[Stripe cofounder]
Thanks! We're always paranoid that we're too slow, for whatever it's worth.
We're still iterating on the supporting models and frameworks. We think a lot about pace layering -- how do we have teams that think on multi-year horizons (infrastructure, security, etc.) alongside teams that are rapidly iterating at much earlier stages of development. I think a lot goes wrong when organizations insist on too much consistency in their operating approach.
We've now identified a few semi-formal mechanisms that have helped substantially. (For example: larger companies tend to want to aggressively standardize roles. This makes sense when operating at scale but less so in the early days. So, we've made it easy to hire for ill-defined roles in nascent product areas -- we figure we can codify those roles later as needed and that doing so sooner is a premature optimization.) We'd certainly consider writing a blog post once we feel that they might be generalizable.
(If anyone reading this is interested in helping us figure things out, please do apply. We're particularly interested in hiring engineers who are interested in these questions.)