Best when
- Use this when generating many solution concepts quickly.
- Use this when early ideation after framing the problem.
- Use this when teams are converging too soon on one direction.
- Use this when you need fresh options under time pressure.
Method
Crazy 8s Ideation
A rapid sketching exercise where each participant creates eight concepts in eight minutes, followed by clustering and selection.
Choose this if
This method works when generating many solution concepts quickly.
- Choose this when generating many solution concepts quickly.
- Back out if participants cannot sketch or note quickly.
Pack = one session you can run now. Workspace = a saved thread for method choice, planning, and shareable outputs.
What Workspace helps you carry forward
- Brief the work
- Check the method fit
- Plan the session
- Prepare outputs you can share
Artifact preview
Idea set
Ideate · Low
Best when vs. Avoid when
Decide fit before you spend session time.
Avoid when
- Avoid this when participants cannot sketch or note quickly.
- Avoid this when the objective requires detailed architecture decisions.
Runbook
The core facilitation flow, with enough structure to run the method confidently in the room.
Standard setup for Mixed in remote delivery.
Total planned time: 45 min
Set problem prompt
What happens
Choose one focused challenge for ideation.
Facilitator does
Write the problem prompt on the board in a single sentence before starting the timer. Verify the room reads it the same way — ask one participant to paraphrase it. If the paraphrase drifts, rewrite before you start. A fuzzy prompt produces sketches that can't be compared.
Participants do
Read the problem prompt silently, then paraphrase it to yourself in one sentence. If your paraphrase is different from what's written, surface that difference now — before the timer starts. A prompt interpreted differently by each person produces sketches that can't be compared or clustered.
Output / signal of success
Idea set
Watch for
Prompt is too broad and participants sketch solutions to different problems.
Recovery tip
Stop the session. Re-read the prompt to the group. Ask one person to paraphrase it, then ask another. If the two paraphrases describe different problems, rewrite the prompt before the timer starts. The time investment in a clear prompt is recouped many times over in a coherent idea set.
Run eight-minute sketch round
What happens
Each person creates eight rough concepts.
Facilitator does
Start the timer, go silent, and do not respond to questions during the sprint. If someone gets stuck, whisper 'any direction is fine, keep moving.' Intervene only if a participant has frozen completely for more than ninety seconds.
Participants do
Sketch quickly and roughly — one idea per panel, eight panels in eight minutes. Don't evaluate while generating. The purpose is volume: get eight distinct directions on paper, not one polished concept. Stick figures and labels beat detailed drawings at this stage.
Output / signal of success
Top concepts shortlist
Watch for
Participants draw one large idea instead of eight distinct concepts.
Recovery tip
Stop the timer. Ask the participant: "Can you give me eight quick rough ideas — they don't have to be good." If they're still producing one large detailed idea after being restarted, let them continue with their concept and invite them to sketch variations in the remaining cells.
Share and cluster
What happens
Present concepts and group related ideas.
Facilitator does
Run a silent gallery first — everyone views all sketches before any explanation. Each person then gets sixty seconds to narrate their strongest idea. Cluster by concept type, not by person. Look for convergent ideas across multiple sketchers.
Participants do
Narrate your strongest idea in 60 seconds only — no setup, no context, just the concept. If the concept needs more than 60 seconds to explain, it's either too complex to communicate quickly or you're over-explaining. Limit yourself to what can be understood from the sketch plus one sentence.
Output / signal of success
Natural patterns are forming without long arguments, and the room can point to why items belong together.
Watch for
Sharing runs long as participants over-explain every sketch.
Recovery tip
If someone is over-explaining or defending their concept during the share round, stop them with: "We're just capturing the direction for now — we'll evaluate in the next step." Evaluation during sharing shuts down contributions from quieter participants before the full set is visible.
Select concepts
What happens
Vote on options to develop further.
Facilitator does
Run dot voting silently against the original problem prompt, not personal preference. After votes land, ask whether the leading concept actually addresses the original problem or just looks interesting. Open convergent clusters for a final heat-check.
Participants do
Vote for concepts that best address the original problem prompt — not the ones you like most aesthetically. After votes land, check your selections against the prompt: does the top concept actually solve the challenge you set, or just look compelling?
Output / signal of success
The room has a clear shortlist and can explain why those choices matter more than the alternatives.
Watch for
Selection reflects aesthetic preference rather than fit with the problem prompt.
Recovery tip
If selection is pulling toward aesthetics or familiarity rather than problem fit, re-read the prompt aloud and ask: "Of your votes, which one most directly solves this?" Remove any votes that can't be justified against the problem prompt before the final shortlist is called.
Prep and materials
Prepare the room, the evidence, and the working surface so the session can stay focused on synthesis.
Minimum viable evidence
A specific, agreed problem prompt written in a single sentence — ideally a How Might We prompt from a prior Problem Framing or LDJ session. Running Crazy 8s without a sharp prompt produces ideas that address different problems and can't be meaningfully compared or selected.
