(image source: Henrik Kniberg)

Henry Ford famously said, "If I asked what my customers want, they would have said faster horses." He gave them cars instead. Today, you should probably give them a skateboard.

Spotify's analogy for MVP is building a skateboard when a customer asks for a car. Instead of setting out to build a car by making a wheel, start with a skateboard.

As you build and iterate on MVP, you're learning whether your product has a market. What Marc Andreessen of Andreessen Horowitz said 7 years ago applies equally today: "The #1 company-killer is lack of market." The skateboard stage tests the riskiest hypothesis : that customers want to get from point A to point B.

The skateboard will still take the customer from point A to point B, which means you can learn whether this is actually something they care about. It's the most minimum stage of your product, but also the most important because it's the entire premise of the undertaking. As you build from the skateboard to the car, you're learning at every iteration and getting closer and closer to product-market fit. Your goal is to build less so you can get your product in your customer's hands sooner, and learn faster.

• Nobody can do anything with a wheel, or two wheels-it's not viable. You can't learn anything about whether they want a car or how they'd use a car by watching them play with two wheels.
• Building a car is equal in time to building 1,000 skateboards-it's not minimal. Trying to jump from nothing to a car first falls into the fallacy of the "big idea." And after all that time and money spent on a car, you have no idea if it's what your customer actually wants.