A Good Company

I was 22 and had been at Thoughtbot for maybe three months. Designer #2 at one of the world's most respected design and development consultancies. They'd built systems for major tech companies, had this incredible reputation for quality, were known industry-wide for their work.

And I thought: I can do this. Why am I waiting?

Most people would call that audacious or naive or both. But I'd seen how a great agency worked from the inside. I'd watched them brainstorm, collaborate, ship. And instead of waiting years to work my way up through their ranks, I decided to start my own thing.

I called my friend Caroline. We'd worked together before. "What if we started our own design studio?"

The next day I gave my notice. We founded A Good Company.

Looking back, I'm not sure if it was confidence or just not knowing what I didn't know. Probably both. But for six years, from 22 to 28, A Good Company was how I learned everything.

How to be a creative director. Caroline was the designer. I was always more interested in the bigger picture—the strategy, the client relationships, the technical side. On client projects, I'd direct her work: what direction the branding should go, how the website should function, what the user experience needed to be. It was probably my first taste of what creative direction actually felt like. Not doing all the design work myself, but guiding someone else toward a vision.

How to manage clients. We worked directly with founders and decision-makers. No account managers, no layers. Aza Raskin sought us out for his Massive Health work. The Collison brothers when Stripe was just three people. Mozilla when they were rebuilding Firefox. These weren't massive corporations with procurement departments—they were smart people at inflection points who needed thoughtful partners.

You learn fast when you're talking directly to the people making decisions. You learn what matters. You learn how to defend your thinking. You learn how to say no when a project doesn't fit.

How to build products from scratch. This is the part I couldn't stop doing. While we were running client work to pay rent, I was obsessively building side projects. The League. A Good Portfolio. Good Timing (a time-tracking app that used natural language). HQueue (an early App Store app for Hulu that never took off but taught me mobile development).

None of them ever really made us money. But I couldn't help myself. Client work taught me creative direction and client management. But product work—designing and building things from the ground up, by myself—that's where I learned to actually make things.

Caroline didn't participate much in the product side. She wasn't interested in the technical work, and these were my obsessions anyway. But that split taught me something important: I was always going to be the person who wanted to design and build. Who couldn't just art direct without understanding the code. Who needed to work across both domains.

The client work was feast or famine. We'd land Mozilla or Stripe or NPR, work intensely for months, then have nothing for a while. Then another big project. Then quiet again. We never cracked the code of making it a predictable business. We were never rich. But it sustained us.

Mozilla hired us to design their extension marketplace and API system. Making complex technical documentation clear for thousands of developers. Creating a marketplace where people could actually find useful add-ons without drowning in options. Brand work at technical scale.

Stripe consulted with us when they were figuring out their first brand. Three people who'd built a payment API and needed to explain it to the world. We pitched a bigger project that didn't end up happening, but being sought out by them at that stage validated something.

NPR and Public Radio Exchange needed a platform redesign. Multi-sided platform thinking: producers, stations, listeners all using the same system with different needs. Content discovery in the public media context.

We also got contracted by the state of California to build a political grassroots network. We built it. It worked. They never finished paying us because politicians cycled out and decided not to keep their end of the bargain. You learn from that too.

After six years, Caroline realized she didn't want to be in web design anymore. She wasn't interested in learning the technical side. Didn't want to be on the client or business side either. She wanted something else from her life.

We parted ways. She went on to become a baker, then a very successful children's book illustrator, then into podcasting and journalism. She's living happily in Maine now, doing all three. I'm genuinely glad she found what she was looking for.

At 22, I thought I could run an agency because I'd watched how Thoughtbot worked for three months. Audacious. Maybe naive.

At 28, I'd directed design work for Mozilla and Stripe. Managed relationships with founders at inflection points. Built The League from scratch—designed it, coded it, grew it to thousands of users. Learned what works when you're talking directly to decision-makers versus when you're building something alone.

Client work taught me creative direction and strategic partnerships. Product work taught me how to actually build things, not just design them. The split—Caroline on design, me on everything else—forced clarity about what I needed: to work across design and code, not choose between them.

Six years compressed what takes most people fifteen. Not because I was special, but because starting at 22 meant learning by doing, every single day, with no safety net and no one else to solve problems I didn't know how to solve yet.

The foundation was set.