Y Combinator: Patrick Collison Running Your Startup
We started working on Stripe in 2010. It took us quite a while to launch because we had to get all these banking partnerships in place. We were working on it for nearly 2 years before launching. [0:11]
When we launched we had about 100 users. Back then that seemed like a big deal. [1:19]
[Why they installed Stripe in person in the early days] We’d suggest people switch to Stripe and they’d say yes that sounds great. But then they’d postpone it and postpone it again. So us going and accosting them in person helped. [2:50]
Stripe is unusual - John and I are brothers and as a result of this relationship we are able to place a lot of trust in each other - we really do run the company together. We really do run the company together. There is no major decision that Stripe has made that we both have not been a part of. [5:33]
How Patrick and John handle disagreements: Our dispute resolution framework is which one of us cares more than which of us holds this title or that. Both of us have significant roles. The fact that I became CEO was semi-random. [5:54]
I think it is important for having an efficient mechanism for reaching a decision - it can’t be just oriented around consensus. That is a recipe for failure. Doing some quasi-democratic voting is probably not a great idea either. [7:06]
We try to make it about which one of us is more passionate about it. That will correlate with the outcome. Wanting something to be passionate can be a self-fulfilling prophecy. So I think it is not irrational to have that consideration. [8:02]
[Early on] we tried really hard to understand - in granular detail - where exactly people [customers] were tripping up at. [13:55]
Example of this: Every time somebody sent an API request to Stripe it sent an email to us. We were looking at every single API request. If we saw users do something weird we’d ask why did they do that. Or think about where our documentation was confusing. [14:47]
Generally speaking pre-product-market fit, metrics are relatively unhelpful. [17:07]
When I look back at life to the things that I am most glad I did - I wasn’t exactly happy while doing them. Often I was stressed out or I had to work very hard. But those are the things that brought the most fulfillment. [18:56]
If something is making you unhappy, if it doesn’t seem promising, or it’s just not working - your time has relatively high opportunity costs. You don’t get to start that many startups in your life- knowing when to call it quits is valuable. . . sometimes you should quit. [21:52]
Book recommendation: Satisfaction The Science of Finding True Fulfillment. [22:59]
It is definitely helpful to have competitors with not very good products. [23:30]
A better mental model to have: You are serving a market. Think of it as finite. There is a percentage of the market you are serving. Whatever percentage you are not serving - well you haven’t built the go-to market functions and organization that is capable of bringing the product to those market segments. The growth curve [or adoption curve] is just a function of that go to market apparatus. It is not some kind of cosmic trajectory. It is very much of your creation and under your control. [29:19]
You see when repeat founders go and start companies they are willing to build the organization ahead of where things are today. Which I think is exactly the right thing to do. They are thinking ok I have the right product now let’s work backward and think about what would the organization look like that was serving the entire market and let’s just start building that organization. . .Because again the growth market is under my control. It’s not 100% under your control but it is much more under your control then people think. [This applies more to b2b than consumer] [30:36]
Aaron Levie of Box stayed at our house warming party until 2 am telling us how much better it was to be building b2b software than consumer software. Consumer software is so hard to predict what consumers want. They don’t even know themselves what they want. Whereas businesses are mostly rational. You can work backwards. [32:44]
There are so many complicated questions where the only thing I am skeptical of is certainty in either direction. [40:17]
We are constantly self-flagellating over how slow we are. We are paranoid about degenerating into an immobile stupor. To any extent that we do get things done or appear fast is because we are very paranoid. [46:44]
The default outcome for companies as they scale is to become more ossified and rigid. Closed to new ideas and things that contradict their prevailing orthodoxies. [47:18]
The most successful larger organizations somehow do succeed at this iterative, repeated, process of augmentation. Amazon and Google being the most prominent examples. [50:02]