Jason Fried on bootstrapping, staying small, and building on instinct

Executive overview

Most software founders default to raising venture capital, assuming growth requires outside money. Jason Fried argues the opposite: constraints force the skills that make a business last.

37signals has been profitable for 24 years with 75 employees, ~100,000 customers, and no investors. The company runs on small teams, six-week cycles, gut-driven decisions, and a flat price structure that eliminates enterprise complexity.

Independence is the root condition — everything else follows from not being beholden to anyone else's timeline or expectations.

Why bootstrapping beats venture for most software businesses

• Software has inherently high margins; Silicon Valley has found ways to eliminate them through headcount and customer acquisition spend
• Venture-backed companies have essentially one outcome available: go huge — the space between small and unicorn is off the table
• Bootstrapping forces founders to practice making money; VC money lets them practice spending it instead
• The more time under the curve making money, the better founders get at the fundamental skill a business needs
• Negative visualization: be at peace with the worst case before you begin — if the downside is acceptable, move forward
• Keep costs ruthlessly low in the early days; control your costs because you cannot control the market

How 37signals stays small and profitable

• 75 people serving ~100,000 customers; competitors with similar customer counts employ 1,500–3,000
• No salespeople, no enterprise tiers, no custom deals — one price, one codebase, one product for everyone
• Maximum price on Basecamp is $299 with unlimited users, eliminating enterprise support complexity
• Constraints and simplicity are features, not limitations — adding people would make the company worse, not better
• Success metric: did we want to do that again? Collectively, did we make more than we spent?
• No OKRs, no revenue targets, no growth goals — just profit and work worth repeating

ShapeUp: how product development works

• Every feature is built by exactly two people: one programmer and one designer
• Work is scoped by appetite, not estimate — six weeks is the maximum budget, not a projection
• Work expands to fill available time; setting an appetite forces the simplest effective version
• If work is still on the "uphill" side of the hill chart at deadline, it typically dies — no extensions
• After each six-week cycle, a two-week cooldown replaces back-to-back sprints; people tighten loose ends and shape the next cycle
• Adopt ShapeUp on low-criticality projects first; failing on something critical poisons future attempts

Gut, instinct, and decision-making

• Every decision is a judgment call — data is one input among thousands, many of which are invisible to the decision-maker
• Asking "how does this feel?" is more honest than demanding proof; humans are primarily feeling creatures
• Hiring for gut: give finalists a paid one-week project, then ask where they'd take it in two more days — look for instinctive riffing, not just execution
• Demonstrating gut-driven decisions from the top creates space for others to trust their own instincts
• Long-term planning is a fantasy; obligations in the future send people down paths they no longer want

Work isn't war

• Business language borrows heavily from warfare: target customers, conquer markets, destroy competition, make a killing
• War metaphors shape how people feel about their work — they create a dark, adversarial lens
• Replace war language with a spirit of creation: make something great, provide an alternative, be proud of the work
• "Underdoing" the competition — less complexity, lower price, simpler product — is a viable and often neglected strategy

Once: a bet against SaaS subscriptions

• SaaS means renting software forever; most of what users pay for month-to-month is unchanged from four years ago
• Once is a new product line: pay once, download the software, install on your own server, no recurring fees
• First product: Campfire, a group chat tool — 90% of what Slack does, at a one-time price under $1,000
• Buyers receive the source code and can modify it; no resale, but full access to a well-built reference implementation
• Target: commoditized software categories still charging luxury prices — provide a high-quality, opinionated generic
• Campfire can run fully air-gapped, making it viable for security-sensitive environments

Principles for building a stay-up

• "Startup" celebrates what's easy — starting; "stay-up" celebrates what's hard — remaining in business for years
• Independence is the root: 37signals lists it as principle one because all other freedoms follow from it
• Simple beats complicated; things are relatively simple until you make them complicated
• Play the infinite game: build a company you want to keep running, not one you're optimizing for exit
• Business advice from Jason's father: no one ever went broke making a profit