I wish I had more time, running to the first meeting of the day. So read Google’s announcement, and then read Danny….
More Numbers from Google, This Time on Invalid Clicks (AKA ClickFraud)
I wish I had more time, running to the first meeting of the day. So read Google's announcement, and then read Danny…….
2 thoughts on “More Numbers from Google, This Time on Invalid Clicks (AKA ClickFraud)”
John, this numbers are for AdWords. I together with some colleagues in Microsoft Research proved that click fraud is self neutralizing and not a major concern for Search ads. This is a theoretical study and should be taken with a grain of salt. But the arguments we made are sound. They were also echoed by Eric Schimdt which you posted on your blog at http//battellemedia.com/archives/002712.php. These numbers, if true, also provide additional support for the arguments.
I am more interested in similar numbers for AdSense. I think this program has scope of major improvements from game theoretic perspective.
FYI, our click fraud paper could be found at http//research.microsoft.com/~nickle/pubs/clickfraud.pdf. The author names are alphabetically listed. The third author was a two year postdoc here who is now a full researcher at Yahoo Research. The first author is also a two year postdoc here. She is an amazing colleague and researcher.
A note regarding the last sentence: Since your blog is widely read by people in her area of research, I took the liberty to advertise my support for her. She is finishing her two year postdoc here and interviewing for various jobs in the area.
It’s nice to see Danny take such an even-handed commentary instead of jumping on Google’s numbers. It’s also high time Google put in IP filtering. For any given business, most click-fraud probably comes from a handful of competitors and could easily be traced by even a semi-competent IT person. Google needs to realize that putting that power in the hands of end-users is only going to save them money and grief.