• The best founders maintain a direct connection to their users throughout the lifespan of their company: no middlemen
  • The Mom Test tells us three common errors that we make while conducting user interviews:
    • Talking about your idea, and not their life.  The goal is to extract information that helps you improve their life, not to talk about ideas.
    • Talking about hypotheticals, instead of specifics.  Never ask, "If I build this, will you like it?"  Instead, talk about specific things that have already happened in their life.  Talk about why they're experiencing pain, and what that pain is.
    • Talking, instead of listening.  You want to bring real facts and hard data about real-world problems to the table.  Not just ideas.  Actionable, practical, solvable pains to be alleviated.
  • Five great questions to ask in any (unbiased!) user interview:
    • What's the hardest part about [doing this thing]?  In the context of dropbox: what's the hardest thing about working together on these lab projects?  The best startups are solving problems that people are facing on a regular basis, or that are painful enough to warrant solving.
    • Tell me about the last time that you encountered this problem. The goal of this question is to extract context surrounding the problem.  It's not to identify the exact problem, but you gain the ability to explain the value of your solution with context that you may not be aware of.  
    • Why was this hard? In general, customers don't buy the "what". They buy the "why". "This will help with this exact problem I had the other day!"  Explaining in terms of technology, "here's the file syncing tool you've always dreamed of", won't persuade.  But framing it as a solution to the actual pain they personally experience in a real situation that you can articulate - that's persuasive.
    • What, if anything, have you done to try to solve the problem? If people aren't already looking for solutions to your problem, it's possible that the problem isn't a burning enough problem to justify trying to solve.  You also want to understand what your solution will be compared against and assess whether the delta is large enough.
    • What don't you love about the existing solutions?  This is how you can begin designing your product.
  • Talking to users is incredibly important at all phases of a company, but there are three critical points:
    • The idea stage
    • Prototype stage
    • Launch
  • Talking to people in-person is a great way to handle this because it's hyper-genuine
  • Take detailed notes while you're interviewing.  Capture as much info as possible!
  • Keep it casual!  And be respectful of their time.  10 minutes should be enough.
  • It's important to identify a high-quality first customer demographic.  You don't want to be led astray by the wrong customers.
  • Find numerical answers to:
    • How much does this problem cost?  Time or money.
    • How frequently does it happen?  More frequent = better problem to solve
    • How large is their budget?


  • For every potential type of user, rank these.  Could help illuminate the path.
  • You want to be iterating towards product-market fit.
    • When you've made something people want
    • When you no longer have to push the product to people, but when it's being pulled out of you.
  • But PMF is a lagging indicator.  It doesn't help you predict how to achieve it.
  • If you're looking for a quantitative way to guide your idea towards PMF, read the blog post that Superhuman wrote on the topic.


  • if you can get >40% of people to day "very disappointed", you've probably got PMF.
  • Don't design by committee.  You can't ask your users what features they want; if you do, ask them if they'd pay for the change as a litmus test
  • Specific, tactical data is far more valuable than fluff or praise.