A casual hit that runs on ad revenue
Dian Dian Interactive publishes a courier-themed management sim where players build up a delivery company from a single office. It found an audience fast. The game reached number four on Google Play's free games chart, passed a million installs, and gathered more than ten thousand player reviews.
The game runs on an in-app advertising model, so revenue comes from players watching ads in the game rather than from purchases. That changes the math on user acquisition. An install only pays off if the team bought the right user in the first place, and if that user sticks around long enough to keep seeing ads. Installs alone tell you nothing about which of those two things happened.
Judging channels by what players actually do
Before, acquisition data and in-game behavior lived apart. Attribution told the team where an install came from. It said nothing about whether that player opened the game again, kept playing, or ever generated ad revenue. Two campaigns could look identical on install count and behave nothing alike.
Using ThinkingAI, the team joined its acquisition attribution with in-app behavior data. Now each channel, campaign, and creative could be judged by how its users actually behaved, not by how many installs it delivered. Sources that looked cheap but brought players who left immediately stopped looking cheap. The team could operate each traffic source on its own terms.
One player, from install to ad revenue
The bigger gain came from mapping a single id across the acquisition side, the in-game side, and the monetization side. With those three connected, the team could follow one player through the whole chain: where they came from, what they did in the game, and how much ad revenue they eventually produced.
That view made ad placement a decision rather than a guess. The team tuned placements by traffic source, so the model earned where it should. Instead of averaging revenue across every user, they could see which sources brought players worth acquiring under an ad model, and fund those.
Deciding the same day, not the next
Reporting used to arrive a day late. The team saw yesterday's numbers today and acted on them tomorrow. On ThinkingAI's real-time ETL and ad-hoc query setup, that next-day delay became real time.
Marketing, product, and operations now work from the same live picture. When a channel starts underperforming or a placement change moves revenue, they see it as it happens and adjust, rather than finding out a day later.