How I Built Ad Fraud Detection Dashboard

Detecting fraudulent website traffic

Online advertising is still a way to monetise millions of websites. So it is not surprising that with the growth of online advertising, there is a stronger temptation to defraud advertisers’ money by fake traffic. Click farms can generate massive amounts of worthless traffic and ad impressions, by automatically clicking on websites and apps in large numbers.

Unless stated otherwise, all pictures in the article are by the author.

The technology needed to create a farm of clicking bots is cheap and available. In one of the simplest models, simply use a cloud computing service such as AWS or GCP, which give you access to a multitude of IP addresses and servers.

Ad networks are trying to combat this problem and penalize dishonest website owners. Knowing this, competitors or pranksters can redirect fake traffic to the victim’s website, causing penalties from the ad network, the threat of lost revenue and additional costs.

In this article, I will show you how to build a fake traffic alert system for ad fraud detection. We will create an engine that, based on server logs and a database of IP addresses, will assess the quality of the traffic that the campaign delivers. Finally, we will verify the performance of our system by commissioning a real campaign and assessing whether the traffic from it was legitimate.