Great insights here to some of the workings of Stripe from Patrick. Though I’ve never been on that ride, the hypergrowth company (even with its flaws and stumbles) is an absolute marvel that it can be made functional at all.

Stripe seems exceptional in many ways, but in others it’s just a solid example of the “doer” culture that great SV tech companies are known for. Stripe’s Atlas product now looks like a polished, full-featured platform for creating a company, but as Patrick notes in this example, even well-funded tech companies have to be frugal, iterative, and resourceful to get ideas off the ground:

When I was on Stripe Atlas, a small, focused team with many high-horsepower generalists who largely didn’t have a huge amount of entrepreneurial experience, part of my job was bringing skills and connections and part was just standing up portions of our offering by sheer force. We wanted helpful advice for founders and didn’t have it; I locked myself in a room for a month and wrote a 30,000 word guide plus the ERB to put it on the Internet. We wanted to inculcate an Atlas community; I installed Discourse, wrote our SSO code for it, sent out invites, and commented on every thread for months.

On the default “bias to action” as part of this culture:

The returns to pushing your cadence to faster are everywhere and they compound continuously, for years. Don’t send the email tomorrow. Don’t default to scheduling the meeting for next week. Don’t delay a worthy sprint until after the next quarterly planning exercise. Design control and decisionmaking structures to bias heavily in favor of preserving operating cadence.

I don’t think Stripe is uniformly fast. I think teams at Stripe are just faster than most companies, blocked a bit less by peer teams, constrained a tiny bit less by internal tools, etc etc. There are particular projects which have been agonizingly long to ship; literally years after I would have hoped them done. But across the portfolio, with now hundreds of teams working, we just get more done than we “should” be able to.