Shape Up Pitch

Shape Up Pitch

Create shaped pitches that fix time (appetite) and adjust scope to fit, rather than estimating how long a fixed scope will take.

Core Principle: Appetite Over Estimates

Wrong question: "How long will this take?" Right question: "How much time do we want to spend on this?"

Appetite is typically 2 or 6 weeks - long enough to build something meaningful, short enough to see the end from the beginning.

Shaping Workflow

1. Set the appetite - How much time is this worth?
2. Gather shapers - PM + Designer + Senior Engineer (knows where the bodies are buried)
3. Define the approach - High-level solution at the right abstraction
4. Identify rabbit holes - Technical risks that could blow up the timeline
5. De-risk - Resolve open questions BEFORE handing off
6. Create artifacts - Breadboards, fat marker sketches, rough architecture
7. Write the pitch - Package for the betting table

The Shaping Session

Who Must Be Present

The engineer is NOT there to code. They're there to identify if the proposed UI implies backend work that doesn't exist or would exceed the appetite.

Shaping Dynamics

• Try to break ideas - If an approach is too complex, don't extend time; change the approach
• Trade actively - "We could do A or B within appetite, not both. Which matters more?"
• Name the rabbit holes - "If we build that onboarding step, it actually branches into three bank integrations"

Artifacts

Breadboards

Show flow and logic WITHOUT visual design. A breadboard contains:

• Places (screens, dialogs, menus)
• Affordances (buttons, fields, links)
• Connection lines (what leads where)

[Appointments Screen]
    |
    ├── [2-month dot grid] → shows availability at a glance
    |       └── tap day → [Day Detail]
    |
    └── [Agenda list] → scrollable list of booked slots
            └── tap slot → [Appointment Detail]

Fat Marker Sketches

Rough UI sketches drawn as if using a fat marker - impossible to add detail. Shows:

• Key screen layout
• Primary action placement
• Information hierarchy

NOT wireframes. NOT Figma files. The lack of fidelity is intentional.

Rough Architecture

High-level view of components:

Components:
- 2-month dot grid view (client-side render)
- Sliding agenda view (lazy-load past 7 days)
- Creation button (triggers existing booking flow)

Data:
- Reads from existing appointments API
- No new endpoints needed

Dependencies:
- Existing booking service (confirmed available)

Pitch Structure

# [Project Name]

## Problem
[Paste the Framed Problem Statement from problem-framing]

## Appetite
[2 weeks / 6 weeks] - [Brief justification for why this budget]

## Solution
[2-3 paragraphs describing the approach at high level]

## Breadboards
[Include breadboard diagrams showing flow]

## Fat Marker Sketches
[Include rough sketches if helpful]

## Rabbit Holes
[List technical risks identified and how they're addressed]

| Rabbit Hole | Risk | Mitigation |
|-------------|------|------------|
| [Risk 1] | [What could blow up] | [How we've de-risked it] |

## No-Gos
[What's explicitly OUT of scope - prevents scope creep]

## Nice-to-Haves
[Things that could be cut if time runs short - the "dials" the team can turn]

The Handoff

Give the team "the whole idea," NOT a backlog of tickets.

The team is responsible for:

• Breaking down into tasks (often using "Nine Boxes" - roughly 9 implementation chunks)
• Deciding implementation order
• Making tactical decisions within the shaped boundaries

Circuit Breaker

If the project isn't finished at the end of the appetite:

• Do NOT extend the timeline
• Cancel it OR pull it back to the shaping table
• This prevents the sunk cost fallacy

Quality Checklist

Before pitching:

• Appetite is explicit (not "as long as it takes")
• Senior engineer validated feasibility
• Rabbit holes identified and addressed
• Artifacts show logic, not just UI
• No-gos explicitly stated
• Team can understand "the whole idea" from the pitch alone

Anti-Patterns

For detailed failure modes and how to avoid them, see references/anti-patterns.md.

Examples

For breadboard and pitch examples, see references/pitch-examples.md.