How Superhuman Built an Engine to Find Product/Market Fit

Jan 25 '20 · 10 min read · 477 views

Key Takeaways


Spend 30-60 seconds after absorbing material to really digest it and see how much you can recollect without referring to source. Then repeat this process every day to see what you can remember.
  • Show/Hide
    • As a founder, Rahul felt it difficult to share problems with the team. These super-ambitious engineers had poured their hearts and souls into the product. I had no way of telling the team we weren’t ready, and worse yet, no strategy for getting out of the situation — which is not something they would want to hear. He felt it was irresponsible to just launch and see what happens.
    Product market fit defined by Marc Andreessen:
    “You can always feel when product/market fit is not happening. The customers aren't quite getting value out of the product, word of mouth isn't spreading, usage isn't growing that fast, press reviews are kind of ‘blah,’ the sales cycle takes too long, and lots of deals never close.
    And you can always feel product/market fit when it is happening. The customers are buying the product just as fast as you can make it — or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account. You're hiring sales and customer support staff as fast as you can. Reporters are calling because they've heard about your hot new thing and they want to talk to you about it. You start getting entrepreneur of the year awards from Harvard Business School. Investment bankers are staking out your house.”
    • The above is a lagging indicator - it is useful to define product-market-fit for companies post-launch. Superhuman was:
      • As a successful founder, no problems getting investor confidence to raise money
      • Actively avoiding press
      • Choosing not to onboard more users than they could handle
    • What you need are leading indicators (pre-launch).
    • Sean Ellis (growth hacker) had found a leading indicator: just ask users “how would you feel if you could no longer use the product?” and measure the percent who answer “very disappointed.” If it was over 40% that respond with "very disappointed", then you have traction.
    • We identified users who recently experienced the core of our product, following Ellis’ recommendation to focus on those who used the product at least twice in the last two weeks. (At the time we had between 100 and 200 users to poll, but smaller, earlier-stage startups shouldn’t shy away from this tactic — you start to get directionally correct results around 40 respondents, which is much less than most people think.)
      1. How would you feel if you could no longer use Superhuman? A) Very disappointed B) Somewhat disappointed C) Not disappointed
      2. What type of people do you think would most benefit from Superhuman?
      3. What is the main benefit you receive from Superhuman?
      4. How can we improve Superhuman for you?
    • With the feedback, Superhuman did the following:
      1. Segment to find your supporters and understand your high-expectation customers
        • With your early marketing, you may have attracted all kinds of users — especially if you've had press and your product is free in some way. But many of those people won't be well-qualified; they don't have a real need for your product and its main benefit or use case might not be a great fit. You wouldn't have wanted these folks as users anyway.
        • As an early-stage team, you could just narrow the market with preconceived notions of who you think the product is for, but that won’t teach you anything new. If you instead use the "very disappointed" group of survey respondents as a lens to narrow the market, the data can speak for itself — and you may even uncover different markets where your product resonates very strongly.
        • So assign personas to the people who answer - very disappointed, and ignore all other personas.
        To start, we grouped the survey responses by their answer to the first question (“How would you feel if you could no longer use Superhuman?”):
        We then assigned a persona to each person who filled out a survey.
        Next, we looked at the personas that appeared in the very disappointed group — the 22% that were our biggest supporters — and used those to narrow the market. In this simplified example, you can see we focused on founders, managers, executives and business development — temporarily ignoring all other personas.
        With this more segmented view of our data, the numbers shifted. By segmenting down to the very disappointed group that loved our product most, our product/market fit score jumped by 10%. We weren’t quite at that coveted 40% yet, but we were a lot closer with minimal effort.
        With this in mind, I sought to pinpoint Superhuman’s HXC. We took only users who would be very disappointed without our product and analyzed their responses to the second question in our survey: “What type of people do you think would most benefit from Superhuman?”
        This is a very powerful question, as happy users will almost always describe themselves, not other people, using the words that matter most to them. This lets you know who the product is working for and the language that resonates with them (providing valuable kernels of insight for your marketing copy as well).
        • Once you identify the HXC, focus solely on them!
        • It's better to make something that a small number of people want a large amount, rather than a product that a large number of people want a small amount. In my view, the product/market fit engine process of narrowing the market massively optimizes for a product that a small number of people want a large amount.
      2. Analyze feedback to convert on-the-fence users into fanatics.
        Find out from your "very disappointed" users
        • Why do people love the product?
        Politely disregard those who would not be disappointed without your product. They are so far from loving you that they are essentially a lost cause.
        That leaves the users who would be somewhat disappointed without your product. On the one hand, the ‘somewhat’ indicates an opening. The seed of attraction is there; maybe with some tweaks you can convince them to fall in love with your product. But on the other hand, it’s entirely probable that some of these folks will never be very disappointed without your product no matter what you do.
        Find out from your "somewhat disappointed" users
        • What holds people back from loving the product?
        Use the top benefit derived from "very disappointed" users.
        • Somewhat disappointed users for whom speed was not the main benefit: we opted to politely disregard them, as our main benefit did not resonate. Even if we built everything they wanted, they would be unlikely to fall in love with the product.
        • Somewhat disappointed users for whom speed was the main benefit: we paid very close attention to this group, because our main benefit did resonate. Something — probably something small — held them back.
      3. Build your roadmap by doubling down on what users love and addressing what holds others back.
        • I eventually came to this realization: If you only double down on what users love, your product/market fit score won’t increase. If you only address what holds users back, your competition will likely overtake you. This insight guided our product planning process, effectively writing our roadmap for us.
        • To stack-rank amongst these initiatives, we used a very simple cost-impact analysis: we labelled each potential project as low/medium/high cost, and similarly low/medium/high impact. For the second half of the roadmap, addressing what held people back, the impact was clear from the number of requests any given improvement had. For the first half of the roadmap, doubling down on what people love, we had to intuit the impact. This is where "product instinct" comes in, and that's a function of experience and deeply empathizing with users. (The HXC profile exercise from earlier helps a great deal with developing this muscle.)
      4. Repeat the process and make the product/market fit score the most important metric.
        As time went on, we constantly surveyed new users to track how our product/market fit score was changing. (We were careful to ensure that we didn’t survey users more than once, so as to not throw off the 40% benchmark.)
        The percent of users who answered "very disappointed" quickly became our most important number. It was our most highly visible metric, and we tracked it on a weekly, monthly and quarterly basis. To make this easier to measure over time, we built some custom tooling to constantly survey new users and update our aggregate numbers for each timeframe. We also refocused the product team, creating an OKR where the only key result was the very disappointed percentage so we could ensure that we continually increased our product/market fit.
        And we’re not done — the product/market fit score is something that we’re going to continue to track. I think it's always useful for startups to look at this metric, because as you grow you'll encounter different kinds of users. Early adopters are more forgiving, and will enjoy your product's primary benefit despite its inevitable shortcomings. But as you push beyond this group, users become much more demanding, requiring feature parity with their current products. Your product/market fit score may well drop as a result.
        However, this shouldn’t cause too much anxiety, as there are some ways around it. If your business has strong network effects (think Uber or Airbnb), then the core benefit will keep getting better as you grow. If you're a SaaS company like Superhuman, you simply have to keep on improving the product as the pool of users expands. To do that, we rebuild our roadmap every quarter using this process, ensuring that we’re improving our product/market fit score fast enough.
        For any founder looking to get out of the wilderness and on the path to the ever elusive product/market fit, I’ve been in your shoes — and I hope you’ll consider retooling this engine in those proverbial startup garages to make it your own. And when you finally hit the product/market fit score you’re targeting, my advice is to push the pedal all the way down and grow as fast as you can. It will feel uncomfortable, but you’ll have the evidence you need to know that you’ll succeed.


