Keep Thinking
Back to Guides
Product18 min read

Building Your MVP

How to build a minimum viable product that validates your hypothesis.

Keep Thinking Team

1What is an MVP?

An MVP is the smallest thing you can build to test your core hypothesis. It's not a crappy version of your product—it's a focused version.

The goal isn't to launch a product. The goal is to learn whether your assumption about the problem and solution is correct.

Reid Hoffman: 'If you're not embarrassed by the first version of your product, you've launched too late.'

2Defining Your Core Hypothesis

Before building anything, articulate what you're testing: 'We believe [target customer] has [problem] and will [take action] to solve it.'

Your MVP should test the riskiest assumption first. What's the thing that, if wrong, means the whole business fails?

Be specific. 'People will pay for this' is too vague. 'Freelance designers will pay $50/month for client management' is testable.

Pro Tips
  • Write down your assumptions before building
  • Rank assumptions by risk and test the riskiest first
  • Define what 'validated' looks like before you start

3MVP Approaches

Landing Page MVP: A page explaining your product with an email signup. Tests demand.

Wizard of Oz: Looks automated but you do it manually behind the scenes. Tests if people want the outcome.

Concierge MVP: Deliver the service manually to a few customers. Learn exactly what they need before automating.

Single-Feature MVP: Build one feature excellently instead of many features poorly.

4Building Lean

Use no-code tools where possible: Webflow, Bubble, Airtable, Zapier. Speed of learning beats perfection.

Set a time constraint. Most MVPs should take 2-4 weeks, not months.

Launch to a small group first. Get feedback, iterate, then expand.

Pro Tips
  • Every feature you add delays learning—cut ruthlessly
  • If you can fake it with a human, do that first
  • Track the one metric that proves your hypothesis

Key Takeaways

1An MVP tests your riskiest assumption, not just your product
2Define success criteria before building anything
3Speed of learning is everything—launch in weeks, not months
4No-code and manual processes are valid MVP approaches
5One feature done well beats ten features done poorly

Put This Into Practice

Ready to apply what you've learned? Get your startup validated by experts.

Submit Your Startup