What an app costs per year after launch, and where the hours go
A freelance iOS developer maps the yearly cost of keeping an app alive: platform deadlines, store paperwork, renewals, small fixes and the monthly bills.

Every November I send the same kind of email to the clients whose apps I look after. It lists what changed on Apple's and Google's side over the past twelve months, what I did about it, and what I expect next. Most replies are a polite thank you. At least one asks the question I hear from nearly every client a year after launch: the app works, so what exactly are we paying for?
The short answer is that the ground under an app keeps moving. An app nobody touches doesn't stay as it was. It slowly falls behind, and then one day it can't be updated at all, usually on the day you need a fix out. In my line-by-line breakdown of what an app costs to build, maintenance got one row and a paragraph. This piece opens that row up.
Same caveat as before: the hours below are illustrations for example apps, sketched the way I would for a client. They are not market statistics, not averages and not a quote. What matters is which kinds of work exist and what makes them grow.
The calendar an app lives on
Most of the work follows the platforms' calendars, not yours. Once you know the rhythm, very little of it should come as a surprise.
| When | What usually happens | What it means for your app |
|---|---|---|
| Late April | Apple's deadline for uploads built with the current Xcode and SDK, in recent years | Your next update needs the new toolchain, and so does every library in it |
| Second quarter | Google's major Android release | New behavior to check when you raise the target API level |
| June | WWDC and the first betas of the next iOS | Time to run the app on a beta and list what breaks |
| August 31 | Google Play's yearly target API level deadline | Updates are refused until the app targets the required level |
| September | New iPhones and the public iOS release | Real users on the new version within days |
| Any month | Library releases, security fixes, store paperwork, crash spikes | A steady trickle of small jobs |
Keeping up with Apple and Google
Apple publishes its deadlines on a page called Upcoming requirements, and the pattern is plain once you read back through it. Since April 29, 2024, uploads had to be built with Xcode 15. Since April 28, 2026, App Store Connect has accepted only builds made with Xcode 26 or later, using the 26 SDKs. Nothing happens to the version already in the store on that date. The trouble comes later, when you need to ship a fix and discover that the project no longer builds with the current Xcode because one dependency hasn't caught up.
Google works the same way through the target API level. Since August 31, 2026, new apps and app updates on Google Play must target Android 16 (API level 36) or higher, with extensions available on request until November 1. Existing apps that don't target at least Android 15 are no longer offered to new users on devices running newer Android versions. People who already installed the app keep it, but the app quietly stops reaching new people on new phones, and nobody sends you an error message about it.
Raising a target API level is rarely one line in a Gradle file. Each Android version changes behavior for apps that opt in, around permissions, background work, notifications or how content is drawn at the screen edges, and someone has to find what moved and fix it. Google moved its major Android release to the second quarter starting with Android 16 in 2025, and that is the release that carries the behavior changes. On iOS, some changes arrive with the new SDK and some arrive the day users update their phones, whether you shipped anything or not.
The expensive part is often not your own code. A small-company app usually depends on a list of third-party packages for analytics, crash reporting, maps, payments, image loading or sign-in. When one of them is slow to support the new Xcode or the new Android target, the whole upgrade waits, or you replace the package. Cross-platform apps add one more layer: the framework itself has to support the new platform versions before you can. Testing on the new OS versions needs real phones too, and if you keep a few Android devices for that, this guide to a small Android test device lab covers which ones earn their place.
Paperwork nobody budgeted
Every year brings forms and declarations that didn't exist when the app was built. A few from Apple's own list:
- Since May 1, 2024, new apps and updates must give approved reasons for a set of listed APIs, including when a third-party SDK inside the app is the one calling them.
- Since October 16, 2024, apps on the App Store in the European Union need a trader status under the Digital Services Act before an update can be submitted. Since February 17, 2025, apps without it have been removed from EU storefronts until the status is provided and verified.
- Answers to Apple's updated age rating questions were due by January 31, 2026, to avoid interruptions when submitting updates.
Google Play has its own version of this: the Data safety section has to stay accurate whenever the data your app or its SDKs collect changes, which means every new analytics or advertising package is also a form to update. Each of these jobs takes somewhere between an hour and a day. None is hard, none is optional, and none is in the original budget unless the maintenance line left room for it.