Try to teach someone the material or write it down (use analogies or acronyms to chunk the information)
  • Launching and seeing what happens is irresponsible. Need to launch when you are certain you have product-market fit. Thus, need to design a strategy around proving fit pre-launch
  • The standard definition of fit is when you have investors at your door, money flowing in faster than you can count it, press raving about you. But these are lagging indicators. Before you reach that stage, you are not certain you're close at all.
  • So, Sean Ellis (growth hacker) delivered an interesting perspective. If your current users are polled (once per user!), what % would say they would be very disappointed if they could not use your product anymore? If that % is over 40%, you have traction and you can go ahead and launch.
  • So, ask the following key questions to your users (2 weeks after they have used the core product)
    1. How would you feel if you could no longer use _________? A) Very disappointed - these are your HXC (high expectation customers) B) Somewhat disappointed - these are your opportunities. C) Not disappointed
    2. What type of people do you think would most benefit from _________? This tells you about them - most people will describe their own persona.
    3. What is the main benefit you receive from _________? Use this combined with A to find out the top features users love. And double down on them.
    4. How can we improve _________for you? Use this combined with B to find out what is holding them back. Be careful to pay closer attention to the answers from personas who enjoy the same benefits with A. There is a likelihood that some B personas will never become A.
  • Assign personas for every survey respondent and (temporarily) ignore all other personas. Marketing brings in a lot of users that may not be ideal, and catering to them is a waste of resources - by the time you deliver the features they ask for, they have already churned out.
  • It's better to make something that a small number of people want a large amount, rather than a product that a large number of people want a small amount.
  • Always revisit your roadmap, and keep the % always in mind.


Get rid of convoluted language, and attempt to ELI5
  • Always look for leading indicators to success, and build your metrics around those. It is hard to build on the basis of - you'll know success when you see it. Ask the right questions, and smartly evaluate answers. Aim to make something a small number of people desperately want more of, than something a large number of people want little of.


Share on