Hitmetrix - User behavior analytics & recording

One Simple Trick To Detect Ad Fraud

Ad fraud is a significant problem, draining corporate profits and battering small startups into submission. In Q1 2018, mobile app marketers endured somewhere between $700 and $800 million of fraud exposure, according to an AppsFlyer report. That’s just one fiscal quarter and one specific area. A May 2017 report from the ANA (Association of National Advertisers) and White Ops projected last year’s fraud losses at $6.5 billion, which the ANA framed as good news due to a decline from $7.2 billion in 2016.

A 2017 QueryClick report on the subject acknowledged that programmatic advertising theoretically connects ad spending with the right audiences, at the right times and places, but in practice, it’s wide open to abuse due to complexity, and a lack of transparency. According to their survey, most brands believe that the ad ecosystem would benefit if independent trade bodies like the Interactive Advertising Bureau had greater authority to punish bad actors.

Some blockchain proponents have suggested that their favored technology could clean up this murky media supply chain, but others remain skeptical. In a Forbes article, author Paul Roberts poured cold water on the idea, stating, “I still cannot find a use with the blockchain to match the analysis and pattern recognition power of machine learning.”

Attribution providers aim to help brands figure out what is working and what leads to junk traffic, so that they can optimize their marketing campaigns accordingly. However, if you’re unsure whether your company is being victimized, there’s a simple, and fast, technique you could try.

Suspicious of ad fraud? Check out user behavior!

Recorded site visitor sessions could alleviate or confirm your suspicions. Strange behavior during a site session is likely indicative of non-human activity.

Having worked on multiple startups and digital marketing efforts in the past, I have seen this method in action. My startup was testing out a new, and somewhat dubious, ad network. Simultaneously, we had installed Inspectlet on our site in order to fine-tune CX/UX.

Inspectlet allows you to record a user’s browsing session on your site. You get to see every mouse movement, click, every letter entered into an input field, and then erased, and reconsidered. From that, you deduce why the visitor did or did not buy your product or services.

Yes, sometimes it feels a little creepy to watch a playback of a real user’s interaction with your site, but it’s a powerful tool.

And it can also expose ad fraud.

 We watched sessions to study the traffic that our campaign produced. We didn’t have any other campaigns running, which made it easy to discern the outcomes of a particular spend.

The questionable ad network charged per click. It generated hundreds of new visitors and all of them were clearly bots. A “user” would hit the page and then the mouse would move in literally the shape of a perfect square, simulating cursor activity. The bot then bounced off the site after 10 seconds. Again and again and again.

There’s a legitimate possibility that the ad network itself was in on it, but these things are difficult to prove because multiple parties can benefit from ad fraud. Bots and botnets are wreaking havoc on the media supply chain.

Online commerce has largely benefited society by connecting entrepreneurs with markets all over the world, but it can also be cutthroat, as demonstrated by the high rates of ad fraud and methods deployed.

Inspectlet offers a funnel analysis tool that can help entrepreneurs to identify prospective customers who added something to their carts, but never proceeded to checkout. Appsee is a similar tool, which lets you see your app through the user’s eyes. It automatically tracks all touch gestures on smartphones and then aggregates them into a touch heatmap. By seeing where users are focusing their attention, businesses can optimize screens, features, and UI elements.

Although these products and features are primarily tools for boosting actual user engagement, they can also be used to expose bots. This lesser-known use might empower entrepreneurs in their fight against fraudsters.

Total
0
Shares
Related Posts