Most app budgets stop at launch day. Then the first surprise arrives: a new Android version, a store warning, a crash only some users see. An app in the store is never quite finished, and the cost of keeping it healthy is the line most founders forget to plan for.
The good news is that the work is predictable once you know what it is. Here’s what monthly upkeep actually pays for, what makes it cost more or less, and how to keep it under control.
What maintenance actually covers
1. OS and SDK updates
Apple and Google release a new version of iOS and Android every year, and each one can change how apps behave. Google Play also requires apps to target a recent version of Android. Fall too far behind and you can’t publish updates until you catch up, which is the worst moment to find out, usually when you need to ship a fix.
Updating for a new OS version is rarely dramatic if it’s done every year. Skipped for two or three years, it turns into a small project of its own.
2. Crash and bug fixes
Real users find problems no test did: a phone model you never tried, a slow network, an unusual way of using a screen. Crash reporting tells you what broke and for how many people, so you fix what hurts users most first.
Set it up so the reports carry no personal data. In HelperBook, my app for household wage records, crash reports carry no names or amounts.
3. Store policy changes
Both stores change their rules regularly: privacy declarations like Google Play’s Data Safety form, permission rules, and review guidelines. When a rule changes, the store expects your app and its listing to follow by a deadline, or the next update can be rejected.
This work is usually small, but it isn’t optional, and it tends to arrive with a date attached.
What makes it cost more or less
There’s no single monthly number, because upkeep grows with the app. The same drivers that set the build cost set the upkeep:
- One platform or two. Two apps mean two sets of OS updates and two store reviews.
- Integrations. Payments, maps, notifications and other SDKs each release updates of their own, and some stop working if you don’t keep up.
- A backend. A server needs its own updates and monitoring, on top of the app.
- How much you’re still changing. Fixing and updating is one thing. Adding features every month is ongoing development, and should be planned as that.
How often should you release?
It depends on the app. An app your team or customers use every day, with features still being added, benefits from a steady rhythm. At one company I introduced automated releases on a bi-weekly cadence, with same-day hotfixes when something broke.
A quieter app can update monthly or quarterly. What it shouldn’t do is sit untouched for a year, because that’s when the OS, SDK and policy changes pile up together.
How to keep it predictable
- Agree it before launch. Decide who looks after the app once it’s live, before you need them.
- Pick a model that fits. A monthly retainer, a set number of days each month, suits steady upkeep. Work billed by the day or week suits occasional fixes.
- Release to a few users first. Both stores let you roll an update out to a small share of users before everyone gets it, so a bad release reaches fewer people.
- Keep the release process automated. When building and publishing an update is routine, small fixes stay small.
If you’re still planning the build itself, start with what drives the cost of building an app in India.