An MVP scope names the one workflow the product has to prove, the users it is for, what is deliberately left out, and how you will know whether it worked. Cutting scope is the most reliable way to reduce what an MVP costs and how long it takes, because scope sets both numbers.
This guide shows how to find that one workflow, how to sort every feature into in, later or out, what a scope document contains, and gives you a template you can copy. For what a given scope costs, see our MVP cost ranges.
Start with the decision, not the feature list
Most scope documents start as a list of features. That is backwards. An MVP exists to answer a question, so the scope starts with three sentences:
- The belief. What you think is true about your users and their problem.
- The result that would disprove it. What you would have to see to stop or change direction.
- The smallest build that produces that result. Everything else is a candidate for later.
Get the order right: settle what you need to learn, then scope the build, and only then compare vendors (our MVP developer comparison lists what to ask them). Skip the first step and you end up scoping a product rather than an experiment.
How to find the one core workflow
The core workflow is the shortest path a user takes from arriving to getting the value you are betting on. To find it:
- Write down who the first users are, as narrowly as you honestly can.
- Write the single job they hire the product to do.
- List the steps from sign-up to that job being done, and nothing else.
- Mark the step where value is delivered. Every step before it is required. Steps after it are usually later.
An illustrative example (not a client project). Suppose the belief is that independent physiotherapy clinics will pay for automatic appointment reminders that cut no-shows. The core workflow is: a clinic signs up, connects its calendar, and patients receive a reminder they can confirm or cancel. The measure is the no-show rate before and after. Online booking, a patient app, invoicing and multi-location reporting are all reasonable features, and all of them can wait, because none is needed to learn whether reminders reduce no-shows.
In, later or out: prioritising features
For each proposed feature, ask whether the first release could still prove or disprove the belief without it. If it could, the feature moves to the later column. Here is that sort applied to the clinic reminder example above. Our MVP development services page shows how we run the same sort on every build.
| Ship in release one | Defer, and name what brings it back | Leave out |
|---|---|---|
| Clinic sign-up and a calendar connection | Online booking: when pilot clinics ask patients to book through the product | A patient mobile app: it answers a different question |
| Reminder sent, with confirm and cancel replies | A second calendar provider: when a paying clinic uses one we do not support | Features copied from a competitor without a user asking |
| A no-show count per clinic, before and after | Multi-location reports: when a clinic group signs up | Marketing tools for clinics |
| Patient data handled to the rules that apply to it | Invoicing: when clinics want to pay through the product |
The middle column is the one that protects the date. Each deferred feature is written down with the trigger that would bring it back, so cutting scope is a decision on record rather than an omission, and nobody has to re-argue it every week.
Security is never deferred, which is why the last row of the first column is there. If the product will hold health data, payment data or anything regulated, the controls that data requires are part of the first release, and they belong in the scope from the start.
What a scope document contains
A useful MVP scope document fits on a few pages. It covers:
- Problem and belief. The question the MVP answers.
- First users. Who they are, how many, and how you will reach them.
- Core workflow. The steps, start to finish.
- Must-have features. Only what the core workflow needs.
- Deferred list. Each item with its trigger.
- Out of scope. Stated plainly. The excluded half is the part that matters most.
- Integrations. Every external system, named, with who owns access to it.
- Compliance and data. What data the product holds and which rules apply to it.
- Success metric. The number that decides continue, change direction or stop.
- Acceptance criteria. What “done” means for each must-have feature.
- Decision owner. One named person who answers product questions.
How scope choices move cost and time
Every line in the scope moves the budget and the schedule. The biggest movers are integrations, regulated data, the number of user roles, a second platform, and how settled the data model is. An integration discovered after the build starts is a common reason a budget grows, which is why the integrations line is filled in before anyone estimates.
We do not quote prices in this guide, because the price follows the scope. For ranges by type of MVP, see the MVP cost ranges, and for the phase-by-phase schedule, see how long an MVP takes.
