Disincentivizing mobile ad fraud
Product Content Strategist
Posted Jul 3, 2017
Marketers are oftentimes incentivized to let fraud happen. Everyone in the office wants to know - how many installs did we get today? This week? This quarter? In some cases, UA managers get paid a bonus based on reaching their install numbers, or their job security is tied to hitting those goals.
This travels up the ranks, from director to vice-presidents, maybe all the way up to the CMO. Sometimes, that devotion to KPIs is tied back to the app’s original funding sources. Funds judge their progress based on the number of users acquired through marketing campaigns, and nobody looks at the rest of their KPIs with much scrutiny. If no one cares about conversion rate, retention rate, cost per acquisition or any other outcome, it becomes very attractive to just allow fraud. To keep everyone happy, marketing managers knowingly buy installs with click spam, month after month, to reach their goals. Within their companies they’re seen as heroes, but fraudulent activity has accomplished most of the work.
After years of distortion from rampant fraud, benchmarks across the industry and those KPIs have become distorted. Without fraudulent means, there’s no other way to reach those benchmarks. Fraudsters have become necessary within the industry, delivering a service that helps maintain the status quo.
But that’s not how we play. We know that changing the role of fraud in the industry is a long game. It depends on players like us disincentivizing the behaviors that encourage fraud and giving you more control over your data and insight into what’s happening inside your app. In your dashboard you have access to a range of data that gives you a robust picture of your app’s health - far more insightful than the number of installs alone ever could be. You also have access to a number of tools to help eliminate fraud from your data set before it ever makes it in - real-time fraud prevention that’s only available at Adjust.
How Adjust’s Fraud Prevention Suite Works
The Fraud Prevention Suite is made of three different tools, or (in reality) filters - Anonymous IP Filtering, Distribution Modeling, and Hyper-Engagement Filtering. Together, they keep your budget, your data and your KPIs safe. However, if you’re in your dashboard, you’ll notice that there are only two toggles for the three products. Here’s why that is:
You can turn Anonymous IP Filtering on or off, separately from the other filter. When enabled, the toggle for the Anonymous IP Filter removes everything from anonymized sources. You’re also able to activate Distribution Modeling and Hyper-Engagement Filtering - but they must be activated together. You cannot turn one on without the other. These filters fight click spamming, a technique used to poach your organic installs. The Hyper-Engagement Filter filters out attribution to sources that send the same click many times before an install.
Meanwhile, Distribution Modeling filters out attribution of installs where statistically the overwhelming majority of clicks show no correlation to their corresponding installs. The reason for this is that if you were to turn on Distribution Modeling or Hyper-Engagement filtering without the other, you wouldn’t create any level of protection. In fact, this is one of the ways you’d be incentivizing fraud. Since click spam has two varieties (high frequency, low reach and low frequency, high reach) both filters need to be activated at the same time. Running just one would provide protection against one variety of fraud, but at the same time would incentivize the use of the other, because it would go completely unpunished.
Many other solutions on the market only detect fraud after it happens. These products run the gamut: some offer basic attribution functionalities disguised as ‘fraud prevention features’. Others rank networks and encourage the client to follow up on their own. Some provide after-the-fact data on possible fraudulent activity, rather than ferret it out as it happens. These tools could be seen as further incentivizing fraud, as they allow marketers to appear as if they are actively doing something while continuing to turn a further blind eye.
You have control over your data at Adjust. Whether you turn on our fraud prevention filters is up to you. Want to find more about how fraud affects your campaigns and strategies? Read our mobile fraud guide here.
If you’re interested in learning more about how the Fraud Prevention Suite works and the technology that powers it, schedule a demo with us. We’ll walk you through it.
Interested in learning more about how Adjust fights fraud? Download our mobile fraud guide to find out how we can help protect your budgets and data.