Shape Up project management for solo developers.
Short answer: Shape Up applies cleanly to solo work once you remove the pieces that assumed a team and replace them with rituals you can run alone. The four phases — pitch, betting table, cycle, cooldown — keep their shape. The ceremonies don't.
This guide is the canonical walkthrough of how to do that — what to keep from Shape Up (Ryan Singer / Basecamp), what to swap, and what to do instead. Read top to bottom or jump to the FAQ if you have a specific question.
The 60-second version
- Cycles — a 2-week default fixed time-box. Defend it.
- Pitches — written before you build. Problem, appetite (T-shirt size), rough solution, rabbit holes, no-gos. Even when you're the only reader.
- Solo betting table — 30 minutes with yourself at the start of cooldown. Cut what you can't defend.
- Hill chart — uphill (figuring it out) → downhill (executing). Replaces percent-done.
- Cooldown — 2–4 days between cycles. Cleanup, retro, capture ideas. Genuinely rest.
- Cut scope, not time — if a scope can't ship inside the appetite, ship a smaller version of the scope.
Why team Shape Up doesn't extend to solo work as-is
Shape Up was written for product teams at Basecamp — designers and developers organized around 6-week cycles, with a betting table run by the founders. Most of the rituals carry weight only because there's a team enforcing them. Try to run them alone, verbatim, and they collapse.
Three pieces are specifically team-shaped:
- The betting table assumes a group of people argue pitches and pick which ones to fund. Solo, you're arguing with yourself, and you'll always lose to enthusiasm. Replace with a written commit checklist plus a self-administered "would I bet money on this?" question.
- Pitch review assumes peers push back on weak pitches before commit. Solo, the pushback has to come from somewhere — an AI reviewer, a written rubric, or a trusted reader who looks at pitches once a cycle.
- Cycle/cooldown rotation assumes some people are working while others are pitching. Solo, you do both phases in sequence, which means cooldown has to be short and rest-heavy or it eats the next cycle.
Everything else — pitches, hill charts, fixed cycles, T-shirt sizes, no-gos, the appetite/scope-hammer relationship — transfers verbatim, and is more useful solo than in a team because there's nobody else to catch your scope drift.
The solo developer's five phases
The phases below extend Shape Up's four (Pitch, Bet, Cycle, Cooldown) with an explicit Think phase up front. Solo developers tend to skip pitching and start coding the moment an idea pops; Think is the parking lot for ideas before they're ready to pitch.
Think — collect ideas before you commit to shaping any of them.
An idea you can't link to anything else isn't ready to pitch. Keep a freeform notebook (markdown notes, wikilinks, tags) where ideas land and accumulate context. Only promote an idea to a pitch when it's connected enough that you can describe its problem in one paragraph.
Pitch — write the brief before you write the code.
Five fields: Problem (what's broken), Appetite (S/M/L/XL — a time budget, not an estimate), Solution (rough sketch), Rabbit Holes (what you'll consciously not solve), No-Gos (explicitly out of scope). Solo pitches can be 200 words; the structure stops you from re-discovering the same blockers at week 3.
Commit — your one-person betting table.
30-minute session with yourself at the start of every cooldown. Read every ready pitch out loud. Ask: would I bet two weeks of my time on this? Cut anything you can't defend. Pick at most two pitches for a two-week cycle.
Build — the cycle is sacred. The hill chart keeps you honest.
A scope at position 30 (still figuring it out) at the start of week 2 of a 2-week cycle is a problem signal, not a status update. Cut surface area. Log unplanned work the moment it interrupts you. Don't extend the cycle when scope grows; cut scope when time runs short.
Cooldown — short, rest-heavy, retrospective-bearing.
2–4 days between cycles. Spend the first day actually offline if you can. Then: small cleanup work (the bug fixes you couldn't justify mid-cycle), retro on the cycle (what shipped, what got cut, where appetite was wrong, what surprised you), and capture any cooldown ideas that came up. Write 2–4 short lessons.
Team Shape Up vs. solo Shape Up — what actually changes
| Element | Team Shape Up | Solo Shape Up |
|---|---|---|
| Default cycle length | 6 weeks | 2 weeks (4 max for L pitches) |
| Cooldown | 1 week, parallel to next cycle | 2–4 days, sequential |
| Betting table | Founders + leads, group decision | 30-min self-check at start of cooldown |
| Pitch review | Peer review before bet | AI reviewer or written rubric |
| Pitches per cycle | 2–4 | 1–2 |
| Estimating | T-shirt sizes (appetite) | T-shirt sizes (appetite) — same |
| Hill chart | Per scope, weekly check-in | Per scope, daily check-in (faster feedback) |
| Unplanned work | Logged centrally, owned by lead | Logged honestly; auto-classified from commits |
| Retrospective | Team meeting | Self-review or AI-guided conversation |
| What if a scope slips? | Cut scope or cut the cycle | Cut scope. Always. Re-pitch smaller next cycle. |
Common pitfalls solo developers hit when adopting Shape Up
- Skipping the pitch. The most common failure mode. "It's just me, I'll keep it in my head" — and then 8 days into a cycle, the cycle is doing something different from what you started. Always write the pitch, even if it's 6 sentences.
- Six-week cycles. Inherited from the book and almost always wrong for solo work. Two weeks is the sweet spot; four weeks for genuine L pitches where the work is naturally sequential. Six weeks is too long to defend solo.
- Skipping cooldown. Skipping cooldown to start the next cycle "while the energy's there" makes the next pitch worse — there's no retro to learn from, no rest, and any cleanup debt rolls forward.
