Click Fraud Could Swallow the Internet


Using this headline, Wired are running a fairly comprehensive resume of the current state of click fraud, and click fraud counter measures. Whilst it did not add much to my knowledge of click fraud, it does highlight the dearth of an concrete figures on it - Google and Yahoo do know, but are not letting on -

The amount of click fraud is difficult to quantify; estimates of the proportion of fake clicks run from as low as 1 in 10 to as high as 1 in 2. In a widely cited recent study,, an online marketing research outfit, reported that "as much as 29.5 percent" of the clicks in three experimental PPC campaigns on Google were fraudulent.
Yahoo! doesn't break out PPC earnings separately in its financial statements, but Goldman Sachs analyst Anthony Noto believes that keyword advertising accounted for about half of the company's estimated $3.7 billion in revenue for 2005. PPC is even more lucrative for Google. According to Noto, Google will end 2005 with $6.1 billion in revenue.

Whatever happened to Greg Boser's test ??



I knew that name rang a bell - Anthony Noto & Phone Spam.


Whilst he might be a pain in the whatsits, he is well read in e-commerce

Anthony Noto, Vice President, is the e-commerce analyst at Goldman Sachs. His coverage of the sector includes,,, e-Bay, eToys,, PlanetRx, Priceline, and Webvan. Anthony was ranked number one in the 2000 Greenwich Survey, and runner-up in Institutional Investor poll in his first year of coverage of the sector.

I did a check with someone from Goldmans and was told that he would not have been trying to sell Greg anything, as he was an analyst - the Personal Investment department at Goldmans that do try to sell you things only make the call if you are known to have a net worth of around $30 million. Then there again maybe Greg has been doing well recently.


The amount of click fraud is difficult to quantify

Well then wait until you get some damn facts before declaring estimates of 1 in 10 or 1 in 2 and scaring off all the advertisers.

I'm getting sick of these so-called writers aka bullshit artists that just sit around and do nothing but write popularized scare pieces that the sky is falling.

Yes, fraud happens, but if was that high nobody would use the net to advertise as the crashing ROI would be so obvious it wouldn't be worth using.

Well then wait until you

Well then wait until you get some damn facts before declaring estimates of 1 in 10 or 1 in 2 and scaring off all the advertisers.

I think you'll find that clickfraud scales differently as a problem in different business sectors. I'm sure a number of AdSense publishers who cover different verticals will notice that some keyword areas seem to attract a disproportional amount of clicks compared to others. Clickfraud? No proof, but there is an interesting disparity between verticals.

"For the first time in more

"For the first time in more than eight years, the Chicago Sun-Times has reliable figures for how many papers it sells."

"... discovered that previous management had been inflating the paper's sales figures at least since 1997."

"It later emerged other papers -- including the Dallas Morning News as well as Tribune Co.'s Newsday -- had also engaged in circulation fraud."

And since advertising rates are based on circulation this is very similar to clickfraud in our industry. There is a notable difference though, in the print industry the fraud has been by the publisher alone while online ads are vulnerable to competitors.

here's another little print ploy

While on the subject of fraud (and the-sky-is-falling), here's another one: junk circulation (emphasis mine)

"...And some of that audience wasn’t (cough) really there. TV Guide is getting rid of its junk circulation. For example, my late father-in-law, a dentist, got a weekly pile of at least seven copies of the magazine every week. That was no fluke; my kids' orthodontist got the same pile. And I have little doubt that because they were sent to waiting rooms, those copies were counted with extra readers-per-copy — a standard measure in the (cough) measurement system in publishing. There is a bigger publishing story than TV Guide here. Other publications have similar junk circulation and the Audit Bureau of Circulations just disallowed counting from some questionable sources. This is a ballsy move and I’d expect to see advertisers demand more of the same."

So, did circulation fraud swallow print advertising? No, it was just another cost of doing business.

splogs really to blame, if you believe some people...

Some person I have never heard of has wrote that aparently Splogs are entirely to blame for the fact that PPC is costing them hundreds of dollars and they havent received a sale in 6 months. And of course, to these people, splogs = click fraud (WRONG WRONG WRONG). Full post here. (hat tip to SEOBH).

This Fraud is a Fraud

Wired should be embarrassed. The media and others are working overtime to convince us that advertisers actually *do* give a damn. They don’t and Google’s valuation, growth and sources of revenue (99%, as reported by Wired, from click ads) serve to prove the point.

Advertisers don’t mind click fraud and probably never will.

Still... Wired is absolutely breathless, screaming "It's search giants against scam artists in an arms race that could crash the entire online economy."

The entire online economy? Is Wired seriously asking us to believe that click fraud is anything other than an accepted risk? I ask author, Charles Mann, to consider one simple fact before he proudly hangs a framed version of the story on his office wall. That is, every article ever written (to my knowledge) about click fraud "victims" share commonalities that support my statement - advertisers accept CPC fraud as a cost of doing business.

What about fixable problems?

They worry about click fraud but a couple of my customers have lost thousands of dollars due to technical issues and nobody seems to care about that and unlike click fraud it's easily preventable.

Seems to me the PPC companies are probably going to have to take "smart pricing" to a whole new level and actually offer discounts per visitor based on that person's history of ROI meaning that if you've never ever purchased anything your clicks are worth a lot less than people that regularly buy things online which would drastically and rightly devalue those clicks.

Although it might not stop click fraud, it would take some of the bite out of it and likewise drastically reduce the income on SpamSense sites thus making them less likely to exist.

Wired can be a joke

Wired has a wildly varying quality level.

After an article in one of *my areas* hit Wired, with many really bad "inaccuracies", I chatted with the author. His response? "Gee, thanks alot for that additional info. This was the first time I wrote abut X, but if I ever get to write about it again I will sure and call you up" (not an actual quote).

Less than stellar quality research/reporting and nobody seems to care much.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.