Notes from building Papermark
A few things I've internalized running Papermark, the open-source secure data room I've been building for the last ~2.5 years.
Open source is distribution
When we started, I didn't think of open source as a marketing channel. It was a values thing — I like open tools, I'd used a lot of them, I wanted Papermark to be one. The distribution side was a happy accident that turned out to be the most durable thing we have.
A GitHub repo gets indexed differently than a landing page. People find you through issues, through blog posts other developers write, through forks, through "alternatives to X" lists. Every one of those is a backlink that compounds. I've never paid for a backlink and Papermark has thousands. The repo does the work while I sleep.
The trade-off is real — anyone can self-host, see your roadmap, or copy your code. But that's an imaginary tax. In practice, the companies that want self-hosted will never be your paying customers anyway, and the visibility you get from being public is worth more than the tiny slice of revenue you theoretically "lose."
Ship something every day that someone will notice
Not every day produces a feature. Some days it's a doc, a fix, a tweet, a cold email, a customer call, a new integration. But something should leave the building every day. Momentum is a real physical property of a small company. When it slows, it takes three times the effort to start it again.
The worst weeks at Papermark have always been the ones where I was "working on something big" and nothing actually shipped. The best weeks were tiny, boring, and continuous.
"Overnight" is just compounding you can't see
It took us 18 months to hit $1M ARR. 6 more months to hit $2M. From the outside that looks like a hockey stick. From the inside it's the same motion we were doing in month 3, just with more compound interest on top.
There was no moment. There was no launch that changed everything. There was a year where nothing seemed to be working and I kept going anyway because the alternative was stopping, and stopping felt worse than grinding. Then a feature landed that more customers wanted than we expected. Then the integration with a specific tool brought in a cohort we hadn't reached. Then SEO started paying off. None of those were "the thing." The thing was still being around when they started to add up.
Build for a boring market
Data rooms are boring. That's the feature. Boring markets have real budgets, clear jobs-to-be-done, and customers who will tell you exactly what they need if you ask. I spent years trying to build in markets where the TAM was enormous and the buyer was fuzzy. It doesn't work. Pick something unglamorous with a sharp buyer, ship faster than the incumbents, and you have a real company.
That's most of what I'd tell anyone starting their own thing today: be public, ship constantly, be patient about the curve, and pick a boring market.
I'm writing more of these as I go — follow along on X or LinkedIn if that's your thing.