Our system determines that activity is invalid based on our internal detection algorithms and analysis, as well as comparing IPs, devices and users to hundreds of millions of data points within the PPC Protect system.
Just like your credit card's fraud department we can’t disclose our systems ins and outs (because that would allow invalid users & fraudsters to know exactly how we catch them!) but we take in to account over 150 different factors when identifying fraud, including (but not limited to):
- IP address/network
- Device Type
- Color Depth
...and many more. This helps us build a profile of if a user is a real, genuine user, a bot or an invalid user.
In addition to that, we also take the following in to account in our algorithms:
- Click frequency
- VPN usage
- Known click farm ranges
- Known fraudulent IP ranges
Combined, these help us to build an extremely accurate picture of which users are legitimate and which are invalid. We do also sit on the side of caution - if we aren't sure if it's invalid or not we'll mark it as suspicious and monitor closely, but we won't block it. This is because having the lowest possible rate of false-positives is extremely important to us.