Introduction | Shape Up — Ryan Singer

Summary

Shape Up is Basecamp's product development methodology built around fixed time, variable scope. Teams work in six-week cycles, with work "shaped" by a senior group before being handed off. The key inversion: instead of estimating how long work will take, you decide how much time an idea is worth (appetite), then shape the solution to fit. Teams get full autonomy — no ticket-taking — and a hard circuit breaker kills projects that don't ship in one cycle.

The mental model: appetite over estimates, shaping over speccing, betting over backlogging.

Key Concepts

• Appetite — time you want to spend, not time you estimate. Frames every decision.
• Shaping — senior group defines key elements of a solution at the right abstraction level: concrete enough to act on, abstract enough to leave room for the team.
• Circuit breaker — projects that don't ship in one cycle get cancelled by default, not extended. Prevents runaway investment.
• Hill chart — tracks work on a spectrum from unknown (uphill) to known (downhill) to done.
• Scope hammering — forcefully cutting scope to fit inside the fixed time box.
• Cool-down — two-week break between cycles for ad-hoc work and the betting table.
• Small batch vs. big batch — a cycle can be one large project or several 1-2 week projects.

Relevance

This maps directly to how we should run [[01-projects/squarely-puzzles/index]] development cycles. The "appetite" framing is a better fit for a small team than sprint estimation. The circuit breaker discipline prevents the "projects go on and on" trap the article names.

Shape Up's shaping phase also parallels the [[06-reference/concepts/skills-as-building-blocks]] idea — you compose known capabilities into a solution outline before committing. And the insistence on shipping over discovery echoes [[06-reference/2026-03-31-block-hierarchy-to-intelligence]] — bias toward output, not process.

Open Questions

• What's the right cycle length for a solo-founder + AI-COO team? Six weeks may be too long — could we run two-week cycles with one-week cool-downs?
• How does shaping work when the "senior group" is one person? Can the AI COO do shaping work?
• Should we adopt the betting table format for deciding what Squarely ships next?