The headline number: 168
168 = 12 × 14. That's Google Play Closed Testing's published requirement.
But the way Google computes it isn't "have you accumulated 168 distinct user-engagement-days?" It's stricter:
Required: 12 unique testers, each daily-active for 14 consecutive days
NOT counted: 24 testers active for 7 days each (= 168 too)
NOT counted: 12 testers active for an average of 14 days each (gaps allowed)
NOT counted: 12 testers in days 1-7 + 12 different testers in days 8-14
Why the 14-day continuous window matters
Real user behavior follows weekly patterns — usage spikes on weekends, dips on weekdays, varies by tester's local timezone. A 14-day window captures two complete weekly cycles, giving Google enough variance to detect a real-human signature.
Fraud rings can't fake this easily because:
- Bot accounts produce flat usage patterns (no weekend spike)
- VPN-rotated accounts produce inconsistent timezone activity
- Paid-farm accounts churn out — most don't make it past day 5
Google's machine-learning fraud model is trained against this pattern. Match it, you pass; deviate, you fail.
Per-tester daily activity = what counts
Google's exact definition (paraphrased from their developer docs): a tester "counts as active for day N" if they:
- Have the test build installed
- Open or use the app at least once in day N's 24-hour window
- Did not uninstall before day N ends
TestHive's daily check-in mechanism mirrors this — each tester submits a screenshot per day showing they used the app. The screenshot is the verifiable artifact that proves daily activity to Google when you submit your final report.
The math when you drop testers
If a tester drops out on day 9 of a 14-day window:
Days 1-8: tester counted as active (8 days credit)
Day 9 onward: tester not counted
Your group: now 11 active testers on day 9
You're below 12 testers on day 9, so Google resets the 14-day counter to zero. The 8 days of credit don't carry forward; you have to start over with 12 active testers on day 1.
This is why tester reliability matters more than tester recruitment. Getting 12 testers is easy; keeping 12 testers active for 14 consecutive days is the actual hard problem.
How TestHive aligns with tester-day
Every TestHive Campaign aligns 1:1 with Google's tester-day model:
- 12 testers required to start (matches Google's minimum)
- 14-day continuous window (matches Google's window)
- Daily payouts gated on approved check-ins (incentivizes daily activity)
- Replacement tester within 24h if anyone drops out (your counter doesn't reset)
That last point is the one developers underestimate. Without auto-replacement, even with 12 highly engaged testers, you have a ~15% chance of someone dropping out for personal reasons during the 14 days. TestHive's incident-response queues a replacement before the day-N counter trips.