Don’t use the phrase “Minimum Viable Product”
I’m a big fan of Eric Ries and the Lean Startup. However I really wish he didn’t popularise the phrase “Minimal Viable Product”. I believe it has led to a lot of misconceptions, and ultimately bad product development approaches. It’s a key step in solution discovery, but it is often used to describe solution delivery. Teams would benefit from using more specific terms for what they are trying to achieve.
The misconception
Many people think a Minimum Viable Product is the first thing that you should release to a customer. Fair enough, that’s exactly what it sounds like. If that’s you, I don’t blame you.
But that’s not what Eric Ries meant by it. To him, an MVP is the smallest thing you can do to maximise learning. That might not be something you release to a customer, and certainly wouldn’t be a fully formed product.
“The lesson of the MVP is that any additional work beyond what was required to start learning is waste, no matter how important it might have seemed at the time.” — Eric Ries, The Lean Startup
But people often think that an MVP is actually the simplest version of the product that you would sell to a customer. It often has all the functionality of the final feature, but often not the scalability. Sadly, I have seen many times when the team releases their “MVP”, and then they get told to focus on building something else. No learning has been used to improve the product, and it has probably created a load of technical debt.
However, the objective of the MVP is to learn. Therefore, to stop the above from happening, it is actually beneficial to build something that can’t be sold! This ensures that we will gather some learning, and use that to build a better product.
How to do it properly
In order to maximise learning, we should try to understand where our biggest risks are. This shows us what is the most important thing to learn. Once we understand what we need to learn, then we can work out what is the simplest thing to build in order to learn that.
“As you consider building your own minimum viable product, let this simple rule suffice: remove any feature, process, or effort that does not contribute directly to the learning you seek.” — Eric Ries, The Lean Startup
For example, we might think that our biggest risk is regarding usability. In which case, we may just build a clickable prototype in a tool like Figma, and test it with 5 users.
Alternatively, we may think that our biggest risk is around our ability to build something in a new technology. In order to learn about this, we may just build a quick “proof of concept”, which we don’t test with any users at all, but we just prove to ourselves that we can do it.
Or we might think that our biggest risk is whether anyone wants our new product/feature idea. In which case, we might run a “fake door test”, where we put a link/button to our new feature, and track how many people click it. We don’t actually build the feature, we just show a well crafted “let-down” message and offer the user the chance to be the first to hear about the feature when it comes.
In all of these scenarios, we aren’t building the product/feature. We are just building the easiest thing that maximises learning. So there is no risk of some senior leader telling us we need to move on to something different. But all of these are great MVPs.
So what should we call it?
The actual aim of the MVP is to test our riskiest assumptions. So I just prefer to call it a “Riskiest Assumption Test”. You could also just call it an experiment, because that’s what it is!
In reality, I often just use the more specific term for the experiment, such as proof of concept / fake door test / usability test.
I think it is also important to work out what we call the first version that we release to customers. It’s important not to use MVP for this, so what should we call it?
Well, I think the following are useful:
- Alpha version: the first version for testing
- Beta version: version for testing the product at scale
- Minimum Marketable Product: the version that you would be willing to start shouting about
Conclusion
I love the Lean Startup, but the “Minimum Viable Product” is not a good name for what was intended. I recommend that you work out what you need to learn, and run experiments to test exactly that. When it comes to building out the first viable product you think you can sell, just call that a Minimum Marketable Product.
