Polygraph: Blocking IP Addresses Misses More Than 95% Of Click Fraud
August 2022 by Polygraph
Polygraph, a cybersecurity company specializing in click fraud detection and prevention, is warning advertisers not to rely on IP blocking to prevent click fraud.
Click fraud is a sophisticated cyber-crime stealing tens of billions of dollars each year from online advertisers. Criminals pose as legitimate website owners, and use bots - software pretending to be human - to generate fake clicks on adverts. For each of these clicks, the advertisers pay money to the advertising networks, and the money is then shared with the criminals. Multiply this by millions of clicks per day – and thousands of scam websites – and the money being lost by advertisers is huge.
A common approach for dealing with click fraud is to block "bad" IP addresses from seeing or clicking on advertisers’ ads. The logic goes like this: if we know these IP addresses have been used to generate fake clicks in the past, we might be able to prevent future click fraud if we block the IPs from seeing people’s adverts.
According to Trey Vanes, chief marketing officer at click fraud detection firm Polygraph, blocking IP addresses from seeing or clicking on ads provides a false sense of security and offers little value.
"We wanted to quantify the effectiveness of blocking IP addresses as a click fraud prevention strategy," said Vanes. "We randomly selected 10,000 IP addresses which have previously been used by click fraud bots. What we discovered was less than 20% of click fraud IP addresses are used more than once. That means if you were to block all 10,000 IP addresses from seeing your ads, you’re going to miss more than 80% of all click fraud."
To make matters worse, advertising networks like Google Ads only allow advertisers to block 500 IP addresses at any one time.
"Since Google Ads only lets you block 500 IPs from clicking on your ads," continued Vanes, "that means you can only block 500, or 5%, of the 10,000 click fraud IP addresses. We already know only 20% of those IP addresses will be used more than once to commit click fraud, so that reduces the effectiveness of IP blocking to 1%. In other words, by blocking the IP addresses which have previously been used to commit click fraud, you’re still going to miss 99% of the fake clicks on your ads."
Vanes recommends a different approach for dealing with click fraud. "We’ve been extremely successful at removing almost all click fraud from our customers’ ad campaigns. We use a four step process which is both simple and highly effective.
"Click fraud isn’t random, but instead targets high value ad keywords. Polygraph monitors which keywords are being targeted by click fraud criminals, so we can warn our customers whenever their ads are at risk. By simply removing the high risk keywords from their ad campaigns, our customers are able to avoid being targeted.
"We monitor most click fraud gangs, so we can see which websites they’re operating. We provide this list to our customers, so all they have to do is block these websites from seeing their ads. This strategy alone dramatically reduces the risk of click fraud.
"Our bot detection software is world class, better than most ad networks, so when we detect a bot clicking on our customers’ ads, we’re able to provide the date and time of the click, where the click came from, and why it’s fraudulent. Our customers then pass this information to the ad networks to get a refund. We have helped advertisers get six figure refunds from ad networks. "Finally, since Polygraph is able to see every fake click, we know which ad networks are doing a good job at stopping click fraud, and which ad networks aren’t. We advise our customers to use this data to move their ad spend away from the poor performing ad networks.”
According to Vanes, if advertisers still want to use IP blocking as a click fraud prevention strategy, Polygraph can help. “We understand some customers want to use IP blocking, so we offer this service, if requested. However we strongly recommend our customers follow our advice and don’t rely on IP blocking to stop click fraud.”