How to Plan an MVP

Aug 3, 2019 00:00 · 496 words · 3 minute read Startup School 2019

These are my notes from “How to Plan an MVP”, by Michael Seibel. An entry of my Startup School 2019 series where I follow the curriculum in Y-Combinator’s Startup School

Michael Seibel co-founded (Later became twitch)in 2007, and is now a Partner and CEO of the YC accelerator program.

Goal of Pre-Launch Startup

  • Launch something quickly - You don’t need the version of your product that everyone can use. You just need a version that anyone can use. It WILL be bad.
  • Get initial customers - You’d be surprised how many companies fail before getting even seeing a single person interacting with their product.
  • Talk to customers and get feedback - See How to Talk to Users

Hold the problem you’re solving tightly, hold the customer tightly, hold the solution you’r building loosely.

Founder’s often try to figure out what else they can use their solution for when they fail.

I’ve built a screwdriver but it’s not good at screwing anything in. I wonder what other problems it might solve? Maybe it can be a cooking utensil?

No, don’t do this. Keep the mechanic (user), keep the problem, fix the screwdriver! Keep working on it until it solves the problem. The problem is the important part.

Lean MVP (in most cases)

  • Very fast to build (weeks not months)
  • Very limited functionality
  • Appeal to small subset of customers

Founders want to address all their users problems, and all their potential users. Instead focus on small set of initial users, and highest order problems. Ignore everything else.

This is just a starting point. It’s not special.



  • No payments - had to pay in host in person.
  • No map view - copy paste into google maps.

Twitch (

  • One channel - watch justin’s life or leave
  • low resolution video
  • no video games.

Stripe (/dev/payments)

  • No bank deals - processed payments in a very “startupy way” instead
  • No integration API - They came to your office to integrate it for you because they were desperate for customers, plus wanted to find bugs.

All of these are billion dollar companies, all of these we can agree started with something pretty bad. Don’t fall in love with your MVP.

None of these examples are the initial vision of what the product ended up being.

You wouldn’t fall in love with a paper you wrote in the 1st grade. And that’s often the impact your MVP will have in the end.

It’s just a starting point.


Founders have a misconception of what it means to launch. They see big companies launch with a lot of press, and a lot of buzz, and think that’s what it means to launch.

Do you remember when Google launched? Facebook? Twitter?

Launching is not that special. Let’s put the “press” launch off, and do the “get any users at all” launch as soon as possible.

It’s easier to learn from your customers with an MVP in front of them.


tweet Share