App monetization: subscriptions, fees, ads or freemium
Compare app monetization models, subscriptions, transaction fees, ads and freemium, and learn how to choose based on the way your app delivers value to users.
Learn why people abandon apps, how to measure user retention with cohort analysis and which onboarding and notification changes bring users back for more.

People abandon an app when they do not reach the promised value quickly, when it never becomes a habit, or when it annoys them more than it helps. To improve user retention, measure exactly when people stop, broken down by cohort, and fix that specific point: onboarding, reasons to return or notification overload. Spending on acquisition before solving this is pouring water into a leaking bucket.
Download counts and total registered users say nothing about retention. The useful metric is cohort retention: of everyone who installed the app in the same week, how many were still active on day 1, day 7 and day 30.
Two details make a big difference. First, define "active" as someone who took a valuable action, such as booking, buying or logging something, not just opening the app. Second, measure at the product's natural frequency. A messaging app should be used daily; a tax filing app, once a year. Judging both by the same yardstick leads to wrong conclusions.
Here is an example with made-up numbers, just to show how to read it:
| Install cohort | Day 1 | Day 7 | Day 30 |
|---|---|---|---|
| Week 1 | 40% | 18% | 9% |
| Week 2 | 42% | 20% | 11% |
| Week 3, after new onboarding | 51% | 27% | 15% |
What matters is the shape of the curve and the comparison between cohorts. A curve that drops and then levels off means a group found real value in the product. A curve that slides to zero means nobody stays, no matter how many people come in. And recent cohorts doing better than older ones show your changes are working.
The reasons repeat across almost every kind of app:
That last reason is natural and does not always need fighting. An app for moving house, for example, is used intensely for a short period by design. In that case the goal changes: deliver well and earn referrals.
The first few minutes decide a large part of retention. The job of onboarding is not to present every feature; it is to get the person to their first valuable experience as fast as possible.
To find out what that experience is, compare users who stayed with those who left and look for an action the first group took early and the second did not. It might be completing a first order, following a few accounts or logging a first expense. Then shorten the path to it.
Practices that usually help:
After first value, people need reasons to open the app again. Those come from two places: external triggers, such as notifications and emails, and value that builds up inside the product.
Accumulated value is what makes the app more useful the more it is used: history, saved preferences, connections, balance, progress. Someone with six months of data in a finance app thinks twice before switching.
Notifications work when they are relevant to that person at that moment. Segment by behavior, respect time zones and quiet hours, cap frequency and let people choose what they receive. Ask for notification permission after showing value, not on the first screen. A permission denied early is hard to win back.
Data shows where people stop; conversations show why. Combine both:
Win-back campaigns have their place, but only after you fix the reason people left. Bringing someone back to the same experience that drove them away is wasted money.
Retention improves through cadence, not a one-off push. A simple routine: every week, the team looks at the latest cohort curve, picks the biggest drop, proposes a change, ships it to part of the user base and compares it with the group that did not get it. One hypothesis at a time, measured to the end.
Bridgemoon builds apps and helps digital businesses scale. That is why event tracking is planned from the first version, so user retention can be followed by cohort from launch day instead of being reconstructed months later.
It depends heavily on the category, natural usage frequency and audience, so generic industry figures are of little help. The most useful benchmark is your own history: new cohorts should retain better than older ones. More than any single number, look for a curve that flattens instead of falling to zero.
Both. Relevant notifications at the right moment bring people back to something that matters to them. Generic, frequent ones lead people to switch notifications off or uninstall the app, and then you lose the channel for good.
Retention measures whether people keep using the app over time. Engagement measures how intensely they use it, such as frequency, time spent and actions per session. An app can show high engagement among a few loyal users and low retention overall, so it pays to track both.