Founded in 2009, ZiMAD develops and publishes games for mobile, PC, and social networks.
Their popular games include Magic Jigsaw Puzzles, Dig Out!, My Museum Story, Maggie’s Murder Mystery, Magic Solitaire and Bubble Birds.
After switching to the ironSource mediation platform, the monetization team at ZiMAD removed some ad units and only served rewarded video ads and banner ads on their game Dig Out!. They were interested in re-adding interstitials to the mix, as it was before moving to ironSource mediation, as a way to boost revenue, but were cautious because of the potential impact on retention.
Naturally, rather than rolling interstitials out to all users and hoping for the best, ZiMAD preferred to set up a test with a small percentage of users.
Solution
ZiMAD was pleased to hear about the launch of ironSource’s ad monetization A/B testing tool on the ironSource mediation platform. Finally, they could test the retention and revenue impact of interstitial ads on Dig Out!.
ZiMAD set up their first ad unit test on Dig Out! - showing only rewarded video and banners to users in group A and introducing rewarded video, banners, and interstitials on Day 1 to users in group B. Their key KPI for the test was overall revenue.
“The A/B test was simple to run and even simpler to analyze.”
Results
Ultimately, implementing interstitials alongside rewarded video and banners increased revenue for ZiMAD by 17% and ARPDAU by 18%, with only a 2.5% drop in D7 retention.
There was also very little effect on rewarded video engagement - just a 3% decrease in impressions and an even smaller decrease in user rate (impressions per DEU). Essentially, despite the increase in system-initiated ads, users were still choosing to opt into rewarded video, as these rewards were valuable to them.
“After switching to ironSource's mediation, there was no way we could have known how well interstitials would have performed on our game without the ironSource testing tool.”
Because test controls run server-side in the ironSource mediation platform, minimal code updates were needed from ZiMAD’s R&D team - now it’s easy for ZiMAD to set up future tests quickly.
The dashboard was also a plus, allowing them to review and analyze full funnel test results at a glance, from performance metrics like revenue and ARPDAU to user metrics like engagement and retention. In fact, ZiMAD already has their next few A/B tests planned out, like running a manual waterfall against an automated one.
