Whether you call it click fraud, ad fraud, PPC fraud or something else entirely, invalid traffic that runs up your advertising bill but doesn’t produce sales is bad news. Click fraud can decimate an advertiser’s marketing budget in short order. But there is a way to fight back. It is all in the data.
Web servers are set up in such a way as to create logs of all activity they serve up. By analyzing the data in those logs, advertisers can find clues that suggest click fraud is taking place. The data can be analyzed manually or using an automated software package like Fraud Blocker.
The people behind Fraud Blocker say that click fraud detection software looks at a laundry list of data points in search of anomalies that would suggest click fraud. Here are just four of those data points:
1. IP Addresses
Every device connected to the internet is assigned a public IP address. An IP address is similar to your home address, which tells the postal carrier the town you live in, the street you live on, and the number of your house. He can deliver mail directly to your door utilizing that information. An IP address works pretty much the same way in the digital realm.
Analyzing traffic via IP addresses can uncover click fraud by demonstrating an unusually large concentration of and clicks in a small geographic location. Clusters of clicks from identical or similar IP addresses are pretty clear proof of fraud. However, sophisticated scammers know how to spoof addresses or keep them completely hidden through VPNs. So you need to look at the other three data points as well.
2. Click Timestamp
Every click that redirects a web user’s browser to an advertiser’s website generates a timestamp. Advertisers can know exactly when their ads are clicked. This data point becomes valuable when automated software or human teams notice a large amount of traffic being generated in an unusually short amount of time. Large concentrations of traffic in and around the same time is indicative of click bot fraud.
The click timestamp offers important data on its own. But combining it with the next data point offers an even clearer picture of what might be going on. That next data point is action timestamp.
3. Action Timestamp
Different activities that occur on an advertiser’s website are considered actions. The action timestamp leaves a digital paper trail behind, allowing advertisers to see how long visitors were on their sites and what they actually did. So here’s the deal: if a bunch of traffic comes in very quickly but no action is being taken by supposed visitors, there’s a good chance the traffic is invalid. A lot of click timestamps with no action timestamps suggests either click bots or click farms.
4. User Agents
User agent data reveals the type of device being used to click on ads as well as the browser, OS, and more. This information is useful in the sense that it can be compared to IP addresses. If seemingly different devices are clicking on ads from the same IP address, there is a good chance that click fraud is taking place.
Combining all four data points gives an advertiser the best possible chances of revealing ad fraud. Once it is revealed, it can be prevented through blocking strategies. If there is suspicion a publisher is perpetrating click fraud, the advertiser can sever ties with that publisher. When all is said and done though, it is all about data. Data is the key to exposing and stopping click fraud.