Click Fraud: Now a Mainstream Story
Over the past few months, DM News has published the concerns that my firm, SendTec, has set forth on search arbitrage practices and their potential connections to click fraud.
While it is difficult for advertising firms and advertisers themselves to undertake an investigation thorough enough to get the search engines to listen, the groundswell of our concern on this subject matter has come full circle as the cover story of the Oct. 2, 2006, issue of BusinessWeek, titled "Click Fraud: The Dark Side of Online Advertising" by Brian Grow and Ben Elgin. It details their investigative analysis of search arbitrage tactics gone wrong.
The story details the adventures of two click-frauding search arbitragers who were more than happy to discuss their financial windfall through the simple development of a Web site that has no content, only paid search listings, and employs people across the country to click on their ads.
Everyone in the chain knows this is wrong and unethical, yet they perpetuate this fraud willingly for the almighty dollar.
In the words of David Struck, who made $20,000 in four months doing next to no work at all, "it was way too easy" to essentially fool the engines.
Even more concerning is the budding 23-year-old Hungarian millionaire Roland Kiss, who rakes in $70,000 in ad revenue per month by engaging his "Paid to Read" membership to click on his Web site ads with instructions on where, when and how often they should click. These are strict instructions that must be adhered to in order to receive payment so that the click fraud will go undetected by search engine detection algorithms.
The good news in the BusinessWeek article is that the federal government is finally getting involved to put an end to this con artistry that has exploded since the beginning of 2006. The FBI and U.S. Postal Inspection Service have assigned a task force to begin looking into federal violations, while the Senate Judiciary Commitee has launched its own review into the problem. As with just about all media, a small number of unscrupulous people are giving the larger community a black eye.
While Google's click fraud czar, Shuman Ghosemajumder, professes that "people have a great tendency to exaggerate when they say they can attack Google's service," it appears that the only exaggeration is Google's and Mr. Ghosemajumber's belief that their algorithmic models can figure out anything.
Google's continued refusal to acknowledge that math doesn't always win is disturbing, and it is time that more policing by human beings and a stronger policy outlining acceptable use is needed.
We industry insiders know better, as we have seen the magical fixes to the algorithm within minutes of calling 866/2-GOOGLE to lodge a complaint and request an engineer review. The algorithms are more imperfect than some would like to admit: Simply search on the term "miserable failure" for proof of how the algorithms are manipulated on a daily basis.
Do not consider this a Google-bashing tirade. As an SEM agency, we want nothing more than to eradicate this problem, and we offer our assistance to Google and all the engines to make it happen. The engines need to realize that the solution requires more human intervention.
In my opinion, there are some easy ways to eliminate the problems quickly and easily, without relying on major technology overhauls:
* Ad distributors must be required to pass the referral URL for logging. On a recent pull of Web logs, our firm found that more than 20 percent of the log line listings had the field blank. This would make it easy to identify the bad guys.
* There needs to be immediate flagging, review and reimbursement of traffic from international IP addresses, particularly when the advertiser's ad is supposed to be shown only in U.S. geography. If we say we want U.S.-only traffic, shouldn't we only have to pay for U.S. traffic?
* There must be specific requirements and guidelines on acceptable practices for impression generation. Illegal, false and misleading advertising must result in immediate legal action and program expulsion.
* All entities involved -- the engines, advertisers and intermediaries -- need to jointly develop a blacklist of all suspected click fraud perpetuators and share that information for mutual protection.
* There should be a manual daily review of at least 25,000 partner sites that delivered traffic to see whether they are in violation of any of the guidelines.
* Set up an industry task force that reaches beyond just the search engines bonding together, as advertisers and SEM firms have much more insight into the click fraud problems.
And, from my personal wish list, I'd love to see Google and Yahoo add functionality to their toolbars that allows users to report useless and unhelpful Web pages where page search listings reside, in the same way AOL members can report spam. Simply create an algorithm that scrapes the page, evaluates the balance and presentation of the content, and should something suspect present itself, prompt the user to vote on the value of the page.
This will allow the search engines to quickly weed out the bad guys. No worries, my search engine friends: I promise I won't try to patent the idea. It's all yours for free, as I am a proponent for a click-fraud-free search engine universe.
It's time to finally admit that algorithms alone are not the answer. Human policing, tougher distribution guidelines and strong legal actions against violators are needed.
The search engines need to understand that eliminating click fraud is to their benefit, not the advertisers, and this will lead directly to better bottom lines and higher stock prices. Advertisers have set budgets and ROI goals that must be met, and currently, they are just embedding fraud into their spending level and CPC targets.
Eliminating the fraud means less distribution partner fees and happy advertisers who will increase their budgets.