Stripe doing things that don't scale

In the world of startups, there’s a well-known piece of advice from Y Combinator’s Paul Graham: “Do things that don’t scale.” This counterintuitive wisdom suggests that in the early days of building a company, founders should embrace labor-intensive, manual processes that could never work at scale—but might be exactly what’s needed to get initial traction. Few companies embody this principle more vividly than Stripe, the payment processing platform that grew from a scrappy startup to a $95+ billion company by starting with some decidedly unscalable methods.

The Collison Brothers: Teen Millionaires Before Stripe

The story of Stripe begins with Patrick and John Collison, two remarkably talented brothers from the small village of Dromineer, Ireland (population: 102). Their journey to founding Stripe was anything but conventional.

In 2005, at just 16 years old, Patrick Collison won Ireland’s Young Scientist of the Year competition for creating a programming language. This early achievement introduced him to the programming language Lisp, which connected him to the ideas of Paul Graham, a prominent Lisp advocate who would later co-found Y Combinator.

Patrick’s academic brilliance earned him admission to MIT at 17, a full year and a half before finishing high school. Meanwhile, his younger brother John was taking an “activation year”—an Irish concept where students take time off between middle and high school to explore life beyond academics.

During this time, John began working on an idea for a better eBay, and Patrick decided to help. The project was compelling enough that Patrick left MIT after his first quarter to work on it full-time with his brother. Despite their youth—Patrick was 17 and John was just 15—they were accepted into Y Combinator’s Winter 2007 batch.

Their startup, eventually named Auctomatic, provided tools to help sellers on platforms like eBay and Amazon manage their listings. Within just 11 months, they sold the company for $5 million, making them teenage millionaires. John returned to Ireland to start high school, while Patrick worked for a year at the acquiring company before returning to MIT.

The Genesis of Stripe

While back at MIT, 19-year-old Patrick began tinkering with various side projects. One problem that consistently frustrated him was how difficult it was to accept payments online. The existing solutions required weeks of setup, complex integrations, and negotiations with banks and payment processors. For developers building new applications, this was a significant barrier.

Patrick saw an opportunity to create something like Slicehost (a user-friendly hosting service he admired) but for payments—an abstraction layer that would make it simple for developers to integrate payments into their applications.

During winter break in January 2010, Patrick and John (who was now at Harvard) traveled to Buenos Aires, Argentina, to focus intensely on building the first version of what would become Stripe. They had secured a small investment from Y Combinator, but they were still figuring out exactly how the product would work.

The brothers had no background in finance or the payment industry. What they did have was coding expertise and a clear understanding of the developer experience they wanted to create. They built a set of APIs that abstracted away the complexity of payment processing, reducing what had been hundreds of lines of code and weeks of setup to just seven lines that could be implemented in minutes.

The First Customer and the Birth of the “Collison Installation”

The first customer of Stripe was Ross Boucher, a friend of Patrick and John from their early Y Combinator days. But getting from that first user to a sustainable business required a much more hands-on approach than simply launching a product and waiting for users to come.

This is where the Collison brothers developed what Paul Graham would later dub the “Collison installation” technique. As Graham described it in his essay “Do Things that Don’t Scale”:

“At YC we use the term ‘Collison installation’ for the technique they invented. More diffident founders ask ‘Will you try our beta?’ and if the answer is yes, they say ‘Great, we’ll send you a link.’ But the Collison brothers weren’t going to wait. When anyone agreed to try Stripe they’d say ‘Right then, give me your laptop’ and set them up on the spot.”

This approach was remarkably effective for several reasons. First, it eliminated the friction between interest and implementation. There was no chance for potential users to forget about Stripe or put off the integration. Second, it allowed the founders to observe firsthand any confusion or difficulties users had with the setup process, providing immediate feedback they could use to improve the product. And third, it created a personal connection with early users, who were more likely to provide ongoing feedback and refer others to the service.

The Collison brothers were particularly strategic about who they targeted with these installations. They focused on other Y Combinator startups, recognizing that these companies were likely to be building new applications that needed payment processing and would appreciate a developer-friendly solution. This gave them a concentrated pool of potential users who were accessible and open to trying new technologies.

Beyond Installation: Obsessive Customer Service

The “Collison installation” was just the beginning of Stripe’s unscalable approach to growth. Once they had users on the platform, the brothers provided an extraordinary level of customer service that went far beyond what most companies would consider sustainable.

They personally responded to support emails at all hours, often within minutes. They jumped on calls with users who were having issues, walking them through solutions step by step. They even wrote custom code for specific users to help them solve unique problems.

This obsessive focus on customer service had several important effects. It created intensely loyal early users who became advocates for the product. It provided the founders with deep insights into how developers were using Stripe and what problems they were encountering. And it established a company culture that valued customer experience above all else—a value that would remain central to Stripe even as it scaled to serve millions of businesses.

As John Collison explained in an interview, “For us it was quite visceral: these [payment] products are not serving the needs of the customers, so let’s build something better… If you’re a developer building the next Kickstarter, or the next Lyft, and you have a two-person team, both of you writing relatively complex code and solving complex infrastructural problems, you need a simple payments API that—once installed—doesn’t keep changing.”

This deep understanding of their users’ needs came directly from their hands-on approach to customer acquisition and support.

