Tobias Lütke wanted to change one thing on his online store: the background color of the top frame. That's it. He wasn't asking Yahoo Stores for a new layout or custom code. Just a different color.
Yahoo Stores said no.
The refusal ended up creating a company now worth over $150 billion. Now we all know the company as Shopify.
The snowboard shop that wasn't
Lütke was 24, living in Ottawa without a work permit, trying to run an online snowboard shop called Snowdevil with his friend Scott Lake. Every e-commerce platform on the market in 2004 felt built for a different decade. Miva was clunky and expensive. osCommerce needed a developer just to keep it running. Yahoo Stores let him change a background color and nothing else.
So he wrote his software using Ruby on Rails, a framework almost nobody outside a small developer community had heard of yet. Two and a half months later, Snowdevil had a working store, a shopping cart, and an admin panel Lütke actually enjoyed using.
Then something odd happened. Merchants across North America started emailing about the website itself. A retailer in Indiana wanted to license it. A developer in Toronto wanted to know what was powering it. Barely anyone asked about the snowboards.
Lütke and Lake had built a snowboard company and stumbled onto a software company hiding underneath it. They shelved the snowboards, and Shopify launched in 2006.
The two decisions nobody talks about
Writing decent software wasn't actually the hard part. Plenty of developers in 2004 could have built a working shopping cart. What turned Shopify into a $150 billion company and left most 2004-era shopping cart tools as footnotes came down to two calls that rarely make it into the retelling.
The first was fixing how they got paid. Shopify launched charging a commission on every sale a merchant made. It sounds reasonable on a whiteboard. In practice, it taxed success: the better a store performed, the more it owed Shopify, which meant the company was quietly discouraging the exact growth it wanted to see. In 2007, they scrapped commissions for flat monthly subscriptions, with transaction fees that shrank as merchants scaled up. Revenue and merchant success finally aligned, and growth accelerated almost immediately.
The second call landed in 2009, on the company's third birthday: an API and an app store. This chapter was the most important. Between 2004 and 2008, a wave of scrappy web tools launched, and each solved one problem well. Flickr sold to Yahoo for $35 million. Feedburner sold to Google for $100 million. Digg and Delicious both got broken apart and sold for a fraction of what they'd raised. All useful tools. None of them became platforms.
Shopify's app store encouraged outside developers to build on the platform, turning potential competitors into partners. Every new app made merchants stickier. Every merchant made the app store more valuable to developers. Developers building for merchants, merchants attracting more developers: that loop is the real moat, the gap between owning a tool people use once and owning the infrastructure they build careers on.
The year everything should have ended
None of this felt inevitable while it was happening. In the early years, Lütke sat in planning meetings discussing features that would take a year to ship, while privately knowing the company had less than a month of payroll left. He couldn't say that out loud. One wrong word and the team could have scattered.
Then in 2008, the broader economy collapsed, and Shopify's user numbers climbed. Every VC who'd shown interest pulled their term sheets. At the same time, people who'd just lost jobs across the country started businesses of their own, because a paycheck was no longer guaranteed and a Shopify store cost almost nothing to open. By 2009, the company had hit cash-flow positive for the first time. The event that was supposed to kill Shopify turned into the reason it survived.
Today Shopify processes close to $380 billion a year in merchant sales, powers stores in more than 175 countries, and briefly became the most valuable public company in Canada. All of it traces back to a background color Yahoo Stores wouldn't let a snowboard seller change.
Founder takeaways
If you are a founder or running a startup, here are some key points to consider:
1. Price around what your customer wants, not what you can extract.
Shopify's commission model looked reasonable until it started punishing merchant growth. A pricing structure that penalizes your customer's success eventually caps your own.
2. Open your platform before scale makes it obvious.
The API and app store launched when Shopify had roughly 2,000 accounts, most of them tiny. Waiting until you're big enough to justify a platform usually means a competitor builds it first.
3. Watch what customers do with your product, not just what they ask for.
Nobody at Snowdevil ran a survey asking if the world needed e-commerce software. The signal came from unsolicited emails about a tool nobody was trying to sell.
4. A downturn filters for real demand.
The 2008 crash didn't manufacture demand for Shopify. It revealed how many people were already sitting on a business idea, waiting for a reason to act. Products solving durable problems tend to hold up, sometimes accelerate, when the economy doesn't.
5. Say "I don't know" before you fake it.
When VCs asked Lütke about metrics he'd never used, he wrote the terms down, looked them up overnight, and came back with real numbers pulled straight from Shopify's own database. Investors backed the honesty as much as the business.
Know a founder who needs to read this? Forward it their way.
