Basic launch metrics explained
Basic launch metrics explained
During basic launch your game is live with ads off. We use that traffic to learn whether the game deserves a full launch. Everything is measured automatically; no SDK is required, though the gameplayStart event makes the data more precise.
How long it lasts#
A basic launch ends once the game has been live for at least 7 days and has reached 500 plays, or automatically after 21 days, whichever applies first. Then a reviewer decides: promote to full QA, extend for another round, or end it. Updates you upload during basic launch go live right away. The console's launch card shows days live, plays and each metric against its target. Statistics update hourly, and next-day retention appears a day later than the other numbers.
The numbers#
| Metric | What it measures | On target | Weight |
|---|---|---|---|
| Average play time | Play time per session (sessions are capped at 30 minutes) | 10 minutes or more (half credit from 5 minutes) | 30 % |
| Day-1 retention | Share of players who return the next day | 15 percent or more is full marks; 10 to 15 percent is close; 5 to 10 percent earns a little | 30 % |
| Conversion | Share of views whose session lasts at least a minute | 80 percent or more (half credit from 60 percent) | 20 % |
| Click-through rate | Share of people who saw your tile and opened it, ranked against other games in your category | The 50th percentile or better (half credit from the 25th) | 20 % |
Load time and initial download are not scored, but they drive conversion: see first frame fast.
The portal combines these into a score between 0 and 1 and gives the reviewer a recommendation: promote at 0.75 or more (with play time and retention both above zero), end below 0.4, otherwise extend. With fewer than 200 plays the data is too thin and the recommendation is always extend.
The score is advice, never a decision. A reviewer weighs the numbers together with quality, so meeting every target does not guarantee promotion and missing one does not always rule it out.
Improving each metric#
Conversion is mostly about the first minute: load time, a game that starts quickly, no wall of menus. See first frame fast. If conversion is low and load time is high, shrink the build first.
Play time comes from depth. Add a reason to play another round: progression, unlocks, variety, goals the player sets.
Retention comes from reasons to return: daily rewards, progress saved through the data module, things that change from day to day, and a first session that ends on "I want to try that again".
CTR is the cover and the title, compared with other games in your category. Use artwork experiments once your game is in full launch, and follow the cover specs now.
Reading the dashboard#
- Compare against games in your category, not the whole portal.
- A small sample swings widely. Judge trends over days, not hours.
- Split by device. A great desktop game with poor mobile conversion often has a touch-control or size problem.
- If a number looks wrong, check the preview tool and the console log for errors first.
After basic launch#
A promotion needs the full SDK integration, so prepare it while you wait: gameplay start and stop, ads at natural breaks, cloud save. See the requirements overview.