Certificates, accounts and renewals
Some costs are fixed and easy to forget because they come around once a year. At the time of writing the Apple Developer Program costs 99 USD per membership year, and Google Play charges a one-time registration fee of 25 USD. If the Apple membership lapses, the apps on it are removed from sale. I put the renewal date in the client's calendar and in mine, and I ask whether the payment card on the account expires before then.
Then the things that expire quietly. Apps that still use certificate-based push notifications need a new push certificate every year; token-based authentication keys for Apple's push service don't expire, so I move every app I take over to a key. Signing certificates and provisioning profiles have expiry dates too, and the build pipeline tends to discover them the morning after. The domain that hosts the backend and the privacy policy page renews once a year. None of these is expensive. All of them can stop an app, and the fix is the same: one list with every date on it, kept by whoever is responsible.
Fixes, monitoring and the odd bad day
The rest of the year is small work. A crash that only happens on one phone model. A support email from a user whose bookings vanished after changing the phone's language. A maps or payment provider that retires an API version and gives you a few months to move. Security updates for the server and its packages. A few minutes most weeks reading the crash dashboard, so problems are caught by you and not by a one-star review.
And now and then there's a day that starts with "the app is broken". Usually it isn't the app. A certificate expired, a backend service changed its response, or a store review flagged a policy issue in an update. Those days are why maintenance budgets need slack, not just a list of known tasks.
Illustrative hours for three example apps
Here is how I would sketch a first maintenance budget for three kinds of app. These are illustrations, not market figures.
| Example app | What it includes | Illustrative hours per year |
|---|---|---|
| A content app | Reads articles or menus from an existing system, one cross-platform codebase, few SDKs | 30-70 |
| A booking app | Accounts, calendar, reminders, a web admin, iOS and Android | 80-200 |
| A shop or community app | Payments, chat, several SDKs, two native apps | 200-400 |
What moves a number up or down is mostly countable. Two native apps mean two sets of platform upgrades, two sets of SDKs and two store listings to keep current. Each third-party SDK is one more thing that can lag behind a new Xcode. An app with automated tests can be upgraded and checked in an afternoon; one without them needs a manual pass on every screen. Old code, written for an iOS version nobody supports any more, costs more every year it isn't modernized. I don't use a percentage of the build price to set a maintenance budget, because two apps that cost the same to build can age very differently. If you want to see how features add up on the build side, the app development cost calculator shows the hours per feature.
The bills that aren't hours
Running costs never show up in an hours table, but a client pays them every month: backend hosting, the database, an email or text message service, push delivery if it runs through a third party, crash reporting, a hosted sign-in service, maps usage. For a small app they are modest next to the build. They also never stop, and most of them grow with the number of users, which means they rise exactly when the app is doing well.
Ask your developer for a list with three columns: the service, whose account it sits in, and whose card pays for it. If the answers are "the developer's" and "the developer's", fix that before anything else, because an app whose services sit in a freelancer's private accounts belongs, in practice, to the freelancer.
Budgeting it in five steps
- List every third-party service and SDK the app uses, with the account owner and any renewal date.
- Put the fixed dates in a shared calendar: the Apple membership renewal, certificate expiries, Google's August deadline and Apple's spring SDK date once it's announced.
- Book a check in June: install the new iOS beta on a test phone, run the app and write down what breaks.
- Agree a yearly block of maintenance hours with your developer, with a written line between upkeep and new features.
- Hold back part of that block for requirements nobody has announced yet, because there will be at least one.
If your app hasn't had an update in a year, start with one question to whoever built it: does it still build and upload with the current Xcode and the current Android target? The answer tells you how big the first maintenance bill will be.
Keep reading

Planning a small studio's team offsite: the order I'd do it in
For a small studio heading to the Colorado mountains, shared transport decided early saves flights, rental cars and a frantic group chat.

App monetization strategies compared: subscriptions, one-time, ads
Subscriptions, one-time purchases, freemium, ads and hybrids compared, with store commission basics, free trials and price localization for small app teams.

How much does it cost to build an app? A line by line breakdown
A freelance iOS developer walks through an app budget line by line, from discovery and design to backend, testing, launch and the yearly cost of upkeep.