- Hiding interruptions. Pretending the unplanned hours weren't there breaks the appetite calibration. After 3–4 cycles you should see your real interruption percentage. If you're hiding 30% of your hours, every future cycle is shaped on a lie.
- Extending cycles to make scope fit. The whole point of Shape Up is fixed time, variable scope. Extending the cycle to ship the original scope is just project management with extra ceremony. Cut.
- Treating the hill chart as a percent-done. 50% isn't halfway done — it's "you've finished figuring it out, now you're executing." A scope that hits 50 fast is a green signal; a scope that lingers in the 30s is a cut signal.
How to actually start tomorrow
- Pick one feature you've been wanting to build. Write a one-paragraph problem statement and a T-shirt size (don't think too hard — S=3 days, M=1–2 weeks, L=3–4 weeks, XL=full cycle).
- Write a rough solution sketch. Two paragraphs is enough.
- Write rabbit holes (what you'll consciously not solve) and no-gos (out of scope).
- Pick a 2-week cycle starting Monday. Don't pick more than the appetite supports.
- Each day, mark scope position on the hill (0–100). If a scope doesn't move for 3 days, that's a cut signal.
- End of cycle: hold a retro on yourself. What shipped? What got cut? What surprised you? Write 2–3 lessons.
- Take 2 days off the project. Then commit the next cycle.
For a step-by-step walkthrough that includes wiring a GitHub repo so commits auto-link to scopes, see the tutorial. For the in-app vocabulary and definitions, see the methodology page.
Frequently asked questions
Does Shape Up actually work for one person?
▼
Yes — but you have to remove the team-shaped pieces and replace them. Pitches still work (they force you to write the problem before the solution). Hill charts still work (uphill = figuring it out, downhill = executing). Cycles still work (a fixed time-box you defend). What doesn't transfer: the betting table (no team to bet against), the kickoff meeting (you're already there), and the rotation between cycle and cooldown (you can't 'switch teams' if you're the only one). Replace those with a personal commit ritual and a real-rest cooldown.
How long should a solo Shape Up cycle be?
▼
Two weeks is a sensible default for solo work. Six (Basecamp's default) is too long — you'll lose track of what you committed to and the project will drift. The point of a cycle is to make a small bet you can honor; shorter cycles mean smaller bets, which is what a single developer should be making. 4 weeks works for L pitches; 6 weeks only for XL pitches you can't decompose.
Do I need pitches if I'm the only person reading them?
▼
Yes. The pitch isn't for an audience — it's for your future self three days into the work. The act of writing problem / appetite / rough solution / rabbit holes / no-gos forces you to surface assumptions you'd otherwise discover at week 3. A solo pitch can be 200 words; the structure matters more than the length.
What replaces the betting table when you're solo?
▼
A 30-minute commit session with yourself at the start of cooldown, before the next cycle. Open every ready pitch, ask: "Is this still worth N weeks?" Cut anything you can't defend. Then pick the bets — typically one or two for a 2-week cycle. Do this even if it feels silly; the ritual is what stops scope from drifting.
How do you do the cooldown phase as a solo developer?
▼
Cooldown serves three jobs: rest, cleanup, and reflection. None require a team. Spend a few days on small bug fixes, refactors, dependency updates — the maintenance work you couldn't justify mid-cycle. Capture loose ideas from the cycle into a brain-dump. Then run a retrospective on yourself: what shipped, what got cut, where the appetite was wrong, what surprised you. Write 2–4 short lessons. Solo cooldown is shorter than team cooldown — typically 2–4 days, not a week.
How do you handle interruptions if you're the only one fielding them?
▼
Log them as unplanned work and don't replan the cycle. Every interruption is a data point: did the cycle's appetite turn out to be wrong, or was the interruption manufactured? After 3–4 cycles you'll see your real interruption percentage and can size future bets honestly. Hiding interruptions makes the next cycle's appetite drift.
What if a scope doesn't fit in the cycle?
▼
Cut the scope, not the time. Shape Up's hill chart is a forecast tool, not a status meter — if a scope hasn't moved past 50 (still figuring it out) by the end of week 1 of a 2-week cycle, that's a strong signal to reduce its surface area. Worst case: the cycle ends, the unfinished scope reflows back to the pitch list, and you re-pitch a smaller version next cooldown. Don't extend the cycle to make scope fit.
Do I really need T-shirt sizes (S/M/L/XL) instead of estimates?
▼
Yes. A T-shirt size is an appetite — what you're willing to spend before you'd give up — not a prediction of how long the work takes. The distinction matters because it commits you to cutting scope when reality runs longer than appetite. An estimate commits you to extending the deadline. Solo work especially benefits: no one else is going to enforce the time-box for you.
Where does Shape Up clash with how solo developers actually work?
▼
Three places. (1) Solo developers usually have a backlog of small bugs and dependency upgrades that team Shape Up sweeps into a 'small batch' — alone, you have to actively defend cycle time from these. (2) Solo developers often work irregular hours, so the 6-week cycle stretches into 8 in real time. Shorter cycles fix this. (3) Solo developers don't have peer pitch review, which is what stops weak pitches from getting committed. An AI reviewer or a written review checklist can substitute.
Heliostat: Shape Up for one
The methodology above is what Heliostat implements end-to-end — pitches, hill charts, fixed cycles, an AI pitch reviewer (the Senior), a cooldown retro coach, and an MCP integration so external AI clients can act on your data.