Skip to main content
New: Forge AI docs + Loop PM assistant. 7-day free trial.
Back to Glossary
ResearchM

Minimum Viable Experiment (MVE)

What is a Minimum Viable Experiment?

A minimum viable experiment (MVE) is the simplest test that can answer a specific question about your product, user, or market. The goal is to learn whether an assumption is true or false with the least possible investment of time and resources.

MVEs are not about shipping products. They are about answering questions. "Will users pay for this?" can be answered with a fake door test, a pre-order page, or a concierge service long before you write any code.

Why MVEs Matter

Building a full feature to learn that nobody wants it is the most expensive way to get a "no." An MVE gets you the same answer in days or weeks instead of months, at a fraction of the cost.

The lean startup methodology is built on this principle: validated learning is the measure of progress, and the fastest path to learning is the minimum viable experiment.

How to Design an MVE

Start with your riskiest assumption. What belief, if proven wrong, would kill this initiative? That is what you need to test first.

Choose the simplest test method. Match the assumption type to the right experiment:

Define success criteria before running the test. "If 5% of visitors click the CTA, we proceed. If less than 2%, we pivot." Setting the bar upfront prevents post-hoc rationalization.

Run the experiment and make a decision. Ship or kill based on the result. Do not run the same experiment with different parameters hoping for a better answer.

MVEs in Practice

Dropbox's famous explainer video was an MVE. Before building the product, they tested demand with a video showing what the product would do. The video generated 70,000 waitlist signups overnight, validating demand without writing a line of product code.

Buffer's MVE was a two-page website. Page one described the product concept with a "Plans and Pricing" button. Page two showed pricing tiers with a "Sign up" button. When users clicked sign up, they saw a "we're not quite ready yet" page with an email capture. This validated both demand and willingness to pay.

Common Pitfalls

  • Experiment too big. If your MVE takes more than 2 weeks, it is not minimum. Simplify.
  • No success criteria. Without pre-defined thresholds, you will interpret ambiguous results as positive.
  • Testing the wrong assumption. Make sure you are testing the riskiest assumption, not the easiest one.
  • Not acting on results. Running experiments and ignoring negative results defeats the purpose. If the data says no, pivot.

MVEs are a core practice in lean startup methodology. They are related to but different from MVPs. Specific MVE techniques include fake door tests, Wizard of Oz tests, and prototype testing. Experiment design provides the framework for structuring MVEs.

Frequently Asked Questions

What is the difference between an MVE and an MVP?+
An MVP is a minimal product you ship to users. An MVE is a test that may not involve any product at all. You can run an MVE with a landing page, a manual process, a survey, or a prototype. MVEs validate assumptions; MVPs validate products.
How do you know if your experiment is 'minimum' enough?+
Ask: what is the cheapest way to get a reliable signal? If you can validate demand with a landing page, do not build a feature. If you can test usability with a Figma prototype, do not write code. Always start with the lowest-cost option.
Free PDF

Get the PM Toolkit Cheat Sheet

All key PM concepts, tools, and frameworks in a printable 2-page PDF. The reference card for terms like this one.

or use email

Instant PDF download. One email per week after that.

Want full SaaS idea playbooks with market research?

Explore Ideas Pro →

Explore More PM Terms

Browse our complete glossary of 100+ product management terms.