A founder's guide: how to build an MVP that tests one idea

An MVP, or minimum viable product, is the smallest version of a product that lets you learn whether your key assumption is true. Eric Ries defines it as the version of a new product that lets a team collect the maximum amount of validated learning about customers with the least effort.

That makes an MVP a learning tool, not a cheap version of the final product. It fits Steve Blank's definition of a startup as an organization formed to search for a repeatable and scalable business model: the MVP is how you run that search with real customers.

The MVP should test the single assumption that, if wrong, would sink the business. Usually that is whether a specific group of users has the problem badly enough to use, or pay for, your solution.

Write it as one sentence, for example: "Clinic managers will use a tool that does X instead of their current spreadsheet." Then decide what result would count as success, such as repeat use, a signed pilot or a payment, before you build anything.

Keep only the features the test cannot work without, and move everything else to a later version. A simple way to do this is to list every feature you have in mind and ask of each one: does the user need this to experience the core value?

• Must have: the user cannot reach the core value without it.
• Can be manual: someone on the team can do it by hand for the first users.
• Later: useful, but it does not affect what you are testing.

Use technology your team already knows well and that is widely supported, so that time goes into the product rather than infrastructure. Managed services for hosting, authentication, payments and email save weeks of work at this stage.

Avoid building complex infrastructure before you have users. If you are not sure the product needs custom code at all, a no-code or low-code prototype can be enough for a first test, as long as you accept that a later version may need to be rebuilt. Make sure the company owns the code, accounts and data from day one.

No. An MVP is anything that produces real evidence about your key assumption, and sometimes the fastest test involves little or no code.

Common forms include a landing page that describes the offer and measures sign-ups, a manual service where the team does by hand what the product will later automate, a clickable prototype tested with target users, or a pilot with one customer using a simple version. Choose the form that gives the most reliable evidence for the least effort, then build software when the evidence justifies it.

It depends on scope, and scope is the part you control. A narrow MVP that tests one assumption on one platform is far faster to build than a product with several user types, native mobile apps, payments and integrations.

Work in short cycles with a demo at the end of each one, so you can see progress and cut scope early. If the timeline keeps growing, the usual cause is features that were added without a clear link to the test.

Launch to real people from your target audience, not friends and family, and measure what they do, not only what they say. Decide on a few metrics that match the assumption you are testing.

• Activation: do new users reach the core value?
• Retention: do they come back and use it again?
• Willingness to pay: do they pay, sign a pilot or commit in another concrete way?
• Qualitative feedback: what do they try to do that the product does not support?

After the MVP, decide based on the evidence: keep going and add what users need most, change part of the solution and test again, or change direction. All three are valid outcomes of a good MVP.

Write down what you learned and what you will test next. This record is also valuable for investors, because it shows how you make decisions. See /en/guides/startup-pitch-deck for how to present traction.

Leave out anything that does not help answer the question you are testing. Common examples:

• Full admin panels and reporting dashboards.
• Support for many languages or markets at once.
• Social features, gamification and complex permissions.
• Several payment options when one will do.
• Infrastructure designed for scale you do not have yet.

WeCcelerate is a Venture Builder in Israel since 2018, active in Tel Aviv and Jerusalem and recognized by the Israel Innovation Authority. We work with founders on defining what the MVP must test, scoping it and building it. Our MedTech track with Leumit Health Services gives founders clinical expertise, regulatory guidance, data access and pilot opportunities, subject to the required approvals.