Shape Up at Exiid
Fixed-time appetites for Evaluations and validation sprints: variable scope toward proof, circuit breakers, and kill over creep.
Shape Up is how Exiid scopes fixed-time bets: appetite instead of estimates, variable scope instead of open-ended backlogs. Evaluations and validation sprints are shaped work, not retainers.
Who this is for
For: partners expecting a clear box around time, cost, and deliverables.
Useful when: you need to know how scope expands (and when it stops).
What must be true
Every shaped bet has:
| Element | Exiid mapping |
|---|---|
| Appetite | 2 to 4 weeks (Evaluation) or 3 to 6 weeks (validation sprint) |
| Problem | Transfer thesis or bottleneck map |
| Solution sketch | Artifacts named upfront, not a full roadmap |
| Rabbit holes | Explicit out-of-scope list |
| Circuit breaker | Kill criteria written before work starts |
01 — Appetite, not estimate
We do not sell open-ended hours. We sell a time box with named outputs:
- Model Transfer Evaluation: go/no-go memo, proof metric sheet, risk scan
- Validation sprint: smoke or concierge result against thresholds
- Diagnostic (Build): friction map and scoped next step
If the appetite is wrong, we resize the bet or stop. We do not inflate the timeline quietly.
02 — Variable scope inside the box
Inside the appetite, scope flexes toward proof, not feature count:
- Build stage principle: smallest product that proves transferred behavior
- Kano discipline: must-haves ship; delighters wait for retention signal
- ICE/RICE ranks experiments inside the sprint backlog
03 — When Shape Up ends
Shape Up ends in ship, learn, or kill. Kill goes to the archive with the reason attached. Expand scope only when the numbers justify a new appetite.
Receipts
- Model Transfer Evaluation process
- Partner tracks: Diagnostic and sprint pricing bands
- Method stages: RECON before RAID
- Operating stack workflows