The Power of Word-of-Mouth Growth

Stripe’s unscalable, high-touch approach to early customer acquisition laid the groundwork for powerful word-of-mouth growth. Developers who had experienced the ease of Stripe’s integration and the exceptional level of support became evangelists for the product.

In the developer community, recommendations carry significant weight. When respected developers began talking about how Stripe had transformed their payment processing experience, others took notice. The company didn’t need to spend heavily on marketing because their users were doing it for them.

This word-of-mouth growth was particularly effective because Stripe had targeted influential developers and startups from the beginning. Many of their early users were building products that would themselves become successful, increasing Stripe’s reach as these companies grew. For example, when companies like Lyft, Instacart, and Shopify chose Stripe for payment processing, they introduced the service to thousands of other businesses in their ecosystems.

The founders also contributed to this organic growth by being active participants in the developer community. They wrote blog posts, spoke at conferences, and engaged in technical discussions online. This established them as thoughtful technologists who understood the challenges developers faced, further enhancing Stripe’s reputation.

Expanding Beyond the Initial Use Case

As Stripe began to gain traction, the founders recognized that they needed to expand beyond their initial focus on simple payment processing. They began building additional features and products that addressed adjacent needs of their growing user base.

They launched Stripe Connect, which allowed marketplaces and platforms to manage payments between multiple parties. They created Stripe Atlas, which simplified the process of incorporating a business and setting up a bank account for international founders. They developed Stripe Radar for fraud detection and Stripe Billing for subscription management.

Each of these expansions followed the same pattern as their original product: identify a complex, frustrating process that businesses face, and create a simple, developer-friendly solution. And in each case, they continued to provide the same level of hands-on support and engagement with early users, even as the company grew.

This expansion strategy allowed Stripe to increase its value to existing customers while attracting new ones. A business might initially adopt Stripe for basic payment processing, but then add subscription billing, fraud protection, and other services as they grew. This increased the “stickiness” of the platform and raised the average revenue per user.

The Transition to Enterprise

While Stripe began by focusing on startups and small businesses, they eventually needed to move upmarket to continue their growth. Selling to enterprise customers required a different approach than the “Collison installation” technique that had worked so well with developers.

Enterprise sales cycles are longer, involve more stakeholders, and require addressing concerns around security, compliance, and reliability that might not be top priorities for smaller companies. Stripe had to build a sales team, create enterprise-specific features, and develop the infrastructure to support large-scale customers.

However, even as they made this transition, they maintained their focus on creating a superior user experience. They applied the insights gained from their hands-on early days to build enterprise products that were still significantly easier to use than traditional payment solutions.

They also leveraged their reputation and relationships in the startup world to gain introductions to larger companies. Many of the startups that had adopted Stripe in their early days had grown into substantial businesses themselves, and their positive experiences with the platform made them willing to recommend it to others.

Lessons from Stripe’s Unscalable Beginnings

Stripe’s journey from the “Collison installation” technique to a global financial infrastructure company offers several valuable lessons for entrepreneurs:

1. Do things that don’t scale. The hands-on approach the Collison brothers took to onboarding users could never have worked for millions of customers, but it was exactly what they needed to get their first hundred. Don’t be afraid of manual processes in the early days if they help you gain traction.
2. Focus on user experience. Stripe’s obsession with creating a simple, developer-friendly product set them apart in an industry known for complexity and poor user experience. Identifying and solving genuine pain points can create passionate early adopters.
3. Target the right initial users. The Collison brothers were strategic about focusing on other Y Combinator startups and developers who would appreciate their solution and could provide valuable feedback. Identifying the right early adopters can accelerate product development and word-of-mouth growth.
4. Provide exceptional support. The personal, responsive support the founders provided created loyal users and valuable insights. In the early days of a startup, the depth of relationships with users can be more important than the number of users.
5. Build for expansion. Stripe started with a focused solution to a specific problem but had a vision for how they could expand to address adjacent needs. This allowed them to increase their value to existing customers while attracting new ones.

The Legacy of the “Collison Installation”

Today, Stripe processes hundreds of billions of dollars in payments annually for millions of businesses around the world. The company has raised billions in funding at valuations exceeding $95 billion, making it one of the most valuable private companies in the world.

This massive success grew from those early days when Patrick and John Collison were personally installing their payment system on users’ laptops. The term “Collison installation” has become part of the startup lexicon, a shorthand for the kind of high-touch, unscalable approach that can help early-stage companies gain their first users.

The technique exemplifies Paul Graham’s advice to “do things that don’t scale,” demonstrating how labor-intensive, manual processes can be crucial for getting a startup off the ground. As Graham noted, “Actually startups take off because the founders make them take off. There may be a handful that just grew by themselves, but usually it takes some sort of push to get them going.”

For Stripe, that push was the willingness of its founders to sit down with potential users, take their laptops, and set up the service on the spot. This unscalable approach laid the foundation for what would become one of the most successful and influential technology companies of its generation.

In a world where founders often focus on automation and scalability from day one, Stripe’s story is a powerful reminder that sometimes the best way to build something that can eventually scale to millions is to start by doing things that don’t scale at all. The “Collison installation” technique may seem quaint compared to Stripe’s current global operations, but without it, the company might never have gotten off the ground.