If you think this is }^}+]^ing bonkers fifty years into the widespread use of credit cards and would like to perform an in-flight upgrade of (a fraction of) the global economy, stripe.com/jobs
And believe me the first time you see the stats on error rates it is a very this meme moment:
“What percentage is it Patrick?”
I don’t think we quote it publicly but it is higher than your first guess.
I don’t think we quote it publicly but it is higher than your first guess.
Credit cards are still extremely important infrastructure, and I’m extremely glad the world has them, but they’ve tolerated pre-computer levels of operational errors for far, far too long.
A question I periodically find interesting: “What’s your best estimate of how many people in the world are working on X?”
For authorization rates, I think it is less than a thousand, total. Credit cards process trillions with a T dollars every year.
For authorization rates, I think it is less than a thousand, total. Credit cards process trillions with a T dollars every year.
One of the interesting lessons of working at an infrastructure company is the tension between centralized and decentralized infrastructure.
You might think e.g. “Ahh big federated network minimizes SPOF.”
You might think e.g. “Ahh big federated network minimizes SPOF.”
But plausibly a big federated network has 10k+ nodes which each see a problem costing them $1M a year and independently decide “Well that’s not worth putting an engineering team on.”