Pre-work
How Might We prompts
Optional
Timer
Optional
Dot-voting setup
Optional
Operational context note
Optional
Bring examples, pain points, or service evidence the room can point to quickly.
Room setup
Dot-voting setup
Optional
Remote facilitation choreography
Required
Pre-plan handoffs, breakout usage, and how people rejoin the main board.
Tools / materials
Template with eight frames
Required
Shared digital board with timer and voting
Required
Keep instructions and voting visible at all times.
Templates
Crazy 8s sheet
Run well / recover fast
Keep the output usable, then use the recovery guidance when the room starts drifting.
What strong output looks like
- Make the output legible to both sponsors and delivery owners.
- A well-run Crazy 8s session produces 8 ideas per participant across a range of directions — not 8 variations on one idea. The gallery shows concepts at different scales (small feature, different flow, completely different model) rather than incremental refinements. The voted shortlist can be traced directly to the problem prompt, not to visual appeal or presenter confidence. A weak session produces 1–2 detailed ideas per participant, a gallery that feels like design presentations rather than rough concept generation, and a shortlist that matches the ideas of the loudest participants.
What this does not produce
Crazy 8s produces raw concept material for further development — not validated directions or solution commitments. The ideas generated in an 8-minute sprint are starting points, not proposals. The selection step identifies which directions are worth developing into concept cards, prototypes, or storyboards, not which ones to build.
Red flags to correct before closing
- Over-analyzing during sketch round.
- Letting loud voices dominate selection.
- Re-state decision objective and expected artifact
- Timebox discussion and force owner assignment
- State what the artifact is confident about versus what still needs tailoring.
Failure mode
Participants over-analyse and refine one idea rather than generating eight distinct ones
Why it happens
Design and product participants are conditioned to make their work good before sharing it. Eight rough ideas in eight minutes violates that instinct. The familiar response is to spend the full eight minutes on one polished idea rather than generating eight rough ones. The single idea may be stronger than any one of the eight would have been, but the diversity of directions the session was designed to produce never materialises.
Recover now
If this happens, call the round. Accept the single concept, then immediately set a new timer for three minutes and ask for three more rough sketches in the remaining cells. The pressure is off because one idea already exists — this often unlocks faster, less precious generation.
Prevent next time
Next time, before the timer starts, say: "Cell 1 is your warm-up. Cells 2–8 are where the interesting ideas live. Don't refine cell 1." Give explicit permission to have bad ideas in cells 2–8.
Failure mode
Selection reflects aesthetic preference rather than fit with the problem prompt
Why it happens
After 8 minutes of sketching, concepts that are visually clean or narratively compelling get more votes than concepts that actually address the problem prompt. The voting step becomes a design critique rather than a strategic selection, and the shortlist reflects what looks good rather than what fits the challenge.
Recover now
If this happens, before voting opens, read the problem prompt aloud. Ask participants to review their votes with the prompt visible. If any highly-voted concept doesn't address the prompt directly, ask its supporters to make the connection explicit.
Prevent next time
Next time, write the voting criterion on the board before voting opens: "Vote for the concept that most directly addresses [the problem prompt]" — not "vote for the concept you like best."
Failure mode
Loud voices dominate the clustering and selection conversation
Why it happens
After silent individual work, the shift to group discussion disadvantages quieter participants. In clustering, the most confident voices assert groupings before others have processed the full board. In selection, the first expressed preference anchors the conversation and others adjust toward it. The output reflects the loudest voices rather than the distributed signal.
Recover now
If this happens, enforce silent dot voting before any verbal discussion. The vote distribution is visible — discuss only the most divergent placements rather than reopening selection broadly.
Prevent next time
Next time, use silent gallery viewing before any discussion begins. Each participant reviews the full board privately before any clustering or conversation starts. Opinions form independently before the group dynamic can influence them.
Adjust for your room
Change the shape of the session only after the fit and runbook are clear.
Open adjustment controlsStandard setup for Mixed in remote delivery.
Standard variant for Mixed in remote delivery. Waypoint is adapting timing, facilitation emphasis, prep, and recovery guidance here, but the room still needs final tailoring to the exact stakes and participants.
What to do next
Use this after you have checked the fit and know you want to carry the method forward.
The pack is the fastest way to run this now. Use the links below when you need prep, sharing, or a saved reference point.
References
Sources and references used in this method page.
- GV Sprint - Crazy 8s
Knapp, Zeratsky, Kowitz. Sprint.
- GV Sprint Resources
Sprint ideation and concepting patterns from GV.
Reviewed 2026-02-27 by Discovery Waypoint Editorial Team
Related playbooks
Known method, unclear next move?
These pages translate recognizable external methods into Waypoint-native guidance so the team can borrow the useful part without importing the whole mythology.
