Master the Crazy 8s sketching technique from Google's Design Sprint. Learn step-by-step facilitation, variations, and how rapid sketching produces better design ideas.
Crazy 8s is one of the most effective ideation exercises ever developed for product design. Originally created as part of the Google Ventures Design Sprint, it forces participants to sketch eight distinct ideas in eight minutes. The time pressure is the point: it eliminates perfectionism, bypasses self-censorship, and pushes past obvious solutions into genuinely creative territory.
Unlike verbal brainstorming techniques, Crazy 8s produces visual, concrete concepts rather than abstract descriptions. A rough sketch communicates spatial relationships, user flows, and interface concepts in ways that words cannot. Even "bad" drawings are useful because they make ideas tangible enough to discuss, combine, and build upon.
When people describe ideas verbally, everyone in the room forms a different mental image. "Let's put a big search bar at the top" sounds specific, but each listener imagines a different size, position, style, and surrounding context. Sketching removes this ambiguity. Even a rough box-and-arrow drawing creates a shared reference point that the group can react to, critique, and iterate on.
Sketching also reveals complexity that verbal descriptions hide. It is easy to say "the user just clicks the button and it works." Sketching the actual screen forces you to consider: where is the button? What information does the user need to see first? What happens after they click? What if there is an error? The physical act of drawing surfaces these questions naturally.
Most importantly, sketching is democratic. You do not need to be a designer or artist. Crazy 8s explicitly uses rough, low-fidelity sketches. Boxes, arrows, stick figures, and labels are all you need. The goal is communication, not aesthetics.
Each participant needs one sheet of A4 or letter-size paper and a thick marker (Sharpies work well because they prevent tiny, detailed drawing that wastes time). Fold the paper in half three times to create eight equal panels. Set a visible timer that everyone can see.
Before starting, clearly state the How Might We question or design challenge that the sketches should address. Write it where everyone can see it. For example: "How might we help new users complete their first project within 10 minutes?" The frame should be specific enough to generate focused ideas but broad enough to allow creative solutions.
Start the timer. Each person sketches one idea per panel, spending roughly one minute on each. No talking during sketching. The facilitator can call out the time at each minute mark to keep everyone on pace. If someone finishes a panel early, they move to the next one. If they are stuck, they can sketch a variation of a previous idea; variations count as separate ideas.
The first two or three panels typically produce obvious, safe ideas. That is expected. The magic happens in panels four through eight, when the obvious solutions are exhausted and the brain has to reach further. Encourage participants to push through the discomfort of not having "good" ideas; the constraint is doing the creative work for them.
After sketching, each person presents their eight sketches to the group. The presenter explains each panel briefly (15 to 20 seconds per panel). The group listens without critiquing. This is a "gallery walk" phase, not a debate. Participants can take notes on ideas they find interesting but should not voice opinions yet.
Use dot voting to identify the most promising sketches. Give each participant 2 to 3 dot stickers. They silently place dots on the specific panels (across all participants' sheets) that they find most compelling. The top-voted panels become the starting point for the next phase of work: more detailed sketching, storyboarding, or prototyping.
Thick markers physically prevent detailed drawing. This is intentional. When people use fine-tipped pens, they instinctively try to draw detailed UI mockups, which takes too long and triggers perfectionism. A Sharpie forces large, bold, abstract sketches that communicate concepts without getting bogged down in visual design details.
Any conversation during the sketching phase breaks concentration and introduces anchoring. If someone asks "what should I draw?", the answer is "anything that addresses the challenge statement." The ambiguity is productive. Different interpretations of the same challenge produce diverse ideas, which is exactly the point.
If participants are self-conscious about their drawing skills, the facilitator should sketch a deliberately rough example first. Show that boxes, arrows, and labels are all that is needed. Demonstrate that a stick figure and a labeled rectangle ("DASHBOARD") communicates a concept just fine. The bar is communication, not craftsmanship.
One round of Crazy 8s produces breadth. A second round produces depth. After the first round and dot voting, have participants take the top-voted concepts and do a second round that explores variations and refinements of those specific ideas. The second round typically produces more practical, detailed solutions because participants have already cleared the obvious ideas out of their system.
If eight panels in eight minutes feels too intense, fold the paper into four panels and give four minutes. This gentler pace works well for teams that are new to sketching exercises or for problems that require more spatial detail in each panel.
After Crazy 8s produces a winning concept, give participants 10 to 15 minutes to create a single, more detailed sketch of their best idea. This "solution sketch" can span one to three panels and include annotations, user flow arrows, and key screen states. It bridges the gap between the rough Crazy 8s panels and a proper prototype.
For distributed teams, use digital whiteboard tools like Miro or FigJam. Create a template with eight boxes per participant. Use the tool's built-in timer. The digital version loses some of the tactile energy of paper-and-marker, but gains the ability to share, annotate, and vote on sketches without physical proximity. Turn off cursors during the sketching phase so participants cannot see what others are drawing.
Crazy 8s is most commonly used at the transition between the Ideate and Prototype stages. By this point, you have a well-defined problem statement from the Define stage, user insights from the Empathize stage, and possibly a set of initial ideas from other brainstorming techniques. Crazy 8s takes those ideas and makes them concrete and visual.
It can also be used earlier in the process. During the Empathize stage, you can run Crazy 8s to sketch possible user journeys or pain point scenarios. During Define, you can sketch alternative framings of the problem space. The technique is versatile enough to produce visual thinking at any stage where abstractions need to become concrete.
Do not let people sketch on their laptops or tablets for the initial round. The temptation to use design tools creates perfectionism and slows everything down. Paper-and-marker is faster, more tactile, and more forgiving of rough ideas.
Do not skip the voting phase. Without explicit prioritization, the group defaults to discussing whichever idea is presented last or whichever idea belongs to the most senior person. Dot voting ensures that every participant's judgment carries equal weight.
Do not combine Crazy 8s with verbal brainstorming in the same time block. They serve different purposes and require different mental modes. Run them as separate exercises with a break in between.
The output of Crazy 8s is not a finished design; it is raw material for the next stage. Top-voted sketches become the basis for rapid prototypes, whether those are paper prototypes, clickable wireframes, or functional MVPs. The sketches provide the conceptual direction; the prototype adds enough fidelity to test with real users.
Keep the original sketches. They serve as a visual record of the team's creative process and are invaluable for retrospectives, stakeholder presentations, and future iterations when you need to revisit ideas that were not pursued in the current cycle.
Related guides: storyboarding techniques · value proposition canvas · assumption mapping
Design Thinker Labs Home · All Guides · How It Works · Pricing