← Back to the ledger
Entry · 2003–present

David Heinemeier Hansson

A $15-an-hour contractor in Copenhagen builds an internal tool in ten hours a week — and it becomes a company that turns down every venture capitalist who calls.

FieldSoftware
CompanyBasecamp / 37signals
Read4 min
Job at launchKept day job
Starting capital$0
Tipping pointScratch own itch
RouteSpun out of employer
Industry knowledgeInsider
By the numbers
Starting capital
$0
Time to first dollar
Under 1 month
Peak scale
Tens of millions ARR, never raised
What nearly killed it
No near-death event — notable as a counterexample

Every entry is researched against 30+ structured fields

CH. 01 — The Setup

In 2003, 37signals was a small web design consultancy in Chicago run by Jason Fried. The company had an ordinary problem: managing client projects across email was a mess, and every project management tool they tried was too complicated to use.

David Heinemeier Hansson was a twenty-three-year-old Danish programmer in Copenhagen. Fried hired him as a contractor, reportedly at $15 an hour, to build an internal tool for 37signals' own use. Nobody was building a product. They were solving their own operational annoyance.

DHH worked about ten hours a week — roughly 380 hours for the first version — and has since credited the shortage of time for the result. Constraint forced simplicity.

◆ THE PIVOT — The Tipping Point

The tipping point was the clients. 37signals used the internal tool with them, and the clients asked whether they could buy a version for their own businesses.

That request converted a cost center into a product. Basecamp launched on February 5, 2004, with a price attached. Fried's stated hope was that it might reach $5,000 a month within a year — enough to be worth the trouble. Within a month it had roughly a hundred paying customers. Within about eighteen months it was out-earning the entire web design business, and 37signals abandoned consulting.

There is a second tipping point inside the first. While building Basecamp, DHH built the underlying framework in Ruby, then extracted and open-sourced it in 2004 as Ruby on Rails. He has said there was no plan to release it — the tools simply spilled out of the project. Rails went on to power early versions of Twitter, Shopify, GitHub, and Airbnb.

CH. 02 — Getting Started

The unusual part is what they refused. Around 2005, by DHH's account, close to fifty venture firms made contact, along with acquirers. They turned all of them down.

He has been candid that resisting was not effortless. What made it possible was a small, no-control equity sale to Jeff Bezos — Bezos bought a minority slice from Fried and DHH personally. It gave them enough financial ballast that the large numbers being waved by investors lost their pull. DHH's argument is that once venture or private equity money is in, only three endings remain: implosion, acquisition, or IPO.

His other structural advice is geographic. 37signals was in Chicago, DHH in Copenhagen, and he has argued that physical distance from the Bay Area was what made a different set of choices thinkable.

CH. 03 — The Build

Through the mid-2000s they ran several products — Basecamp, Campfire, Highrise, Backpack — deliberately diversified in case any one faltered. Then in 2014 they reversed it, killing or spinning off the others to focus entirely on Basecamp, and renamed the company after it. The reason was not focus for growth's sake; it was that supporting four products across web and mobile would have forced them to become a larger company than they wanted to be.

The company has been profitable every year, has never taken venture funding, and operates with a few dozen employees generating tens of millions in annual revenue. Fried and DHH published Getting Real, Rework, Remote, and It Doesn't Have to Be Crazy at Work — which turned their operating philosophy into a second distribution channel and made them the public counterweight to growth-at-all-costs startup culture.

LEDGER NOTES — What to Take From It

The mechanism to steal is the internal-tool-to-product path. Basecamp had a validated customer before it had a price, because the customers were the people already watching 37signals use it. That is the lowest-risk product origin available, and it is sitting inside most service businesses unnoticed.

The second lesson is about the Bezos stake, and it is more actionable than it first appears. Partial founder liquidity — selling a small slice without giving up control — is what let them ignore offers that would have ended the company's independence. Founders who cannot afford to say no eventually say yes to something.

Third, note what constraint produced. Ten hours a week and $15 an hour is not a story about scarcity being romantic; DHH's claim is that the limit is why the product was simple enough to sell.

▸ THE PLAYBOOK — Run It Yourself

The framework: build the tool your own business needs, sell it when your clients ask, and stay small on purpose.

Move 1 — Inventory the internal tools, spreadsheets, and processes you have built to run your own work. The one your clients or peers have asked to borrow is a product with demand already proven.

Move 2 — Put a price on it before it is finished and see whether anyone pays. Basecamp launched with a price against a modest target and hit a hundred customers in a month; that is a real test, unlike a waitlist.

Move 3 — Decide your ceiling before anyone offers you money, and get physical or mental distance from whatever scene is telling you to raise. If you ever take capital, structure a small liquidity event rather than a growth round — enough to make saying no affordable.

Budget line: near-zero. The first version cost roughly 380 hours of contract work. The scarce resource is the willingness to stay smaller than you could be.

Sources & verification corroborated
  • Startups.com founder interview
  • 37signals corporate history
  • Ruby on Rails release history

The $15/hour rate and 380-hour build figure are sourced to DHH himself.

Last verified 2026-09-01

Keep going

Get one of these a week.

Same structure every time — the tipping point, the first ninety days, and a playbook with real numbers attached.