Skip to content
← All articles
6 min read

How to scope an MVP: what to build first and what to cut

Most MVPs miss their deadline not because of the code but because the scope kept growing. This is the order of decisions we use to keep the first version small enough to ship and useful enough to learn from.

An MVP tests a hypothesis, it is not a smaller product

A minimum viable product exists to answer one question: will these people use this thing to solve this problem? If the answer depends on twenty features, the question is too big. Write the hypothesis in one sentence before listing any functions, for example: “Dispatchers will accept orders through a web form instead of phone calls.”

Everything in the first version must either help answer that question or be required to make the answer trustworthy. The rest is a later release, not a missing feature.

One user, one scenario

Pick the single type of user and the single path they take from start to result. A client portal with three roles, a dashboard and notifications is three products. A client who signs in, files a request and sees its status is one.

A useful test: can you demo the whole scenario in five minutes without saying “and later this will also…”? If not, the scope is still too wide.

What is usually safe to cut

Early versions rarely need things that look important in a plan:

  • Complex role models — start with one or two roles.
  • Admin panels for every entity — a script or a database console is enough for the first users.
  • Configurable settings, themes and white-labelling.
  • Reports and analytics beyond what the hypothesis needs.
  • Rarely used integrations that can be replaced with a manual step.

What not to cut

Cutting features is fine; cutting foundations makes the second version a rewrite. Even a small MVP should keep a correct data model, proper authentication and access checks, backups of the database, and a repeatable deployment. These are cheap on day one and expensive to retrofit once real users and real data are in the system.

Agree acceptance criteria before development

For each scenario write down how you will know it works: who does what, what the system shows, what is stored. Short, testable statements are enough. They protect both sides: the customer knows what will be delivered, and the team knows when a stage is finished.

Anything that appears later goes into a separate list for the next iteration instead of silently extending this one.

What you should receive at the end

A working service you can show to real users, the source code in your own repository, deployment instructions and a short description of what was deliberately left out. Ownership of the code and the infrastructure should be clear from the start, so that you can continue with any team. How the work then proceeds and what drives its effort is covered in MVP development process.

Questions

FAQ

Didn't find an answer — ask on the call or in Telegram.

Should an MVP include payments?

Only if the hypothesis is about willingness to pay. Often it is enough to take requests and invoice manually at first. Online payments add provider contracts, legal details and failure handling, so they are worth including only when they are the thing you are testing.

Can the MVP code be reused later?

Yes, if the foundations are sound: the data model, authentication and deployment. That is why we cut features rather than quality. Interfaces and rarely used parts may still be rewritten as you learn from users.

How long does an MVP take?

It depends on the number of scenarios, integrations and roles, so we estimate after the scope and acceptance criteria are agreed. A narrow single-scenario prototype takes much less than a product with payments and several integrations.

Related service

Custom MVP development for your business idea

Start with a working flow you can put in front of users. We help define the essentials, design the service and build a first version around a specific business hypothesis.

Learn about the service
Other articles
Contact

Shall we discuss your task?

Thirty minutes on a call — and you understand the architecture, timeline and engagement format. No obligations.

Or message us on Telegram →Follow our Telegram channel →

Enter your Telegram or email so we can reply.

We reply within 15 minutes.