Minimally Viable Products vs. Experiments

I was recently asked what the difference between a Minimally Viable Product (MVP) and an experiment was. I spent some time reading what others had said about this and found that there is general lack of agreement regarding A) what an MVP is and B) what an experiment is. This makes a comparison of the two difficult.

Despite the confusion, I think the comparison is valuable. Both terms are frequently used when discussing lean startup methods. However, in practice, experiments and MVPs may be overlapping but not equivalent.

What is an MVP?

For the purpose of this comparison, I will use the author of Running Lean, Ash Maurya’s, definition:

A Minimum Viable Product is the smallest thing you can build that delivers customer value

I prefer this definition over others which focus on “maximized learning” because it does not lead to confusion with other valid methods of learning which might not be products (e.g., marketing pages).

What is an experiment?

I found it even harder to find a single definition of an experiment. Even Miriam-Webster’s definitions range from simply a “test” to the following:

an operation or procedure carried out under controlled conditions in order to discover an unknown effect or law, to test or establish a hypothesis, or to illustrate a known law

In business, I would argue that it is valuable to use a more scientific definition of experimentation like this. By scientific, I mean the systematic testing of hypotheses using controlled conditions. Testing hypotheses under controlled conditions allow for causal inferences regarding the effect of variables on each other. For instance, a product team might be able to use A/B testing to determine whether a certain button design causes an increase in signups.

So, what is the difference?

In contrast to an experiment, an MVP is something which can provide value to customers. An experiment might not always do this. An experiment might help a team learn about their customers but the experiment itself might not provide value to customers.

In contrast to an MVP, an experiment is a tactical approach to answering specific questions about the relationship between factors. For instance, an experiment can be used to test small changes to a product. This might inform the creation or modification of an MVP.

In a way, the MVP is a company’s theory of what customers value. Experiments can be used to improve this theory of value to customers.

Read my article “Need a startup idea? Start with a problem.”