The best mini-games barely update
Look at the top of the 2025 earnings charts and a pattern appears: the winners are not the ones shipping the most features. They are the ones running the most events. Mobile game in-app purchases hit roughly $82 billion globally in 2025, and the concentration is stark — 135 titles each earned more than $100 million, together accounting for about $50.8 billion. Last War alone took around $2.4B and Whiteout Survival about $2.2B. Their secret is cadence, not codebase.
| 2025 signal | Reading | Implication for mini-games |
|---|---|---|
| Global mobile IAP | ~$82B | In-app spend is the dominant pool |
| Titles > $100M | 135 (combined ~$50.8B) | A few live-ops machines capture most value |
| Top earners | Last War ~$2.4B, Whiteout ~$2.2B | Cadence beats features |
| Avg events / game | 73 → 89 (Oct peak 91) | More events = more monetization surface |
| Hybrid-casual revenue | +75% YoY (downloads −4%) | Events lift revenue without more installs |
Live Ops is a distribution problem in disguise
AppMagic's tracking shows the average number of live events per game climbed from about 73 to 89 over the year, peaking at 91 in October around seasonal pushes. Gossip Harbor runs on the order of 100 events a month. Hybrid-casual revenue grew +75% year over year even as downloads fell −4% — proof that squeezing more value from the existing audience, via events, outperforms chasing new installs.
But here is the catch for mini-games: a live event only pays if it reaches players everywhere the game lives. If each host app requires its own build, its own review, and its own rollout, your "event" becomes a logistics project. The value of an event decays with every day it is live in one host and absent in another.
Three tiers that actually work
The studios that win run a predictable stack:
- Daily — login streaks, timed crates, rotating shop. Keeps the habit alive.
- Weekly — tournaments, rank resets, weekend multipliers. Creates a rhythm players plan around.
- Seasonal — themed passes, collabs, holiday pushes. The big monetization spikes that fund the rest.
The craft is not inventing events; it is reaching the player on time, in every host, with one config.
One runtime, one Live Ops engine
A shared runtime turns Live Ops from a per-host project into a single broadcast. Instead of rebuilding and resubmitting for each app store, the runtime exposes a remote-config surface: the studio authors one event — rewards, timing, paywall, art — and the runtime pushes it to every host simultaneously. No app-store review sits between the idea and the player.
That means:
- One config, every host. The Halloween event is live in all ten hosts at the same minute, not staggered by review queues.
- No review latency. A bug in an event is fixed by config rollback, not a new build.
- One experiment layer. A/B a reward curve across the whole audience on identical definitions, then keep the winner.
What this means for developers and hosts
For a studio, Live Ops stops being the thing that breaks across platforms. You design one event and trust it to land everywhere — so you can run the cadence that the top earners run, without a release manager per host.
For a host app, a runtime with built-in Live Ops is how you keep your catalog feeling alive. Developers ship events to you on the same day as everyone else, and your users see fresh content without you lifting a finger on distribution.
Cross Mini App gives every mini-game one runtime and one remote-config Live Ops engine, so an event is a broadcast — not a build — and the revenue cadence that defines the top of the charts becomes available to every host at once.
