How long does it take to build an app? Time by scope and route

How do you work out how long an app will take?

Calendar time is total hours divided by the hours you can give it each week, plus the waits that no amount of effort shortens, such as store review. Most “it takes 3 to 6 months” answers skip the middle of that sum. Someone building for 30 minutes an evening and someone building full time are not on the same timeline, even with the same app.

The hours in this page are assumptions for planning, chosen to show the arithmetic. They are not measured averages, and your numbers will differ. Treat them as a starting estimate to replace with your own after the first week of building. The fixed waits, such as store review, come from Apple and Google and are cited where they appear.

1. Pick the scope tier that fits your idea.
2. Pick your route and read off the assumed hours.
3. Divide by the hours you can spare each week.
4. Add the fixed waits for the stores you will use.
5. Add a buffer. Pick your own figure; 50% on the total is a reasonable place to start for a first project.

How many hours does an app take, by scope and route?

The hired-developer column matches the assumptions used on the cost page, so you can read time and money together. The self-build column is lower because a first version for a single user needs no design polish, but it assumes you are new to the tools and will spend time learning them.

Two things stretch the hours more than anything else. The first is scope that grows while you build, which is why scoping an MVP comes before a tool is opened. The second is debugging: an AI agent or builder can produce a first screen in an afternoon, then spend days on the one bug that stops it working on a real phone.

How many weeks is that on the calendar?

Find your total hours in the left column and the hours you can give each week across the top. The figures are weeks, from simple division.

A worked example: a simple app built by you with an AI tool, assumed at 60 hours, at 10 hours each week, takes 6 weeks. With the 50% buffer it is 9 weeks. If you have only evenings and manage 5 hours each week, the same app takes 12 weeks, or 18 with the buffer. That is the reason people who “built an app in a weekend” and people who “are still building” can both be telling the truth.

Where does the time actually go?

Not mostly into coding. Here is an assumed split for the 60-hour simple app above, to show the shape. Your split will differ, but notice how much of it comes before and after the build.

Skipping validation saves 8 hours and risks all 60. The detailed workflow for the build phase is in the AI coding agent guide, and the full path is in how to make an app.

How long do the app stores take?

A web app has no store, so it has no wait. For native apps, add these fixed delays.

Apple’s guidelines, which reviewers apply, say an app should be a final version without placeholder text and tested on a device. A rejection for incomplete work or missing details costs you a round trip, so budget for one. The App Store publishing guide lists what to prepare, and the Android guide covers the closed test. The Apple guidelines are worth reading before you submit.

Plan the Google 14 days from the day you have a build ready for testers, and line up the 12 people early. Friends and relatives who agree in advance are a start; ask them while you build.

How do you plan a timeline for your own app?

Copy this, fill in your own figures, and check it against reality after the first week. If you are behind, cut scope before adding hours.

App: [name]
Scope tier: [simple / medium / large]
Route: [AI agent / no-code / hired]

Assumed total hours: [number]
Hours I can work each week: [number]
Weeks of work: [hours / weekly hours] = [weeks]
Buffer (start at 50%): [weeks]

Fixed waits:
- Apple review: [2 days, if iPhone]
- Google closed test: [14 days, if Android and a new personal account]
- Time to find 12 testers: [days]

Total calendar time: [weeks]
Date I will check progress against this estimate: [date after week 1]
What I will cut if I am behind: [feature 1, feature 2]

Keep the estimate honest by tracking actual hours for the first week. If a phase ran at double your assumption, double the remaining phases of the same kind.

How do you finish sooner?

• Start from a build prompt, not a blank page. Every idea on the board comes with a generated one, and the free AI Coding Prompt Builder writes one from your answers.
• Choose the web first. It skips the store waits entirely.
• Build one thing. Time grows faster than features, because every feature interacts with the others.
• Fix the order of work. Build the screen where the user does the job first. Sign-in and settings come last.
• Recruit testers early. If you are going to Android, ask the 12 testers while you build, not after.
• Say no to polish. A first version that works beats one that looks perfect and is not finished.

If you run out of time before you run out of enthusiasm, you can still get the idea built: post it on the board, free and with no account, so a builder with spare weeks can take it.