Design Sprint Exercises: Sketching Solutions and Running Structured Voting
Teaches you how to facilitate Crazy 8s rapid ideation, detailed solution sketching, silent heat-dot voting, and the Decider's final call so your sprint team converges on the strongest prototype-ready concept without groupthink or design-by-committee.
Start with individual Crazy 8s sketching (eight variations in eight minutes) to generate volume, then have each participant create a detailed three-panel solution sketch anonymously. Display all sketches in an art museum walkthrough, let the group place heat-map dots on compelling elements silently, and finish with the Decider making a final call on which concepts to prototype. This sequence prevents groupthink while surfacing the strongest ideas through structured, silent evaluation.
Outcome: You produce a single winning solution sketch (or at most two competing concepts for a Rumble) that the entire team has evaluated fairly, ready to hand off to storyboarding and prototyping with clear confidence in the direction.
Prerequisites
- A completed sprint challenge and target map from Day 1 (see mapping-and-defining-sprint-challenges)
- Lightning demos or How Might We notes capturing existing inspiration
- A designated Decider with authority to make the final call
- Basic understanding of the Google Design Sprint five-day structure
- Paper, thick markers (Sharpies), and small round sticker dots for each participant
Overview
Sketching and voting form the creative engine of the Google Design Sprint. They occupy Day 2 (Sketch) and part of Day 3 (Decide), and their purpose is deceptively simple: generate many possible solutions individually, then converge on the best one without the loudest voice in the room steering the outcome. The exercises themselves, Crazy 8s, solution sketches, art museum, heat-dot voting, and the Decider vote, are specific design sprint exercises engineered to separate idea generation from idea evaluation. When run correctly, the sequence produces a concrete, prototype-ready concept backed by genuine team signal rather than consensus by exhaustion.
The core problem these exercises solve is convergence quality. Brainstorming sessions in most organizations devolve into a few dominant speakers pitching ideas while everyone else nods along or stays silent. The Google Design Sprint sidesteps this entirely by making every participant sketch alone, present anonymously, and vote silently before any discussion happens. The result is that a quiet backend engineer's concept gets the same evaluation as the VP's pet idea. The structured voting layer then compresses dozens of ideas into a clear signal about which elements resonated most, giving the Decider enough information to make a confident call.
The artifact you produce at the end of these exercises is a winning sketch, a detailed three-panel concept that captures a user's interaction with the proposed solution in enough detail that someone who was not in the room could understand the idea. If the vote produces two strong but incompatible directions, you may run a "Rumble" where both get prototyped and tested head-to-head. Either way, the output feeds directly into storyboarding and prototyping, giving those phases a clear starting point instead of a vague consensus.
Success looks like this: every participant, regardless of role or seniority, produced at least one detailed sketch. The voting surface clearly shows clustering around specific elements. The Decider made a call within minutes rather than hours. And the team leaves the room aligned on what to build next, not because they all agreed, but because the process gave every idea a fair shot and the strongest one emerged visibly.
How It Works
The sketch-and-vote sequence works because it separates four cognitive tasks that most teams mash together: divergent generation, detailed articulation, silent evaluation, and authoritative decision. Each task has its own exercise, its own rules, and its own output. Combining them into a single brainstorm session is what produces groupthink, anchoring bias, and the HIPPO effect (Highest Paid Person's Opinion). The design sprint exercises keep them apart on purpose.
Crazy 8s handles pure divergent generation. By forcing eight rough sketches in eight minutes, it pushes participants past their first (obvious) idea and into unexpected territory. The time pressure is the mechanism: you literally do not have time to self-censor or polish. Most people find their first two or three sketches are predictable, sketches four through six start getting weird, and sketches seven and eight sometimes contain the breakthrough concept precisely because the safe ideas have already been exhausted. The key assumption is that volume precedes quality in early ideation, which is well supported by design research.
Solution sketching shifts the task to detailed articulation. Each participant picks their most promising direction (often combining elements from multiple Crazy 8s frames) and creates a three-panel sketch with enough detail that a stranger could follow the concept. The sketch is anonymous: no names on the paper, no pitching allowed. This anonymity is critical because it forces the idea to stand on its own visual clarity rather than the persuasiveness of the presenter. The assumption here is that good ideas should be self-explanatory. If a concept requires a five-minute verbal explanation to make sense, it probably is not clear enough to prototype and test with real users.
The art museum and heat-dot voting phase handles silent evaluation. All sketches go up on a wall, and the team walks through them in silence, placing small sticker dots on any element (not just whole sketches) that they find compelling. This silent treatment prevents anchoring: nobody hears someone else's enthusiasm and adjusts their dots accordingly. After the silent round, the facilitator leads a brief "speed critique" where dots are tallied, the creator is revealed only after discussion, and the group captures standout ideas. The clusters of dots become a visual heat map of collective intelligence.
Finally, the Decider vote converts signal into action. The Decider, who was designated at the start of the sprint, reviews the heat map, listens to the speed critique summaries, and places a single large "super vote" sticker on the concept the team will prototype. This is not a democracy. The Decider bears organizational responsibility for the outcome and uses the team's signal as input, not as a binding referendum. The reason for a single authority is speed and accountability: committees negotiate, Deciders decide. The result feeds directly into storyboarding the user journey and ensures the team moves into prototyping with clarity.
