23 per cent of ad networks show ad fraud levels over 20 per cent

Almost a quarter of ad networks demonstrate significant levels of fraud, with 35 ad networks showing levels of 50 per cent or greater, a new study by Tune has revealed. The research analysed 24.3bn clicks on more than 700 ad networks, and demonstrated that ad fraud remains an endemic problem in the industry.

Estimates of the cost of ad fraud vary wildly, with digital security company WhiteOps suggesting that $7.2bn (£5.6bn) was lost last year, while ad verification firm Adloox claim that marketers will lose $16.4bn to ad fraud this year. That’s a lot of wasted inventory and fraudulent clicks.

Tune made use of dozens of measures to identify various forms of mobile ad fraud, including timing, geography, location, volume, device fingerprints and more. The study found that across all ad networks, the average level of fraud is 15.17 per cent.

More concerning, however, was the number showing significant levels – 20 per cent or higher. 23 per cent of those studied showed levels of fraud above 20 per cent, while a number of outliers showed huge levels, with eight ad networks dealing with 100 per cent fraudulent inventory.

While most ad networks don’t show anything near this level (over half of the 500+ ad levels with significant traction and traffic had less than five per cent fraud) it still shows that a number of bad actors remain in the ad tech industry.

You can find out more about the results, and Tune’s advice for safeguarding against ad fraud, by downloading the research for free here.