37signals Shape Up
Borrow an operating slice · Primary source
Treat Shape Up as an operating system for scoping and delivery, not as a workshop you can casually run in a single afternoon.
Best opened when
Teams that repeatedly start work too vague and let scope sprawl silently.
What to steal first
Appetite-driven scoping and shaping before commitment.
What not to copy
It is not a discovery workshop method.
Closest move
Core mechanism
Shape Up is a full product operating system built around three interlocking ideas: shaping work before committing to it, betting on that shaped work at a deliberate table, and running bounded six-week cycles where the team owns the execution. It exists because 37signals believed that most delivery problems are actually scoping and commitment problems — work that starts too vague, expands mid-cycle, and hides uncertainty behind percent-complete fiction.
The part that transfers most cleanly is appetite-driven scoping. Appetite means treating the time available as a fixed input that constrains the scope, rather than treating scope as fixed and negotiating time after the work starts. A team that can hold a six-week appetite and negotiate scope against it — without adding back the vague features that were removed during shaping — has captured most of the leverage.
What does not transfer easily: the full model requires real authority to say no mid-cycle, which most organizations cannot support without structural change. The betting table works only when the people at the table have genuine authority to un-commit from low-value work. The hill chart framing is only useful if the team can tolerate acknowledging uncertainty publicly. Adopt the parts that your operating context can actually support.
Adopt the minimum useful slice
What to adopt
The pitch document and the appetite — the two elements that do the most work before delivery begins. The pitch defines the problem and a rough solution concept. The appetite defines how much time the problem is worth before solution design begins.
How to start
Before your next significant piece of work is scoped, write a Shape Up pitch. Two pages maximum. Include: the problem (what is going wrong and for whom), the solution concept (rough direction, not a spec), the appetite (how long are we actually willing to spend — 2 weeks or 6?), and the rabbit holes (what are we explicitly not building in this cycle). Share it with decision-makers and bet on it explicitly rather than adding it to a backlog.
Signal it's working
The appetite conversation changes. Instead of 'how long will this take?', the question becomes 'how much time is this problem worth?' When scope creep is named and rejected using the appetite rather than discovered at the end of a cycle, the slice is working.
First failure: Running shaping and delivery in parallel. Shape Up only works when the people doing the shaping are not simultaneously trying to deliver. If the same people are writing pitches and building features, neither gets the attention it needs.
What good looks like
How it worked there
37signals built Shape Up at a small company where the same people who shaped the work also built it, and where leadership had full control over the betting table. The model was designed for a team that could make clean commitments and honor them — one where a six-week cycle meant a real six-week cycle with no scope creep.
That context does not transfer. Most product organizations have external stakeholders, governance requirements, and mid-cycle requests that make a clean betting table impossible. The organizations that have successfully borrowed Shape Up ideas — appetite framing, shaping packets, hill charts — have done so by translating the intent into their own operating constraints rather than by copying the full system. The full system is coherent at Basecamp's scale and culture; it is incoherent in organizations with separated product and engineering authorities, variable team ownership, and quarterly OKR commitments.
What not to copy · Failure modes
Teams import the Shape Up vocabulary while keeping the same vague scoping habits underneath. They run 'shaping sessions' that produce loosely defined work packets. They call phases 'cycles' without changing what gets committed or who controls scope mid-cycle. The labels create the impression of operating change without the substance. Shape Up is a set of operating assumptions — about how decisions are made, who has authority to say no, and how progress is read — not a vocabulary kit.
The appetite limit is treated as an estimate, not a constraint. The whole leverage of Shape Up comes from treating time as fixed and scope as what adjusts. Organizations that adopt Shape Up while still treating time as soft and scope as negotiable mid-cycle have not adopted anything. They have decorated their existing planning habit with new words. The betting table only works if the organization can actually honor the cycle boundary.
It is used as a project management replacement rather than an operating model decision. Teams sometimes adopt Shape Up to fix a planning tool problem — frustrated with their backlog or their sprint rituals. But Shape Up is not a Jira replacement. It is a question about how your organization decides what to build, who shapes it, and how progress is understood. The planning tool question follows from those decisions; it does not drive them.
Weak signals to watch for
Closest Waypoint move
Primary route
Use this when the work still lacks a strong enough frame to be shaped responsibly.
Use this when you need a scoped, shaped problem ready for delivery and can accept skipping the broader strategic framing the pitch assumes you have already done.
technique
Useful when the team needs a simpler commitment conversation before changing its whole delivery model.
Sources and confidence
Primary source
37signals
37signals
Discovery Waypoint
Reviewed by Discovery Waypoint Editorial Team · 2026-04-04