mobile ad fraud

Mobile Ad Fraud: 4 Best Practices to Fighting Click Spamming

By Andreas Naumann | November 8, 2017

Research from The&Partnership and Adloox, an ad verification company, estimates that advertisers could stand to lose $16.4 billion this year due to fraudulent traffic and clicks manufactured by bots on both desktop and mobile. Mobile ad fraud is unfortunately on an upswing heading into 2018, and is currently projected to rise to $50bn by 2027.

Fraudsters use several tactics to steal ad dollars away from advertisers, including click spamming (the most common), click injection, and fake installs. In this post, let’s take a closer look at click spam and what you as marketers can do to protect yourself from it.

Paying Fraudsters for Organic Users
Click spamming (also known as click flooding) is a type of mobile fraud that sends a large number of fraudulent clicks on an ad without the user knowing, in hopes of capturing the last-click prior to an eventual “organic” install. Click spamming is so rampant in mobile that any app with a sufficient user base is very likely wasting ad dollars buying their own organic installs.

A subset of this growing problem is the push by ad networks to over-credit and over-attribute network traffic. Some ad networks will also execute clicks for the app user without their knowledge to upgrade the network’s chance of getting installs attributed to them. For example, video networks send clicks instead of impressions to vastly improve their attributed conversion count. Without an effective real-time filter against click spamming, advertisers are fooled into believing those numbers.

Any actions taken by publishers or networks that create clicks where no human user clicked ad media with the intent to find out more about the advertised product is in breach of the IAB’s measurement guidelines and should be considered fraudulent. Background clicks on websites hosting illegal content are also considered fraudulent because the end-game is the same: cheating attribution providers into attributing an install.

4 Best Practices to Fighting Click Spamming

So, we’ve identified the problem. But what’s the solution? A mixture of vigilance and use of Adjust’s Fraud Prevention Suite can go a long way. Beyond that, here are four tips to follow when it comes to taking on ad fraud.

Check channels which are too good to be true
If you have cheap, new channels outperforming the known, more expensive ones, pay close attention to the full user conversion funnel. It could be a sign of click spam inflating results, as well as quality of the installs. It could also be fake installs inflating the volume on one or more sources.

Checking the funnel from beginning to end is a very effective method of identifying click spam and click injections. To find exploited incentivized campaigns and fake installs it makes sense to expand this exercise to the backend of the funnel as well.

Monitor traffic patterns
When statistical traffic patterns don’t match the promised traffic type, this can be an indicator of ad fraud. For instance, if supposed high-quality display traffic gets very low conversion rates, there could be fraud involved.

This same approach can be taken in an effort to unearth compliance issues on your campaigns. For instance, non-incentivized traffic that’s converting well above typical performance benchmarks should be checked for low post-install metrics and benchmarked against trusted sources to find out if incentivized traffic is being mixed into the source traffic.

Get granular
Drill down as much as possible on your acquired traffic, such as looking specifically at publisher IDs, to see if different low-quality sources have been combined on a campaign to form a statistically average campaign. For example, review a mix of click spam and click injections, or click spam and fake installs or incentivized traffic. Combinations like these can average out to build something that looks like a strong and well converting high-quality campaign. This goes double for checking campaign performance in a graphical analysis. So whenever you create histograms to look for indicators for click injections or click spam, try and make sure you do so on the deepest granularity available.

Diversify your portfolio
Business relationships and loyalty are important for a reason and I will not try to argue against them. What is important though is to be aware that perpetrators of ad-fraud are not loyal or discriminating. They will take money from any advertisers that hand it out and they will go through any exchange or network to get access to new or better revenue streams.

This manifests in the complete uselessness of blacklists of any kind that are supposed to help advertisers manage their fraud levels. It is a game of whack a mole on a massive scale. Most fraudsters operate dozens of accounts with the same amount of networks and exchanges simultaneously to mitigate their risk of financial loss if found out.

Lastly, look at the networks and partners you are buying traffic from and evaluate the quality of the traffic they are generating. If a network shows a high amount of fraudulent traffic, work with them to remove it. And if you are unable to, drop them. But understand that as networks become more savvy in removing fraud from their networks, fraudsters will abandon them and look for new sources of income, making the network once again a potential good source of traffic and quality users.

Guest Blog: Well versed in large-scale data analysis, Andreas Naumann from Adjust brings more than nine years of experience working with advertisers in fraud prevention at multiple European leading ad networks including Zanox, Trademob and Glispa.

Headquartered in Berlin, Adjust is a mobile attribution and analytics company that provides app marketers with a comprehensive business intelligence platform.