New Product Development Cheatsheet

This is intended to be a useful knowledge base for anyone building new things, particularly new products. Assume you are building a new software product from zero.

By definition, new product ideas don't have “product/market fit”. This is, in my opinion, the canonical definition of product/market fit. The critical things to focus on first are:

  1. What is the shortest path to P/M fit?
  2. What is the opportunity cost for this idea?

And (1) is much more important than (2) because often big new things look like toys at first.

The shortest path to P/M fit is not a straight line because new ideas have to get tested by reality (eg. showing an early prototype to a potential user) and usually new ideas have to go through hundreds of adjustments as they hit reality before they can achieve P/M fit. So the right thing to optimize for, in fact the only thing to optimize for ahead of achieving P/M fit, is to increase the number of learning moments per week. A team that learns 5 things per week is going to move 10X faster than a team that only learns one thing every 2 weeks.


  1. Pre-Launch
    • Things to obsess about
      • First questions to answer
        • Who are you initially building for? Definition should be as close to a “SQL Query” as possible. The definition should help teams orient where to find the people we intend to serve in the world. Also important is “initially”. This will grow/expand over time. But better to be hyper focussed up front.
        • What is the minimum scope that will make those initial users in (1) love the product/service? Better to have 1 user falling over in love vs 10 users expressing mild interest.
        • If (1) and (2) work, how could this new product/service get big? This is where some vision setting comes in and acts as an important “opportunity cost” filter for ideas.
      • The only thing you should be doing is talking to users or building
        • Building customer acquisition counts too, Zapier built SEO landing pages before we even had a public product
      • Increase the number of learnings moments per week
        • Launch early – Whether it's an alpha, beta, etc. Launch early. If you're not embarrassed by what you launch, you've waited too long.
      • Do things that don't scale
        • Put your personal phone number on the landing page
        • Manually onboard every single user, use the software for them
        • Have them pay you manually, reconcile revenue manually
        • Manually find initial users on forums, existing user base, etc.
    • Measures to pay attention to
      • Actual external users
        • UUs (beat zero)
        • Usage per UU (beat zero)
        • Retention per UU (beat zero)
      • Qualitative feedback from users of problem importance
        • Evidence of high motivation / excitement for problem to be solved
        • Look for revealed preferences, do they proactively ask for access every day after hearing about the new product?
    • Time in stage
      • 6-12 months
    • Team size
      • Just the founding team (“co-founders”):
        • Product/Technology/Design/Builder
        • Market/Customer/GTM
      • No more than one team (~7 people)
  2. Post-Launch, Pre-P/M Fit
    • Things to obsess about
      • Get to P/M fit: 10% WoW on usage/revenue is the bar to hit
      • Find a Go To Market acquisition channel that works
        • Most products that get big have a single dominant channel they acquire new users from (examples: SEO, content marketing, channel partners, WoM referrals, PLG referrals, advertising)
      • Keep doing things that don't scale
        • If you have 10 users on day 1, 10% WoW means getting only 1 more user by day 7. Very doable and you can still do super manual things to find/onboard new users.
        • Support your own product. Avoid the temptation to put people or processes between users and builders at this stage. Super critical because rapid iteration is still necessary.
      • Willingness to pay
        • Do not give it away for free at launch. Charge something. Very important as it reveals true problem importance and value.
    • Measures to pay attention to
      • Growth rate
        • Consistent 10% WoW on usage/revenue is the bar to hit for P/M fit
      • Usage
        • UUs
        • Usage per UU (going up as you add features/value)
      • Retention
        • Weekly Cohorts (increasing retention over time w/ new features?)
    • Time in stage
      • 12-24 months (if going beyond 12 months, avoid the “one more feature” trap. Make significant pivots or de-invest.)
    • Team size
      • Up to one team (~7 people)
  3. Post-P/M Fit
    • Things to obsess about
      • Growth. Easy to get distracted at this stage. Keep making sure the #1 priority is to increase growth.
    • Measures to pay attention to
      • Growth rate
        • Hold 10+% WoW as long as possible. Silly thought experiment, if you do this for 36 months in a row your new product is bigger than Google.
      • Monthly metrics
        • MAU/WAU/MEA (some time series measure of engaged user)
        • ARPA (average revenue per account, increase it with tiering)
        • NRR (net revenue retention, goal is to get 100+% monthly NRR)
      • Pricing and packaging
        • With significant users and revenue, can start optimizing price points and models to figure out the elasticity curve.
    • Time in stage
      • Until you run out of ideas and growth stalls.
    • Team size
      • Throttle off. Grow the team as necessary to drive usage/revenue growth