The impression you did not sell
A mini-game can have a perfect core loop and still earn poorly — if its ad inventory is filled badly. Every impression that goes unsold, or sells at a low bid, is revenue lost forever. The craft of ad monetization is therefore less about the game and more about the plumbing: how many networks compete for each slot, and how intelligently they are ranked.
Bidding has become the default
Real-time in-app bidding now accounts for roughly 80% of a typical ad setup, because it lets advertisers bid on value-based KPIs like CPA and ROAS through machine learning. But the winning pattern is not bidding or waterfall — it is both. According to Tenjin's 2025 monetization report, about 85% of top-grossing mobile games run hybrid mediation, where bidding runs first and a waterfall catches the leftovers. That hybrid approach maximizes both fill and price.
| Setup | What it optimizes | Typical result |
|---|---|---|
| Manual waterfall | eCPM order only | Stagnant, lost impressions |
| In-app bidding | Real-time competition | ~80% of setups |
| Hybrid (bidding + waterfall) | Fill + price | 85% of top games |
The measurable lift
The gains from moving off a manual waterfall are concrete. A mid-sized puzzle studio that switched to automated bidding reported a 15% increase in eCPM and a 20% increase in ARPDAU within 30 days. Across cross-border casual teams, adding mediation with a multi-layer waterfall is documented to lift eCPM by 15%+ and fill rate by 12–22%. The lever is competition: more demand sources bidding on each impression.
Format choice matters too. Rewarded video consistently posts the highest eCPM of any ad style — roughly $18–25 in the US market versus $3–5 in India — while interstitials and banners sit lower. Mini-games that front-load rewarded ads and cap frequency keep players happier and units more valuable.
One runtime, one mediation layer
The practical blocker for most studios is integrating and tuning a mediation stack inside every host app's ad SDK. A shared runtime collapses that: the game requests an ad through one interface, and the runtime runs the bidding-and-waterfall logic centrally, swapping demand partners without touching game code. The same optimized stack serves every host, so a tuning improvement in one place lifts revenue across the whole catalog.
What this means for your roadmap
If you build mini-games, treat mediation as a first-class system, not an afterthought — segment by player LTV tier, device, and session depth, and optimize for ARPDAU rather than eCPM alone. If you run a host app, providing a centrally-managed mediation layer through the runtime means every title benefits from best-in-class fill without each team becoming an ad-tech expert.
Cross Mini App runs one mediation layer for the whole runtime, so a single optimization compounds across every game and every host.
