Skip to content
Founder Tools

Founder storiesDeveloper products

How the Collison brothers built Stripe with their first users

Patrick and John Collison put a small payments product in production well before Stripe’s public launch. Early users showed them what to build next.

From idea to public launch
2009
Patrick and John discuss the idea.
Early 2010
A customer uses Stripe in production.
Summer 2010
The founders go full time.
September 2011
Stripe opens to the public.

Early milestones as recounted by Patrick Collison.[1]

Lessons for founders

  • Stripe had a live customer in early 2010 and launched publicly in September 2011, according to Patrick Collison.
  • An early user’s follow-up requests exposed important jobs beyond accepting a charge.
  • Design a short path to the first valuable outcome and watch where real users stop.

A payments problem they had experienced

Patrick and John Collison had dealt with the friction of moving money online while working together on Auctomatic. In a later conversation with Y Combinator’s Root Access, Patrick traced the decision to build Stripe to a conversation with John after a 2009 Startup School event. The brothers started work while in college, then went full time in the summer of 2010.

Patrick recalled that 280 North was using Stripe in production by January 2010. At first it could do little more than charge a card. That was enough for their customer, Ross Boucher, to reveal the next steps: viewing charges, issuing a refund and receiving the proceeds. Each request came from someone already trying to finish a real job.

Sources1

The private beta was a learning loop

Stripe launched publicly in September 2011. Patrick described adding customers throughout the private beta and learning from their requests as the team put reliability, security and financial infrastructure in place. New customers gave the founders repeated opportunities to see whether an earlier fix made the next integration easier.

For a founder, the useful distinction is between release size and contact with users. A narrow release can produce regular evidence if customers are doing real work with it. Waiting behind closed doors produces a different kind of risk. Pick a small group you can support, give them a clear outcome and set a date to review what happened.

Sources1

What the small demo communicated

Stripe engineer Michelle Bu revisited the company’s 2011 homepage in a retrospective about the payments API. The original example occupied nine lines and demonstrated a successful charge. It gave developers a direct way to see what the product could do before they worked through a full integration.

That demo answered a buyer’s first question: how quickly can I get something working? It also left a second question to the customer’s next steps: can I rely on this for the whole job? A useful onboarding flow has to answer both. The opening demo needs a fast win, then clear guidance for the less glamorous states a customer will eventually face.

Sources2

Run your own first-success review

For a developer tool that turns usage logs into alerts, give one customer a test input and ask them to produce their first alert while you watch. Note where they hesitate, whether they can tell the action succeeded and what they need next. If they ask who received the alert or how to stop duplicates, write down the underlying job before adding a setting.

Onboard the next customer with the revised instructions and compare the two sessions. Record how long the first useful result took and which steps still required help. The point is to turn your support time into a better default path, one repeatable step at a time.

Tools for this article

  • Stripe Payment Links

    Try the modern no-code checkout flow when testing a paid pilot with an early customer.

Sources

2 references