WebProNews

A couple of numbers regarding click fraud have been in play and widely discussed: ClickForensics' 14.2 percent claim and Google's internal assessment of less than 2 percent. Those numbers don't tell the whole story either.

Google Defends Against Click Fraud Report
We heard from Google's Shuman Ghosemajumder after a recent discussion of his response to a report from ClickForensics on click fraud. The debate about click fraud has been ongoing, as our readers, many of whom use PPC campaigns, have observed for some time.

Shuman said a key point made in his first blog post demonstrates why we can't directly compare the 2 percent figure to the 14.2 percent. "It's an apples to oranges comparison," he said. Here's the point he made that merits further discussion (his reference to "those problems" means inaccurate and inflated click counts):

Even if they fixed those problems, they're not actually measuring click fraud

Even if they were counting clicks correctly, they are still trying to measure only activity (attempted click fraud) and not advertiser impact (actual click fraud). That is, even if they corrected the basic engineering and accounting problems contributing to the above problems, they would still be counting clicks we filter (and do not charge to advertisers) in their click fraud estimates. They admit this.
"This is a key point -- even if there was nothing wrong with their data collection (which of course, we've demonstrated there is) ClickForensics' 14.2% figure is their estimate of attempted click fraud, not actual click fraud," said Shuman. "So calling this a click fraud number is incorrect."

He moved on to discuss attempted click fraud in his email, calling it an "open loop number" because "there's no limit to how much someone can try, using computers." Even without any filtering in place, Shuman cited how an advertiser's budget would stop those clicks from registering eventually.

Google of course does filter those efforts, and that is the point of contention. The company considers actual click fraud, not attempted click fraud as third parties do when forming their numbers.

"Here we have an attempted click fraud number being reported as an actual click fraud number," said Shuman. "This is certainly not accurate."

The essential issue as we still see it comes back to a need for auditing these figures independently. Last summer, when the Lane's Gifts v Google case was taking place in an Arkansas court, Google did have an audit of its procedures performed.

Steps taken by Google as observed by Dr. Alexander Tuzhilin, Professor of Information Systems at NYU, left a Fundamental Problem in place, which was summarized like this:

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."
Tuzhilin determined Google had made reasonable efforts to combat click fraud. There were points he made that Google did not completely agree with, though they were satisfied with the overall tone.

Now the industry, particularly its most visible member, needs to hammer out the third party auditing process through the IAB working group. With so much money in play, a mechanism to measure click fraud independently, instead of the back and forth discussion taking place on blog posts and company reports, has become essential.

---
Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

David Utter is a staff writer for WebProNews covering technology and business.