Minimum Viable Products emerge from Experiments

Bob Jansen
January 15, 2016

Robbert wrote about most Minimum Viable Products (MVP) likely being an experiment. Lean Startup has been around for a while now, most of it is super helpful for startups. Not the term MVP. Lean Startup is meant to limit risks for your entrepreneurial endeavours. However when the MVP comes up, team focus easily shifts from learning to building.

Over time, we’ve seen people take different approaches to Lean Startup. The most common is the half-assed version where teams prepare all experiments up-front, build all the product they need to execute all tests. They execute their full plan of experiments without adapting to learnings. It’s interesting how people manage to turn an iterative process as Lean Startup into something with a waterfall twist.

Welcome to the world of the solution driven startup. You’re now in the danger zone. It’s easy to spot. When your spending 80–90% of your time talking about the next product feature; you’re there.


Balancing between learning and execution

Effectively balancing a team between execution and learning mode is hard. To us, this all comes down to building something people want. It’s just such a waste of effort to work on something that nobody ends up using. Sometimes making things for the sake of making things is fun. But the majority of the time should be spent on things that improve the life of users.

Much of my involvement in the Lean Startup scene evolves around Experiment Design. Which I think is great, because it is easier to get right than just talking MVPs. On the downside, people turn their startup into an overly scientific effort and try to validate *everything *up-front. I call this dry-running Lean Startup.

You can’t validate anything without shipping product

Startups are striving to achieve ‘product-market fit’. Originally coined by Marc Andreessen, it’s the sweet spot where there is so much demand for your product; you can’t keep up. You can’t hire fast enough; you can’t build fast enough. For every customer request you answer, three more pop-up.

Before reaching product-market fit, your job as a founder is to proof that you created a product loved by people (whether it solves a problem or provides a value). Testing problem-solution fit is done by measuring customer experience and happiness after using a version of your product. The closer you get to the customer experience you aim to provide, the better.

Mapping the MVP to the Business Model Canvas

“Where are your customers coming from?” this is one of the biggest questions startups fail to answer correctly. However, it’s the most important one to answer. As it defines vital parts of your product and marketing efforts.


MVP on the Business Model Canvas.

I use the above mapping to the Business Model Canvas to explain MVP to startup teams. Assumptions around these items of the business model help to focus on what matters. You can create the most awesome product, but without a channel you can attract people from, what good is it?

The focus on *customer relationship *and *channel *is vital to proof that there is enough demand to for your product. If you’re wondering whether revenue streams should be part of your MVP, it depends on the type of business your building. If your startup’s business model is typical SaaS model, yes you would ideally test for revenue as fast as possible. Building a market place? Focus on growing both sides of your model and their interaction first.

What to test first?

In my projects, I prefer to start with exploring the customer segments that are in my market. First with some desk research and then by talking to customers. I’m painting my ‘day in the life’ of a customer there. With this in mind, I start running some channel & proposition experiments.

On a high-level it works more or less like the beautifully hand-drawn arrows below. First I learn about the customer and focus on qualitative learnings about the value I try to provide.


These are the best arrows I could do using my trackpad

Skateboard, then Bike, then a Car

Maybe some of you have seen this image regarding building your MVP (aka the first version of your product). The only takeaway for most people is to build a simpler version of the product. What tends to be forgotten is the MVP you ship should still take the user from A to B (to stay close the example of skateboard to a car).


The first version of your product provides your users or customers with their desired outcome. The goal of doing this is to measure both their interest and happiness after using your product or service.

With limited resources you need to be smart where you spend your time. Don’tspend your time *just *talking to customers. When spending too much time doingthis, it can become an obsession and you might forget to work on your solution.Switch between the two.

Build the mindset of experimentation. After deciding what you need to learn,think of ways in collecting data to validate or invalidate assumptions. Design afew experiments and run them. As you progress, the need to start delivering onyour proposition becomes a need. You’ll notice your experiments are driving thedevelopment of your MVP. Your need to learn forces you to focus on the thingsthat matter.