ad again to see the content a second time:
Therefore, in some cases the true intent of a click can be identified only after examining deep psychological processes, subtle nuances of human behavior and other considerations in the mind of the clicking person. Moreover, to mark such clicks as valid or invalid, these deep psychological processes and subtle nuances of human behavior need to be operationalized and identified through various technological means, including software filters. Therefore, it is simply impossible to identify true clicking intent for certain types of clicking activities and, therefore, classify these clicks as valid or invalid.
It appears that click fraud detection can reach a certain point of accuracy, but no greater, based on Tuzhilin’s assessment. “It is impossible to measure the true rates of invalid clicking activities, and all the reports published in the business press are only guesstimates at best,” he also wrote.
—
Tag:
Add to 
Bookmark Murdok: 
David Utter is a staff writer for Murdok covering technology and business.
