Click frauds, it seems, have hit a new high in the online ad space. US-based online traffic quality management firm Click Forensics, Inc. reported that click fraud rates for pay per click, or PPC, ads touched 17.1% in the fourth quarter of 2008, against 16.6% a year earlier. PPC ads are Internet ads for which advertisers pay their host only when the ad is clicked on.
Chandrakanth B.N., co-founder and president, global services, Theorem Inc., a global media operations firm, enumerates the different types of click frauds pertaining to both display and search ads: “Manual clicks on an ad to purposefully increase the ad spend; deliberate clicks on an ad to increase profits by site owners hosting the ads; and automated clicking tools, bots, or other deceptive software. This hurts the advertiser by clicking on their ads with no intention of purchase.” Bots are software applications that run automated tasks on the Internet.
Manual click frauds are perpetuated by “click farms” that are present in India, too. Anshuman Misra, vice-president, business development, India, LogicServe UK Ltd, a search engine marketing firm, says a click farm consists of a group of people who sit and click on ads on particular sites and are paid to do so. Many of the part-time job ads pasted on Mumbai’s local trains may well be referring, in fact, to the click frauds perpetrated by click farms.
Also Read Marion Arathoon’s earlier columns
Although Google, Yahoo and other search engines have implemented automated systems to guard against click fraud by rivals or corrupt online publishers, the situation is expected to only get worse. Click Forensics has cited three contributors in a report: More dollars filtering into PPC; increased competition in the click farm industry and rise in cybercrime and botnet (or malicious software) activity during the economic downturn.
Roy de Souza, founder and CEO, Zedo, Inc., the world’s third largest ad serving network, tells me that click fraud is usually associated with ads served on thousands of publisher sites by Google Inc.’s ad network. “Almost anyone can run Google ad network ads on their sites. The site is paid every time a user comes to the site and clicks on a Google ad. Some site owners, therefore, click on the Google ads on their own site even if they are not interested in the ad. Further, some site owners build software that clicks on the Google ads on their site,” says de Souza. “These clicks are not real clicks by users who are interested in the ad. Therefore, they are known as fraudulent clicks.”
Click fraud actually short-changes everyone. It tarnishes the image of the entire online publishing community. Ad networks have to continually fund technologies to nab offenders. And since various sites all around the world engage in click fraud, advertisers everywhere stand to lose a lot of money. Misra says the more the fraud, the more genuine advertisers have to pay for each conversion/sale.
Click fraud varies widely from industry to industry and depends on several variables, including category of ads, number of competitors in the field, and cost per click. For example, a competitor is much more likely to maliciously click on your ads if you are paying $2 (Rs99.8) per click than if you are paying $0.20 per click, says Chandrakanth.
There is, however, a perception that online ads in India are actually more vulnerable to click fraud.
“The cost of a click is much higher in the US, so advertisers there could lose a lot of money. Google is, therefore, probably more vigilant in searching for fraudulent clicks in the US,” says de Souza. “The price per click is lower in India and so less money is lost by advertisers and Google may, therefore, not be as vigilant in preventing click fraud in India.”
Says Parminder Singh, business head, Google India Pvt. Ltd: “We believe we lead the industry in terms of our innovation in this area as well as the effectiveness of the protection we provide, but we also look forward to continuing to innovate and invest in this area.”
He underlines that click fraud protection is something Google takes very seriously. Its three-stage system for invalid click detection:
• Real-time filters—automated algorithms that filter out invalid clicks in real time
• Offline analysis—automated algorithm and manual analysis more focused on AdSense networks and
• Reactive investigations—all advertiser inquiries are investigated by a click quality team.
Tougher surveillance is clearly needed online and offline. De Souza mentions that Zedo employs sophisticated algorithms. It also keeps lists of IP addresses (computers) that are known to engage in click fraud and disregards clicks from them.
It looks for click patterns that are suspicious—such as one click exactly every 60 seconds. No individual clicks exactly every 60 seconds all day, so these clicks are probably fraudulent.
The easiest way to spot click fraud is to look out for computers that are always clicking but never buying the product advertised, he advises.
Marion Arathoon is Mint’s advertising editor. Your comments are welcome at firstname.lastname@example.org