What I learned from being one of the first PMs at Stripe

Jen Huang
Co-founder & CPO, Fluxon
September 22, 2026
Share article
What I learned from being one of the first PMs at Stripe
Contents
Summarize this article

Copy the prompt, open the AI tool, then paste and send.

Open AI tool ↗

Ten years ago I joined Stripe as their fifth product manager. I was fresh from Google with a proven product-building playbook. It didn't work at Stripe.

Search Ads was already running at enormous scale, and one bug could cost the company billions. The PM function was mature. Google knew how to train PMs, how to run a launch, and who decided what. I learned how to study the data closely, navigate a complex organization, and build alignment across teams. A good idea wasn't enough. You had to prove it and bring people with you.

Stripe worked differently. Many of the teams I joined had never worked with a PM before, so there was no shared expectation of what I was there to do. There was much less process. Products went out faster than I was used to.

At first I thought the structure was just missing, and that part of my job was to supply it. It took me a while to see that most of it actually wouldn't have been useful. Google's process existed to protect something that already worked at enormous scale. When you're building something completely new, you have less to protect and far more to learn, and process added too early gets in the way of the learning.

When you're building something completely new, you have less to protect and far more to learn, and process added too early gets in the way of the learning.

So I had to work differently. Get comfortable with ambiguity, and be willing to invent an approach instead of going to look for the standard one. One example is regulation. Payments is heavily regulated, but the rules rarely tell you how a product should work. They tell you what you need to know about a customer. They say much less about when you have to ask. We might need certain information from a customer. But do we need it immediately at sign-up? Or could we ask once they're actually processing meaningful volume? Those kinds of choices made a real difference to how easy Stripe was to use.

It’s one reason I think PMs benefit from seeing very different kinds of companies up close. A lot of what feels like “the right way” is really just the right way for a particular company, at a particular stage.