Why your app lost its users in week one
You cannot fix retention you have never measured.
Most app installs stop being used within a week. The fixes with the largest effect are reducing time to first value, delaying the signup wall, requesting notification permission at a moment that makes sense, and using push for genuine user moments rather than announcements.
Key takeaways
- Measure day 1, 7 and 30 retention by cohort before changing anything.
- Every screen before first value costs installs — count them and cut.
- Ask for notification permission after you have shown value, not on launch.
- Push is a channel you can burn once; announcements burn it fastest.
Retention is the only app metric that compounds. Installs can be bought; a user who comes back on day thirty cannot.
Instrument this first
- 01Install → first open. A surprising share never open the app after installing.
- 02Onboarding funnel, step by step. You need the drop-off per screen, not the total.
- 03Activation. Define the single action that means the user got value, and measure the rate.
- 04Retention by cohort at day 1, 7 and 30, split by acquisition source.
- 05Notification opt-in rate, because it caps every re-engagement effort you will ever run.
Time to first value
Open your app as a new user and count the taps before something useful happens. If the answer is more than three or four, that is your retention problem, and no amount of push notifications will compensate for it.
The signup wall
Requiring an account before showing anything is the most common and most costly default in app design. Let people see the product, then ask for an account at the point where they need one — saving something, ordering something, syncing something. The conversion to registered user typically goes up, not down.
Push that people keep switched on
- Moments, not announcements. Order shipped, class starting, item back in stock.
- Frequency caps, enforced globally rather than per campaign.
- Deep links to the exact screen the notification is about; a push that opens the home screen is a broken promise.
- A preference screen, because the alternative to letting someone mute one category is them muting everything.
What good looks like in India
| Metric | Reasonable | Investigate below |
|---|---|---|
| Install → first open | Over 85% | 70% |
| Onboarding completion | 60–80% | 40% |
| Day 1 retention | 25–40% | 15% |
| Day 7 retention | 12–25% | 8% |
| Day 30 retention | 6–15% | 3% |
Categories differ enormously — a daily-use utility should sit far above a high-consideration purchase app, and comparing yourself to the wrong category is how teams talk themselves out of a real problem.