Design Sprint

Use this when the team needs to go from problem to tested prototype in a compressed timeframe. Produces a facilitation plan for a design sprint (1-5 days depending on scope) with exercises, templates, and decision frameworks for each phase. Based on the GV design sprint methodology, adapted for modern product teams.

Framework attribution: Built on the GV design sprint (Jake Knapp). Kate runs sprints inside the Agile UCD discovery/framing arc she coached at Pivotal Labs; see knowledge/agile-ucd-reference.md.

Related skills: Use /facilitation-guide for general facilitation techniques. Generate prototypes with /artium-prototype or /client-prototype. Analyze test results with /usability-findings.

Process

Step 1: Gather inputs

Ask the user to provide:

1. Sprint challenge -- what problem are you trying to solve? What's the big question?
2. Duration -- full 5-day sprint, compressed 3-day, or 1-day "lite" version?
3. Participants -- who's in the room? (Roles, count, decision-maker identified?)
4. Target user -- who will test the prototype? (Persona or segment)
5. Constraints -- technical, business, or timeline constraints the team must respect
6. Existing research -- any prior interviews, analytics, or insights to inform the sprint
7. Logistics -- in-person or remote? What tools does the team use? (Miro, FigJam, physical supplies)

Step 2: Generate the sprint plan

Based on duration, produce a detailed facilitation plan:

5-day sprint plan

## Design Sprint Plan -- {{challenge name}}

### Sprint question
(The specific question this sprint will answer)

### Day 1: Understand & Define
**Goal:** Align on the problem and pick a target.

| Time | Exercise | Purpose | Facilitator notes |
|------|----------|---------|------------------|
| 9:00 | Sprint overview & rules | Set expectations | No devices, no HiPPO, vote not debate |
| 9:30 | Lightning talks (15 min each) | Share what we know | Business, tech, user, data perspectives |
| 10:30 | How Might We notes | Capture opportunities | Write HMWs on stickies during talks |
| 11:00 | HMW clustering & voting | Prioritize opportunities | Dot vote, top themes go on map |
| 11:30 | Journey map (current state) | Shared understanding of user flow | Rough, not polished |
| 1:00 | Sprint target selection | Pick the target moment | Decision-maker picks, team commits |
| 2:00 | Expert interviews (if applicable) | Fill knowledge gaps | 15-20 min each |

### Day 2: Sketch
**Goal:** Generate diverse solutions individually.

| Time | Exercise | Purpose | Facilitator notes |
|------|----------|---------|------------------|
| 9:00 | Lightning demos | Inspiration from existing solutions | 3-min demos of comparable products |
| 10:00 | Sketch warm-up: Crazy 8s | Rapid ideation (8 ideas in 8 minutes) | Quantity over quality |
| 10:30 | Solution sketch (detailed) | Each person's best idea, fully rendered | 3-panel storyboard, work alone |
| 1:00 | Continued sketching | Refine solution sketches | Anonymous -- no names on sketches |

### Day 3: Decide
**Goal:** Choose which solution to prototype.

| Time | Exercise | Purpose | Facilitator notes |
|------|----------|---------|------------------|
| 9:00 | Art museum | Silent review of all sketches | Sticky dot questions and highlights |
| 9:30 | Speed critique | 3 min per sketch, facilitator narrates | Sketch author stays silent |
| 10:30 | Straw poll | Everyone votes for favorites | Dot voting, no discussion yet |
| 11:00 | Supervote | Decision-maker picks the winner(s) | Can combine elements from multiple sketches |
| 11:30 | Storyboard | Map the prototype flow screen by screen | 10-15 frames covering the test scenario |

### Day 4: Prototype
**Goal:** Build a realistic-enough prototype to test.

| Time | Exercise | Purpose | Facilitator notes |
|------|----------|---------|------------------|
| 9:00 | Assign roles | Divide work | Makers, writers, asset collectors, stitcher |
| 9:30 | Build | Create prototype screens | Figma, Keynote, HTML -- whatever's fastest |
| 3:00 | Stitch & review | Assemble screens into clickable flow | Test internally, fix gaps |
| 4:00 | Write test script | Prepare interview guide for Day 5 | Tasks, not questions |

### Day 5: Test
**Goal:** Watch 5 users interact with the prototype.

| Time | Exercise | Purpose | Facilitator notes |
|------|----------|---------|------------------|
| 9:00-12:00 | User interviews (5 sessions, 45 min each) | Test the prototype | Interviewer + note-takers in a watch room |
| 1:00 | Debrief | Capture patterns | Each person lists top 3 observations |
| 2:00 | Pattern mapping | Cluster observations by screen/task | Green (worked), red (failed), yellow (unclear) |
| 3:00 | Decision time | What did we learn? What's next? | Go / iterate / pivot |

3-day compressed plan

(Combine Day 1 + 2 into one day, Day 3 morning + Day 4 into one day, Day 5 stays)

1-day lite plan

(Morning: problem + sketch, Afternoon: decide + quick prototype, Next day: test 3 users)

Step 3: Customize for context

Adjust the plan based on:

• Remote teams: Add Miro/FigJam board templates, adjust timing for screen fatigue (more breaks, shorter exercises)
• Existing research: Skip lightning talks if the team already has shared understanding; use the time for deeper sketching
• Technical constraints: Add a "feasibility check" exercise on Day 3 before the supervote
• Large groups (8+): Split into small groups for sketching, reconvene for voting
• AI-assisted sprint: Add AI where it has the least cultural friction and the highest payoff first: preparation (synthesizing prior research into the lightning talks) and post-sprint synthesis (clustering test observations on Day 5). Add AI to prototyping next (see Day 4). Save AI-assisted ideation for last, once the team trusts the tool, so it augments the sketches rather than replacing the thinking.
• AI-accelerated prototyping (Day 4): Turn solution sketches and storyboard frames into clickable mockups with an AI prototyping tool instead of hand-building every screen. This buys back hours on the tightest day and lets the team test higher fidelity. The storyboard is the spec, the AI renders it, the team fixes what feels wrong.
• Async pre/post-work: For distributed teams, move lightning talks and Day 1 research to async pre-reads, and run the Day 5 debrief as async observation logging that the team clusters together live. Reserve synchronous time for sketch critique, the supervote, and testing.
• Liberating Structures for decisions: Swap open debate for structured participation at the decision points. Use 1-2-4-All to converge on the sprint target, and Roman voting (thumbs up/down/sideways) or 25/10 Crowd Sourcing alongside the straw poll so the quiet experts weigh in before the supervote.

Step 4: Prepare materials

## Materials Checklist

### Physical (in-person)
- [ ] Large sticky notes (4x6) for solution sketches
- [ ] Small stickies for HMWs
- [ ] Dot stickers (voting)
- [ ] Markers (thick Sharpies -- forces big thinking)
- [ ] Timer
- [ ] Whiteboard or large paper for journey map

### Digital (remote)
- [ ] Miro/FigJam board with exercise templates
- [ ] Video conferencing with breakout rooms
- [ ] Shared timer (visible to all)
- [ ] Prototype tool (Figma, Keynote, or HTML scaffold)

Step 5: Review

Ask the user:

• Does the timeline work for your team's availability?
• Is the decision-maker identified and committed to attending?
• Do you have access to 5 target users for testing?
• Should I generate the Miro/FigJam board templates?

Output location

Present the sprint plan as formatted text in the conversation. For team distribution, export to Notion or create a Miro/FigJam board with exercise templates.