The Numbers That Decide
Every mini-game lives or dies on a handful of ratios, and most teams are guessing at what good looks like. Benchmarks are not gospel — a puzzle game and a battle royale are different animals — but you should know the shape of a healthy curve before you call a launch a failure.
Q: What is a realistic retention benchmark, and why does day 1 matter most?
For a decent mini-game, D1 retention in the 35 to 45 percent range is solid, D7 around 15 to 20 percent, and D30 in the 6 to 10 percent band. Day 1 is the loudest signal because it tells you whether the first session delivered on the promise of the ad that bought it. If D1 is under 25 percent, the problem is almost never your day-7 content — it is a mismatch between what you advertised and what you shipped. Fix the hook before you fix the meta.
Q: How should I read ARPU and ARPDAU without lying to myself?
ARPU is revenue per user over a chosen window; ARPDAU is revenue per daily active user, and it is the number that actually maps to live operations. A healthy mini-game might run ARPDAU between 3 and 10 cents depending on geography and ad load, but the figure is meaningless without context. Compare ARPDAU against your user-acquisition cost per install, and remember that blended averages hide the whales. The honest version of the metric is LTV against CPI, not a vanity daily number.
Q: What is a good LTV to CPI ratio, and when do I panic?
The rule of thumb is LTV to CPI of at least 1.2 to 1, and preferably closer to 1.5, to leave room for the 30 percent platform cut, payment fees, and the fact that half your cohorts will underperform. You panic when payback period — the days to recover CPI — stretches past 60 to 90 days on a genre that loses most of its users by day 30. At that point you are not acquiring customers, you are financing them.
Q: Do benchmarks differ enough by region that I should track them separately?
Heavily. The same game can show 40 percent D1 in Southeast Asia and 25 percent in North America, and eCPM can differ threefold between the US and emerging markets. If you blend those into one dashboard you will make wrong calls. Track retention, ARPU, and CPI per cohort and per geo from day one, because the decision to scale a campaign is always a regional one, not a global one.
| Metric | Healthy mini-game | Time to worry |
|---|---|---|
| D1 retention | 35 to 45 percent | Under 25 percent |
| D7 retention | 15 to 20 percent | Under 8 percent |
| D30 retention | 6 to 10 percent | Under 3 percent |
| LTV to CPI | 1.2x or higher | Near or below 1x |
Benchmarks tell you whether you are in the game, not whether you have won it. The teams that scale are the ones that watch the curve move, not the ones that hit a number once.