How to save big by blocking fraudulent installs and in-app events

Background

Fetch Rewards is a mobile shopping platform that rewards shoppers for buying the brands they love.

The Fetch Rewards app gives users the easiest way to save on everyday purchases by simply scanning their receipt. For their brand partners, the platform allows them to understand a 360-degree view of shopping habits, and to meaningfully reward a customer’s individual loyalty.

Shopping apps are already at a higher than average risk of fraud – with a 46% fraud rate compared to a 22% fraud rate across all app types. The Shopping category sees nearly $1 billion in financial exposure to install fraud every year, second only to Finance.

The Fetch Rewards UA Team focused the majority of their spend on top-of-the-funnel activities.

The key to Fetch Rewards’ success depended on driving enough users to integrate the app into their purchasing behavior in order to become “qualified” – defined as a user who scanned more than one receipt more than once per month. The team found that too many users were becoming disqualified after their initial app download.

To realize a viable number of qualified users with the drop-off rate they were seeing required running campaigns on a large scale.

Challenge

Although Fetch Rewards recognized that large media partners drove a high volume of traffic, they didn’t want to put all of their eggs in one basket. Instead, they embraced a diverse campaign strategy, which included programmatic, affiliate, content, and influencer campaigns.

This strategy provided them with the volume they needed to be successful at scale, but left them exposed to ad fraud.

Fetch Rewards’ previous attribution provider did not sufficiently protect them from fraud.

The team was running reactively. If they noticed a spike in overall fraudulent activity, their only recourse was to reach out to every one of their network partners and hope that one raised their hand to claim the fraud. This was not a scalable solution. Fraud was eating at Fetch Rewards’ advertising budget, polluting their data, and wasting time that they didn’t have.

Solution

Protect360, AppsFlyer’s fraud protection solution, offered real-time and post-attribution fraud detection ensuring Fetch Rewards only paid for legitimate installs. Protect360 leverages AWS cloud storage and computing solutions to apply advanced machine learning algorithms on the industry’s largest mobile attribution database.

Protect360’s custom click-to-install time (CTIT) setup meant that the fraud detection rules could be configured to reflect their app’s unique exact time-to-download and install requirements, providing protection against even the most sophisticated fraudulent bot behavior.

Perhaps more importantly, the Fetch Rewards team not only benefitted from full insights into blocked installs and fraud savings through their AppsFlyer dashboard, they were also able to proactively communicate this with their network partners. Real-time reports were shared directly with all of Fetch Rewards’ network partners, acting as a trusted source of truth for all parties.

This increased transparency was a game changer when it came to taking joint action against fraud, from removing fraudulent actors to processing refunds for affected traffic.

Real-time and post-attribution fraud detection from AppsFlyer allowed Fetch Rewards to run user acquisition campaigns at scale without fear of malicious actors.

Results

In 12 months, Protect360 saved Fetch Rewards more than $750,000 in fraud, $590,000 in realtime fraud savings, and $175,000 in post-attribution fraud savings.

During that same period they saw a total of 330,000 blocked fraudulent installs, 150,000 blocked post-attribution fraudulent installs, and 160,000 blocked fraudulent in-app events.

Looking for an MMP? Get your free assessment template today

Thanks for your download!