Google Play’s 12 Testers for 14 Days, as a Solo Developer
On the App Store, Fuelvo went from upload to public in a few days: one review and a manual release. On Google Play the same app needed a month, and most of that time wasn’t review. It was a rule that every new personal developer account has to go through.
The rule
Since November 2023, a personal Play developer account created after 13 November 2023 can’t publish to production straight away. According to Google’s help page, it first has to run a closed test with at least 12 testers who have been opted in continuously for at least 14 days, and then apply for production access. Testers who opt in and leave before 14 days don’t count, and someone who leaves and rejoins starts their 14 days again.
Organisation accounts are exempt. For a solo developer without a company, there’s no way around it, so it’s better to plan for it than to resent it.
How it went for Fuelvo
| Date (2026) | Step |
|---|---|
| 6 October | 1.0.0 uploaded to the internal and closed tracks |
| ~8 October | Closed test starts with friends and family |
| 13 October | Meanwhile, the iOS app goes live on the App Store |
| 22 October | 14 days complete; production access requested |
| 25 October | Access granted (Google says to allow about a week) |
| 28 October | Version 1.1.0 at 100 % in production, after a 10 → 50 → 100 % staged rollout |
A few decisions made the test work:
- Recruit well above 12. About 20 people joined. Some testers stop using the app, change phones or leave the test without noticing, and if the count drops below 12 the clock no longer counts. With 20, a few drop-outs didn’t matter.
- Use a feedback channel people already use. The opt-in link and the instructions went out over WhatsApp, and that’s where the feedback came in. A form nobody opens produces nothing to report.
- Keep shipping during the test. The test didn’t freeze development. Version 1.1.0, with the reworked price map, went to the closed track halfway through, so the testers ended up testing the version that would reach production.
- Keep everyone in until Google answers. The application asks about the test that has already happened, but there’s no reason to let the group break up before the answer arrives.
What the test was actually good for
I started out treating it as a formality. It turned out to be the first time the Android app ran on phones I didn’t own, with real fixes from real locations, on networks I don’t control.
-
Android vitals had real data: 0 crashes and 3 ANRs across about 20 users. An ANR is the kind of thing an emulator rarely shows you.
-
Play pointed out something I hadn’t configured. On each upload, the Console warned that the bundle contained native code without debug symbols. The native libraries come from dependencies (SQLite’s bundled driver among them), and without symbols a native crash in Play vitals shows up as addresses. The fix is one line in the Android Gradle Plugin, and it ships in the next update:
android { buildTypes { getByName("release") { ndk { debugSymbolLevel = "SYMBOL_TABLE" } } } } -
It forced a rehearsal of the full release path: upload, track, promotion, release notes in each language, rollout percentages. By the time production arrived, none of it was new.
The production access application
The form has three sections: the closed test, the app, and how ready it is for production. Among other things, Google asks you to summarise the testers’ feedback, how you collected it, and what you changed because of it. Two things helped:
- Answer in English and be specific. Name the versions the testers received and what changed in each one; that reads better than general statements.
- Answer honestly about the size of the test. Friends and family is a legitimate source of testers for a solo developer; inventing a larger beta programme isn’t.
The answer came in three days.
Planning a launch around it
For a new personal account, the minimum time to production is:
14 days of testing + the application review (Google says up to about 7) + the review of the production release, and the 14 days only start once 12 people have opted in.
That changes the plan:
- Open the closed test as soon as there’s a stable build, not when the app is “finished”. The clock runs while you keep building.
- If the launch is on both platforms, start on Android first. Fuelvo’s iOS app was live on the App Store for two weeks before the Android app could even ask for production.
- Don’t commit to a launch date or any marketing until production access is granted. Until then the date isn’t in your hands.
The rule is clearly designed to filter out low-effort apps, and for a solo developer it costs about a month. But it isn’t wasted time if you use it as what it is: a real closed beta, run before the public sees the app.