The seven lines of code that became Stripe

When Patrick and John Collison started what would become Stripe, they didn't lead with a pitch deck. They led with a curl command.

The entire onboarding flow for a competing payments product in 2010 looked like this: fill out a 40-page merchant application, wait two weeks for underwriting, pass a PCI compliance audit, integrate a 200-page SDK, deploy. Best case, you were six weeks from your first transaction.

The Collison brothers shipped a private beta where the entire integration was a single POST request:

``` curl https://api.stripe.com/v1/charges \ -u sk_test_xxx: \ -d amount=2000 \ -d currency=usd \ -d source=tok_visa \ -d description="Test charge" ```

Seven lines. No paperwork. No audit. Get an API key, charge a card, done. They handwrote sign-up emails to specific developers they wanted on the platform. When a developer hit a bug, Patrick would respond in under an hour — often with the fix already deployed by the time the developer hit refresh.

The insight wasn't "payments should be easier." Every payments competitor said that. The insight was "payments should be a thing engineers can ship in a single afternoon, before asking permission from anyone." Stripe wasn't competing with Authorize.net on price or features. It was competing with "we'll just hold off on payments until next quarter."

The seven-line example became the entire homepage. For five years, the Stripe landing page above the fold was a code snippet. Not a hero shot. Not testimonials. A code snippet — because the target user trusted code more than copy.

The broader lesson is about who you optimize the first impression for. The Collisons knew that if a developer ran the curl command and it worked, that developer would later become the person inside their company who picked the payments stack. Stripe's growth strategy for five years was: make one engineer happy, let them carry you in.