A major gaming publisher had recently released a new title. This title’s initial user engagement and retention looked promising. In an effort to scale their active user-base, they planned a massive, new, strategic and highly targeted campaign. This campaign launched on a Thursday, with the first stage projected to run for two to three weeks.
This advertiser’s new campaign launched on a Thursday morning. Over the first 24 hours, the app’s marketing team and their AppsFlyer Success Manager paid close attention to their campaign performance, hoping to catch any issues before heading out for a well-deserved weekend. Looking at their real-time data in AppsFlyer, the team noticed that the campaigns did not appear to be performing as expected. For example, while one network’s reported installs were generating strong in-app engagement, this advertiser’s app did not report an overall growth in in-app engagement as expected.
The team began reviewing this gaming app’s full raw data reports and discovered a number of anomalies. For example, while one network had a particular success generating new installs, many of these installs came from suspiciously similar devices. Another network looked to be driving strong in-app engagement, but much of this engagement seemed to come at the expense of existing users.
A few hours of investigation revealed that the first network had a few publishers utilizing illegal bots, and the second network was misattributing re-engagement ads as new installs.
By quickly identifying and addressing their fraud, this advertiser saved $300,000 over their first weekend alone.