Michael Seibel - How to Plan an MVP

Hold the problem you're solving tightly, hold the customer tightly, but hold the solutions you're building loosely.

• Launch quickly, even if it is imperfect. 
• Get some initial customers. Get anyone using your product and see if they can get value out of the product.. You don't have to think about how you get everyone using it. 
• Talk to your users after you've launched your MVP to get feedback.
• Iterate. Continue to improve on your solution until it actually solves a problem in most cases.

• Build fast. Do it in a few weeks, not months. For a heavier MVP, such as building a rocket or biotech, your MVP can start with a simple website explaining what you do. This can help when you talk to people so that they have something to refer back to.
• Limited functionality. You need to condense down what your user needs into a very simple set of things. Founders may want to address all of their users problems and for all their users. But it is better to focus on a small set of initial users and their biggest problems. The rest should be ignored until later.

Many founders see how big companies launch and then assume they have to launch in the same way. However, a launch simply means to start getting customers.

The only way you can know if your product solves their problem is to put the thing in front of your customers. If it doesn't solve the problem, you know right away.

• Timebox your spec. Your spec is the list of things you need to build before you launch. If you want to launch in three weeks, then focus on the stuff you can build in three weeks.
• Write your spec. If you don't write things down, you may change them and not even realise it. If you write it down, you can be honest with yourself that you're changing it.
• Cut your spec. When you're a week into your three-week sprint, you may realise that you have too many things in your spec. Cut the things that are not that important.
• Don't fall in love with your MVP. It will probably change as you adapt.