This is a common question we hear when meeting entrepreneurs, and on that I came across on Quora this weekend. The question asked “How long is too long to release a minimum viable product?” and specifically referred to how Blekko was in development for over three years before launching. Naturally, I felt the need to chime in and then subsequently turn this into a blog post.

I think we need to defer to Gandalf on this one:

"A wizard is never late. Nor is he early. He arrives precisely when he means to."

And so it should be with a Minimum Viable Product. Every startup is different and only the founders know precisely when is the right time to release their MVP upon the world.

To give a more actionable answer: a MVP by definition should require theminimum (really putting some extra emphasis on minimum here) amount of effort required to gain the maximum amount of validated learning to test the core assumptions.

That’s too many buzz words for me in one sentence, but I think it’s required to properly answer this question. The entire purpose of the MVP is to test the crazy assumptions that the founders have been dreaming up. Most of the people that approach us at NYC Dev Shop never explicitly identify the main assumption that they are trying to validate. That’s a problem.

Without knowing exactly what your core assumption is, how do you know what your MVP should do functionally let alone how long it should take to build before you release it? You are basically saying, “I want to build this solution even though I’m not really sure that I’m solving a problem for anyone but myself (sometimes not even for yourself).”

Once you have your core assumption identified, your decision making process is easy:

  • Will my MVP in its current state allow me to test my core assumption? If no —> keep building
  • Will this new shiny feature provide me with more validated learning? If no —>  defer to Phase 2
  • Will spending countless hours on design and branding truly help me understand what people think about my MVP? If no —> be proud that you are embarrassed by your design

I had someone approach me with the idea of building a new MVP every single week for 10 weeks. I initially thought that the idea of throwing a bunch of stuff on the wall to see what sticks was a clever way to go about finding something amazing. After thinking through it a little more, I decided that I was very much against this. Putting an arbitrary time limit on an idea isn’t the correct approach at all. A week might be too long for a bad idea and certainly might not be long enough for a good one.

At the Dev Shop, our sweet spot is 3 weeks to 3 months of development, but of course it is entirely dependent on the project. Sometimes you can get it done in a weekend at a hackathon. Other times if you’re building a search engine or a Tesla, it could take a bit longer.