How to define your first MVP

One core journey, a clear hypothesis and the features needed to test it.

A modular prototype with a core blue block and transparent future modules

Start with one user task. Keep the features needed to complete the journey and define the evidence that will help test your hypothesis.

01. Define the hypothesis

Write down who has the problem, how they solve it today and what you want to change. Choose a specific audience and testable situation rather than a platform for everyone.

Replace “build a booking service” with a testable assumption: customers of a small studio will select an available slot themselves if they can see the service, price and booking confirmation. This is an illustrative example, not a research finding. It identifies the audience, action and reason. Then decide what to observe: whether people find a slot, complete a booking and need help at particular steps.

02. Describe the journey to an outcome

For a booking service, the journey might be choosing a time, confirming and receiving a notification. Account features or recommendations can wait if the journey works without them.

In this example the core path is: choose a service → see available slots → leave contact details → receive confirmation. An administrator also needs to maintain the schedule, change statuses and contact customers. The MVP must support both sides. A polished customer screen cannot test the idea if the team cannot fulfil bookings or availability is out of date. Plan the operational journey alongside the public interface.

03. Prioritise the features

Ask of each feature: can we test the core assumption without it? Don’t postpone data security or the correctness of critical operations simply to reduce scope.

Ask of each feature: can users finish the core journey and can we test the assumption without it? If yes, consider it for a later release. Loyalty rewards and recommendations may wait; checking whether a slot is already taken is fundamental. A temporary manual step can be reasonable when its workload is understood and it does not create hidden risks or a misleading experience for customers.

  • Essential: enables the core journey.
  • Next: improves usability after validation.
  • Later: does not affect the current hypothesis.

04. Agree on how to evaluate the launch

Observe whether users complete the main action, where they stop and what they say. Compare the evidence with your hypothesis rather than registrations alone. Use the results to define the next version.

Write the validation plan before development: whom to invite, what task to give them and how observations will be collected. Do not substitute sign-ups for completed bookings if bookings are the hypothesis. Compare behaviour with interviews: a participant may say the flow is clear but stop at confirmation. In a small pilot, recurring obstacles and reasons for abandoning matter; conclusions about an entire market need broader evidence.

05. Limit the pilot, not the quality of the core journey

A first version can serve one city, team or service type. This reduces complexity without removing essential checks. Authentication, correct data storage, clear errors and recovery still matter when the product depends on them. MVP means a limited agreed scope. Before launch, name the people handling enquiries, fixing issues and deciding when a pilot should pause. Smaller scope should leave the core service usable and supportable.

06. Decide what happens after validation

After the pilot, choose an action: develop the journey further, change the assumption or set the idea aside. Keep observations, technical issues and feature requests separate. A request should not automatically become a feature; first establish whether it supports the core task. The next release needs a short priority list, acceptance criteria and a record of the evidence behind each decision. That makes learning visible rather than merely extending the backlog.

Example priorities for a booking MVP

FeatureFirst releaseReason
Service and slot selectionYesThe core customer action.
Availability validationYesPrevents double bookings.
Confirmation and admin workflowYesThe team can fulfil the booking.
Rewards and personal recommendationsLaterNot essential to this initial hypothesis.

Sources and further reading