Lessons from real companies

I could write these lessons as a list of abstract principles and they would read fine. But principles detached from where they were learned tend to flatten into platitudes — "listen to your customers", "measure what matters" — the kind of thing that sounds true everywhere and changes nothing anywhere. So instead I've organised what I know by the place I learned it: a two-sided marketplace at Just Eat, a churn-and-renewal business at Sage, a support queue at Echo-U, a venue SaaS platform I founded and scaled myself at Wocal, and a carpooling platform I'm building right now at Co-Ride.

Each of these is a different kind of company with a different kind of pressure. Just Eat taught me about liquidity and network effects at a scale I couldn't have simulated. Sage and Echo-U taught me what actually decides whether an enterprise customer renews, and how often that decision has nothing to do with the product roadmap. Wocal taught me what it costs — in time, attention and money — to build and sell something from nothing, and what a founder does differently from an employee with a similar title. Co-Ride is teaching me, in real time, how much of that earlier learning transfers to a different market and a different kind of network.

What carries across all of them is less a technique than a discipline: watch what people actually do, not what they say they'll do; be honest about which metric you're really optimising for; and treat every "obvious" feature request as a hypothesis to test rather than a instruction to follow. The detail below is where that discipline came from.