MVP scope template
Copy this table into your own document and fill in the right-hand column.
| Section | What to write |
|---|---|
| Belief | We believe that [users] will [behaviour] because [reason]. |
| Disproving result | We will stop or change direction if [measure] is below [threshold] after [period]. |
| First users | [Who], about [how many], reached through [channel]. |
| Core workflow | Step 1 to step N, ending when [value is delivered]. |
| Must-have features | Only features the core workflow needs. |
| Deferred | [Feature]: brought back when [trigger]. One line each. |
| Out of scope | [Feature or area], stated plainly. |
| Integrations | [System]: purpose, data exchanged, who owns access. |
| Compliance and data | Data held, rules that apply (for example HIPAA or PCI DSS), controls in the first release. |
| Success metric | The one number, how it is measured, and when it is read. |
| Acceptance criteria | For each must-have feature: done means [testable statement]. |
| Decision owner | [Name], who answers product questions within [time]. |
The template, filled in
The same clinic reminder example, completed. The numbers are illustrative, not from a client project.
| Section | Filled in |
|---|---|
| Belief | We believe independent physiotherapy clinics will pay for automatic appointment reminders because missed appointments leave paid hours empty. |
| Disproving result | We will change direction if pilot clinics see no fall in their no-show rate after eight weeks. |
| First users | Ten independent clinics, reached through their practice-management software community. |
| Core workflow | Clinic signs up, connects its calendar, patients get a reminder, patients confirm or cancel. Value is delivered at the reply. |
| Must-have features | Sign-up, calendar connection, reminder sending, confirm and cancel replies, a no-show count. |
| Deferred | Online booking: when clinics ask for it. Invoicing: when clinics want to pay in the product. |
| Out of scope | A patient app and marketing tools. |
| Integrations | The clinic calendar: appointment times only; the clinic owns access. |
| Compliance and data | Patient names, phone numbers and appointment times; the health privacy rules that apply are confirmed before the build. |
| Success metric | No-show rate per clinic, read weekly, decided at week eight. |
| Acceptance criteria | A reminder reaches the patient before the appointment, and a reply updates the calendar. |
| Decision owner | The founder, who answers product questions within one working day. |
Before you share the scope with a vendor
A good vendor will push back on the scope, especially the out-of-scope list. Ask them what they would cut, which integration worries them most, and who owns the code and the deployment credentials. Our page on what to ask any MVP developer lists the questions that matter, and MVP, prototype and proof of concept explains which one your scope actually describes.
Frequently asked questions
What is an MVP scope document?
A short document that names the question an MVP answers, its first users, the one core workflow, the features it needs, what is deferred and what is out, the integrations and data rules involved, and the metric that decides what happens next.
How do I decide which features go into an MVP?
Ask whether the first release could still prove or disprove the belief without that feature. If it could, the feature waits. Write each deferred feature down with the trigger that would bring it back.
How can I reduce MVP development costs?
Narrow the scope. Fewer features, one platform first, as few integrations as possible, and integrations named before the estimate. Scope is the biggest cost lever because it sets both the price and the schedule. Our MVP cost guide shows how each choice moves the price.
How much should I spend on an MVP?
Enough to answer the question the MVP exists to test, and no more. Write the disproving result first, then scope the smallest build that can produce it. Our MVP cost ranges show where different kinds of MVP typically land.
Should security be in an MVP?
Yes. Enough security and data handling to be legitimate is part of the first release. If the product holds regulated data, the controls that data requires are in scope from day one.
Scope your MVP with us
Bring the idea and the belief behind it to a free scoping call. We will separate the core from the nice-to-haves, name the integrations and compliance work that move the budget and the date, and write down what waits. See MVP scoping and development.
Sthenos is an SBA-certified Women-Owned Small Business and Economically Disadvantaged Women-Owned Small Business (WOSB and EDWOSB, active through March 2029), registered on SAM.gov (UEI ULUZGKWYCJB7, CAGE 9RX92), under primary NAICS 541511, Custom Computer Programming Services.
