Concept

Shape Up

conceptshape-upproduct-developmentprocess37signalsbasecampmulti-source

Shape Up

The product-development methodology built and used at 37signals (Basecamp, HEY), published free at basecamp.com/shapeup. It replaces estimates, backlogs, and tickets with fixed appetites, shaped work, and small autonomous teams working in cycles. Developed by Jason Fried and Ryan Singer, among others.

Appetites, not estimates

The foundational inversion. An estimate asks how long a feature will take — and because work expands to fill the time available, estimates slip. An appetite is how much time you are willing to spend; you fix the time and vary the scope to fit. The ceiling is six weeks: short enough to see the end from the beginning, long enough to ship something meaningful. Not everything gets the full six — some work has a one-week appetite. The discipline forces the team to find the simplest effective version.

The six-week cycle and cool-down

Work runs in cycles, not sprints. Fried is pointed about the word: a sprint leaves you exhausted and unable to repeat it, whereas cycles recur like seasons, so the team is built to do it again. Each cycle is followed by a two-week cool-down: teams self-direct, tidy what shipped, fix small bugs, and scratch itches, while a few people shape the next cycle’s pitches.

Teams of two

Every feature is built by two people — one designer, one programmer — with full latitude to decide how. They are not handed a spec or tickets; they receive shaped work and create their own task breakdown. This keeps decisions fast and the team close to the work.

Shaping

Before a cycle, the work is shaped: designed at the right level of abstraction — concrete enough that the team knows what to build, loose enough that they decide how. Singer’s emphasis: shaping is collaborative and live, not documentary. The right room is a product person, a senior engineer who ‘knows where the bodies are buried’, and a designer; the output is an idea expressible in fewer than ten moving pieces. The session’s real job is to surface unknowns before the clock starts — find the rabbit holes and time bombs before engineering time is committed (the grumpy plumber who insists on seeing the pipes before quoting).

Hill charts and trading concessions

Basecamp visualises a task as a hill. The left slope is problem-solving (‘still figuring out how’); the top is the turn; the right slope is pure execution. As work proceeds, teams and shapers trade concessions — negotiating scope so the work stays alive without creeping. At the deadline: work near the bottom of the right slope can get a couple of extra days, but anything still on the left side almost certainly dies. Killing it is the system working — it prevents the never-ending, morale-destroying project.

The PM moves upstream

Singer’s observation: in a Shape Up team at stride, the product manager spends less time chasing the build and more time narrowing the problem — understanding business context, negotiating scope with leadership, and choosing which slice of a customer struggle is worth solving. Jobs-to-be-Done research is the framing step before shaping begins.

Adopting it — don’t turn on a dime

Both founders warn that Shape Up is hard to adopt and routinely fails when forced. Fried’s physics metaphor: momentum is mass times direction; a company that has worked one way for years cannot turn the corner immediately. Practical advice: try it first on a low-criticality project (expect to be bad at it, like a beginner on a guitar), then expand as it works. Take whatever parts fit — sometimes just the six-week cycle, or just appetites-over-estimates. Singer adds that Basecamp’s own advantages (designers who code, no sales org, founder in every decision) are rare, so most teams must explicitly bridge the product–engineering wall during shaping rather than rely on organic integration.

Sources