Hiten Shah

creator of KissMetrics

Lean Startup poster-boy

Jul25'2011 talk

  • lots of failures
  • MVP: focus on testing assumption
  • daily.io
    • daily site stats email
    • put up Mini Sales Site
      • no dev
      • just a pitch (Splash Page), no screenshot or other specifics
    • put up survey
    • 50% said they only use free tools
      • killed this idea
  • KissMetrics
    • took 3iterations (aka 2 Pivot-s) to get right
    • first dashboard iteration
      • aiming for Social Application Developer persona
      • took 6 months
      • learnings
        • people want customized reports
        • SADs don't like to pay
          • so going to target "everyone"
    • 2nd iteration
      • took 9 months (6 to build, 3 to run/test)
      • learned
        • people don't know what to measure
          • so customization is barrier
        • not all customers are profitable
        • customizability requires hand-holding
      • new target: marketers
        • focus on actionable metrics
        • want to improve conversion
        • willing to pay if it gets results
    • 3rd iteration
      • took 1 month
        • they set that as constraint/requirement
        • initially focused on developers - background tool
  • KissInsights
    • their new product, building with new process.... (see also Lean Canvas)
    • start with problem hypothesis
      • (x) kind of people people have problem doing (y)
    • wanted to learn
      • what doing now?
      • what are other tools leaving on the table
      • who is involved
      • how frequent/severe is the pain
    • learning process: 3wks
      • 20 phone interviews
      • 3 in-person interviews with paper prototypes
      • 2 landing pages
      • few hours to build MVP
      • built a little more, tested with 8 alpha testers
    • learned
      • people aren't doing customer research
      • they want private feedback
      • need easy-to-understand reporting
    • status
      • profitable, still running
      • but still mainly focused on KissMetrics
      • Jun'2012 update: bought by Sean Ellis' group
  • process learnings
    • your customers have the answers
    • get everyone dealing with customers/market
      • gives them context
      • even if they're just listening in when you do interviews
    • metrics
      • actionable: learn, make decisions
        • focus on funnel rates - AARRR
    • 2 teams
      • Problem team
        • what are the most important problems to solve?
      • Solution team (engineers)
        • find the MVP for a given problem
      • loop as fast as possible
    • the only path to perfection is iteration

Edited:    |       |    Search Twitter for discussion