Click Fraud Getting Worse, Especially on Content Networks

Click Forensics, a Texas-based company that tracks click fraud using detailed campaign data from more than 3,500 marketers, reported that industry-wide click fraud increased modestly in Q1 to its highest level ever: about 15% of all clicks, versus 14% in Q4 2006.

More ominously, click fraud on "content networks"--the third-party advertising solutions that support an ecosystem of thousands of small content companies--increased a more significant 3 points, to 22%, from 19% in Q4.  This trend is dangerous for small content providers in addition to search engines.  A continuation of this trend will soon result in more than a quarter of all content-network clicks being considered fraudulent, a level that could begin to cut significantly into the revenue of smaller content providers.

Also significant: Fraud on high-priced keywords--those over $2 a click--rose to 22% from 21% in Q4, confirming the theory that higher priced keywords are more susceptible to fraud than average- and low-priced keywords.

From the release:

“It appears that click fraud perpetrators are becoming more sophisticated even as search providers step up their efforts to fight click fraud,” said Tom Cuthbert, president and CEO of Click Forensics, Inc. “Click fraud seems to be following a similar path as other online fraud schemes such as spam and phishing - the problem is growing as fraudsters fine tune their methods.”

Comments

[The future is now or coming soon to a Theater ner you.]

Henry,
On a totally different topic.

If you are going to be on the Google earnings call today I would like to offer you a question to ask. In their TV test with Echo they are bringing efficiencies to a broken model. TV needs a search model they sends ads only to viewers that match a marketers profile. This can be done by delivering multiple ads in a single commercial pod, each ad appearing only to viewers that match the profile. The proposed test will measure commercials but will not in any way deal with the reach and frequency problems now facing TV. As I discussed in a previous post, the issue Google will face using low rated cable networks is redundant frequency. The more a spot runs on the same niche network the more the same viewers will see the ad. Their gross impression total will be skewed with a small number of viewers representing a disproportionate number of impressions. Do they intent to go to true addressability where like viewers can be aggregated into large anonymous clusters? There are two companies that can do that today, INVIDI Technologies Corporation and OpenTV. They approach this solution differently but appear to be the only ones that have the technology and IP to make TV a relevant advertising going forward. One of these appear to be an ideal takeover prospect for Google.. They won't tell you but I would love to hear their answer.

[Robert]

Henry,
while the sub-title of your blog says.."Analyzing the Internet Business: companies,
trends, research and more....", it appears that the content of such analysis has clearly
changed, and in my eyes the content has clearly diminished in its usefulness.
would you care to comment on this: for instance, you no longer predict earnings,
margins, sales, and the like. to me, your blog lately has become just an internet gossip page, nothing much is original, simply regurgitated opinions of others.
you know this. we all see the transformation. The comment section of each posting used to have substantial follow-ups by your readers. Now, as in the posting above, there have been no follow-ups.
I respect your opinions and wonder if
in a previous comment, you have discussed this.
I guess I want to know if this blog is relevant any more.

[Victor]

http://investor.google.com/releases/2007Q1.html

Looks like rev beat estimates by a little bit, but EPS is a blowout. Margins seem to be improving over at Google...

[Victor]

Wow, these numbers are amazing. These guys are going to do $16 a share in 2007. that's HUGE