Monetizing ad fraud: the fraudsters perspective

The IAB and Ernst & Young estimate that $4.6 billion is lost due to ad fraud / non human traffic annually in the US. This is only one estimate, there are numerous estimates, with some reaching over 10 billion. However, most of these figures are somewhat down to guesswork, due to the ever changing nature of ad fraud. Ad fraud is an arms race that is constantly evolving with improved levels of sophistication and new bot networks popping up daily.

But what makes ad fraud so attractive to fraudsters? Why do they choose this medium of crime over others? One of the main attractions is the vast sum of money that can be made from ad fraud, with fraudsters having the potential to pocket billions. There is also no scaling issue with ad fraud, take stealing cars for example, you need to constantly find good cars to continue with your chosen method of crime. Whereas ad fraud is often viewed as a money tree for cyber criminals. As soon as cyber criminals start generating traffic and getting paid for it, they can keep doing that indefinitely. The only restrictions for fraudsters are the technical limits of their botnet.

Fraudsters continue to target online advertising due to the limited risk of penalty associated with ad fraud. Unfortunately ad fraud falls in the grey area in many jurisdictions, making it very hard to prosecute fraudsters. This results in many fraudsters viewing ad fraud as very low risk, with very high payout potential.

To understand why ad fraud has flourished in online advertising we need to look at how we measure the success of advertising. Part of the reason ad fraud has become such a big problem is because historically rather than brand lift or conversions, the industry has been looking at the wrong metrics. For too long the focus has been on using things such as click through rate (CTA), or time spent on a page as a proxy for effective advertising. The problem with using CTA, time spent on a page or any of these types of behavioural metrics is that they’re very easy to game.

There are three main methods that we employ to try and tackle the ever evolving problem of ad fraud: behavioral and network analysis, browser and device analysis and malware and targeting analysis.

Whilst measurement and analytics companies have built sophisticated products that tackle the threat of ad fraud, to truly tackle ad fraud we need to see a shift in how we measure as an industry. Don’t you want to stop wasting thousand of dollars of ad spend on non human traffic that will never lead to a conversion? To tackle ad fraud the industry needs to stop relying on measurement metrics and make a move towards measuring brand lift and conversions. In doing this you will help to ensure that your advertising spend isn’t being wasted on non human traffic, and instead goes towards influencing real humans interested in purchasing your products.