Fradulent Clicks on the Internet

Came across an interesting article today on TechCrunch that talks about the growth in click-fraud rate. According to the research, nearly 2 out of 3 clicks on Google or Yahoo ads are fraudulent. While that statistic is daunting, it’s not surprising. Many Internet sites have built their revenue models around serving ads. The formula is simple: generate good content; optimize it for search engines to drive hit rate; serve up ads; rinse and repeat. Execution is often not as simple as one must have compelling content as a starting point. Digg and Plentyoffish have been some of the top revenue generators with reported earnings as high as 901k per month from Plentyoffish.

The increase in click-fraud rate presents a big problem for the advertisers that write those checks. One of the comments in the TechCrunch article stated “Yes, the rate will go higher and higher. No one can stop it, not even Google.”. While I don’t fully agree with that statement, the problem does exist and will require a system-wide solution. I can think of areas that are starting points to solve the problem:

• early detection at ISP edge to prevent fraudulent traffic from leaving the network (value-add for ISPs)
• an intrusion detection engine at the ad provider data centers to pool stats on ad clicks (apply the same approach on worm/virus detection)

This is an area where Cisco can make an impact as often times it’s Cisco gear that provide transport for these fraudulent traffic. While the fraud-click prevention space may not appear significant right now, the dependence on the intelligent network will continue to grow as our computing world shift away from desktop computing to service-based internet computing.