It doesn't matter how you define MVP
I got 22 definitions of MVP and 'bitch' ain't one.
One of the questions I get asked the most is: “Jeroen, what is an MVP according to you?”. In this post, I will explain why this question is the wrong question.
I needed to define MVP for next week’s piece on the advice ‘Build an MVP ASAP’. This got a bit out of hand, and I figured a standalone piece on defining MVP was worth creating.
👨🔬 Defining using science
One of my favourite aspects of science is defining things properly. Oscar Wilde famously said ‘To define is to limit’. Clear boundaries allow for clear discussion.
MVP, coined by Frank Robinson in 2001, stands for minimum viable product. The term is popularised by Eric Ries and Steve Blank, perhaps you read either of their books.
What an MVP is depends on who you ask. A 2016 literature review identified over a hundred articles with a clear definition, containing about a dozen unique definitions. They rank up in five categories: minimum features, minimum requirements, smallest possible implementation, minimum effort, and minimum value organisation.
🎯 So, which one is it?
So which one defines MVP? All of them and none of them. With ideas such as the MVP, anyone with a blog can add to the definition of it 👀 🌝. If a blog gets picked up, a definition gets hold of an audience.
More often than not, these social constructs live a life of their own. Ironically, the more people try to define such an idea, the more it creates a distance between what the idea describes and practice, rather than reducing it. I’ve seen this for definitions of design, entrepreneurship and opportunities.
Where in natural science, our definition of water has been refined over the past thousands of years. The ancient scholars thought it was one of the core elements. We now can say that pure water always contains 2 hydrogen molecules and 1 oxygen molecule.
But we can’t reduce an MVP to atoms like we can reduce water to its atoms. You can’t define MVP by its material parts, like it always has a login button or it always has an ‘on-button’. Because MVP is an idea applied to something people make, it’s hard to define it as concrete as something like water.
🥅 Your purpose is leading, not the MVP definition
So rather than being more concrete, you see that the MVPs definitions become pluriform, depending on the use of the idea. Depending on where you are in your startup journey, you require your MVP to achieve certain things.
That’s why I don’t think ‘What is the right definition of MVP?’ is an interesting question for an entrepreneur’s journey. Nor is it a question that guides you in your journey.
A more interesting question is: ‘What is your MVP’s intended use?’ Do you want to see if you can sell it, do you want to gather feedback?
An even more interesting route is to go one step back, before deciding on an MVP, asking yourself: ‘What is it that I need to learn?’ Sometimes, an MVP is a suitable answer. Very often, something else.
📝 My definition of MVP
My above writing suggests that I still have an implicit definition of MVP. It’s not just any research device or any prototype. I define MVP as such:
The minimal thing that actually delivers sufficient value to your customer, deployable in the context of the customer, that you can sell.
For me, MVP is all about proof of value and sellability.
Proof of value: Does it generate value in the context of the customer? The fact that a prototype works in your test environment doesn’t equal value.
Sellability: Can you sell it, i.e. is a customer willing to pay for this value? Is it enough value?
So a smokescreen landing page is not an MVP. Even though you can test sellability, you can’t test value delivery.
What about prototypes? Prototypes are a much broader category, that for me includes any materialisation of an idea aimed at learning something. A prototype deployed in your customer’s usage context might prove it generates value. But if you didn’t sell it, you didn’t prove that you can sell it.
So, is an MVP a prototype that works in the customer’s context that you sold? Yes, for me it is.
MVP examples that fit this description
Breeze’s (above) dating app Whatsapp based MVP that got people on dates and the hosting bars paid for it - Full story here
Flowline records padel matches. They made a self-serve prototype using a Raspberry Pi and a GoPro that they hung up at a padel court. People actually used it and paid for it - More here
AirBnB famously put up the airbeds in their apartments. People paid for the stays - More here
🔥 MVP as proof of value generation and capture
The MVP lies at the core of your startup. In the end, your startup should generate value, that you can capture for profit.
MVPs don’t necessarily scale well. Breeze’s WhatsApp-PDF hacky MVP couldn’t handle more than 200 users. ‘Do things that don’t scale’ still applies.
Your MVP should be able to demonstrate the value delivery and actually selling the MVP so you can capture profits.