When you create a way to jump the liquidity in a marketplace, you can overtake the incumbent. AirBnb did this against Homeaway, by making people’s own homes inventory. Similarly, Doordash does this against GrubHub; by allowing restaurants to do delivery they can also take the other restaurants (Sarah Tavel on ILTB).
How to build a > $1bn dollar company? It comes down to understanding how to maximise engagement. Level 1 - grow engaged users completing a core action (Facebook = friending, Pinterest = pinning). Often the core action is most correlated with retention. Ask yourself, what features do you need to be to encourage the core action? Level 2 - Retaining engaged users. This comes down to creating accruing benefits and mounting losses within the product. Take Evernote, the more notes you make (core action) the more value you get (all your notes) and the harder it is to leave (my notes!). Level 3 - Self-perpetuating virtuous loops. Does your users engagement fuel your company forward? A network effect is one example (more pins > richer graph > better discovery). But there are also growth loops (discover pin > send pin > acquire user) and re-engagement loops (re-pin > notify user > re-engage user). You should maximise these naturally occurring loops. Level 3 is hardest. Evernote succeeds on 1 and 2 but fails on 3. Tinder has some loops, but also has some off ramps (a successful date means you’re off the platform). The most important thing to measure is cohorts - for a group of users, the number of them completing an action (both absolute and percentage). Cohorts show size, engagement and retention (Hierarchy of Engagement).