Skip to content

Choosing an idea

How do I validate an app idea before building it?

Lumos · Updated

Start with a person and a problem you can investigate. Find out what happens today, what gets in their way, and what they already do about it. Use what you learn to choose a small test of your idea.

Try writing: “I want to help [person] with [problem].” Make the person specific enough that you could actually talk to them. Then separate what you've observed from what you're guessing.

What should I check?

  1. Ask about a recent time the problem happened. What was the person trying to do? What stopped them?
  2. Look at their current workaround. Have they spent time or money dealing with it? What do they like about the way they work now?
  3. Write down what would make you reconsider your idea. Perhaps the problem happens rarely, or your explanation for it is wrong.
  4. Choose a small experiment that helps you learn what remains uncertain. That might be another conversation or a first version someone can try.

Imagine a freelance designer waiting for client feedback. A portal is one possible solution. First, find out whether the difficulty is getting a clear decision, finding scattered comments, or something else. Each answer points towards different work.

Someone describing a problem is useful evidence. You'll still need to learn whether they'd change how they work to use your solution and, if you plan to charge, whether they'd pay.

Y Combinator's guide to getting and testing ideas also starts with the problem and treats the first product as a way to learn.

In the Builders Handbook

Start with finding problems, then work through your assumptions and talking to people.