The entire sequence within the Google Design Sprint is designed so that even if the Decider picks a concept that did not receive the most dots, the process still feels fair because every participant had an equal, silent, anonymous opportunity to contribute and evaluate. Legitimacy comes from the process, not the outcome.
Step-by-Step
Step 1: Set the Stage with Lightning Demos and Recap
Before anyone picks up a marker, spend 15-20 minutes reviewing the sprint challenge, the target map from Day 1, and the Lightning Demo notes the team collected. Post these artifacts on the wall where everyone can see them. Walk through the How Might We notes clustered on the map, reminding the group which part of the user journey they are solving for. This recap ensures every participant has the same context loaded into working memory before they start sketching.
Without it, people sketch solutions to different problems, and the voting phase becomes an apples-to-oranges comparison. Hand out thick markers (Sharpies work well because they prevent tiny, detailed drawing that slows people down) and letter-size paper to each participant.
Tip: If you are running this after a break or on a new day, do not skip the recap even if the team protests that they remember everything. The 15-minute investment prevents 45 minutes of misaligned sketches.
Step 2: Run Crazy 8s for Rapid Divergent Sketching
Each participant takes a single sheet of paper and folds it into eight panels (fold in half three times, then unfold). Set a visible timer for eight minutes. The rule is one rough sketch per panel, one minute per panel. Announce transitions at each minute mark.
Emphasize that these are rough concepts, not polished drawings. Stick figures and boxes are perfectly fine. The goal is quantity and variation: eight genuinely different approaches to the sprint challenge, not eight refinements of the same idea. Participants work in complete silence and do not look at each other's papers.
When the timer ends, each person has eight candidate directions, most of which they would never have explored without the time pressure.
Tip: Many participants freeze at panel one because they want to draw something 'good.' Tell them explicitly: 'Your first two sketches will be obvious. That is fine. Push past them. The interesting ideas start at sketch four or five.' This reframing dramatically reduces blank-page paralysis.
Step 3: Facilitate a Quiet Reflection Period
Give the group five to ten minutes of silent review time. Each person looks at their own eight panels and circles or stars the one or two concepts (or elements from different panels) they want to develop further. Encourage participants to think about which idea best addresses the sprint challenge, not which one looks prettiest. This reflection step is easy to skip but doing so means people rush into the detailed sketch without filtering, which leads to weaker solution sketches.
Remind the team that they can combine elements from multiple Crazy 8 panels into their solution sketch. The output of this step is a personal decision about which direction to develop, made privately before any social influence enters the picture.
Tip: If someone says all eight of their sketches are bad, reassure them that even a single interesting element from one panel is enough to build a solution sketch around. Perfection is not the standard here.
Step 4: Create Detailed Three-Panel Solution Sketches
Each participant now takes three sheets of paper (or one large sheet divided into three panels) and creates a detailed, self-explanatory solution sketch. Allow 30-45 minutes for this step. The three panels should show a sequence: how a user encounters the solution, how they interact with it, and what outcome they reach. Each panel needs a descriptive title at the top.
Participants should use annotations, callout text, and arrows to make the sketch comprehensible to someone who has never heard the idea explained verbally. Emphasize that the sketch must stand alone because there will be no pitch. Collect all sketches and do not put names on them. Stack them face-down so nobody sees anyone else's work during creation.
Tip: The biggest quality lever in this step is the 'no names, no pitching' rule. Remind participants: 'If your sketch needs a verbal explanation to make sense, add more annotations until it doesn't.' This forces clarity that directly translates to prototype quality later.
Step 5: Run the Art Museum Walkthrough
Tape all solution sketches to the wall in a row, spaced evenly. Give each participant a sheet of small, round sticker dots (20-30 dots per person works well). Set a timer for the silent review, usually 10-15 minutes depending on the number of sketches. Participants walk along the wall like an art gallery, reading each sketch silently and placing a dot on any specific element they find compelling.
They can place multiple dots on the same sketch if multiple elements resonate, or cluster dots on a particular panel or annotation. The only rule is silence. No pointing, no whispering, no eye contact that signals agreement. When the timer ends, the wall displays a visual heat map of collective interest that emerged without any verbal negotiation.
Tip: If you catch participants trying to peek at where others are placing dots, gently redirect them to a different part of the wall. Dot placement is contagious, and even a small amount of social signaling can create false clustering.
Step 6: Conduct the Speed Critique
Now the facilitator walks the group through each sketch in order, spending no more than three minutes per sketch. The structure for each sketch is: the facilitator narrates what they see (not the creator), the group calls out standout ideas, the facilitator captures these on sticky notes above the sketch, concerns or questions are noted briefly, and finally the creator is revealed and gets 60 seconds to clarify anything the group missed. The creator only speaks at the end, and only to clarify, not to sell. This sequence matters because hearing the facilitator's narration exposes whether the sketch is truly self-explanatory.
If the facilitator misreads the concept, it signals a clarity problem that would also affect real users during testing. The sticky notes above each sketch become a caption that enriches the heat-map data for the Decider.
Tip: As facilitator, practice narrating sketches in a neutral, descriptive tone. Say 'This panel shows a user receiving a notification' rather than 'This is a really clever notification idea.' Your tone shapes the room's evaluation even in a structured format.
Step 7: Execute the Decider Vote
Give the Decider three large, visually distinct sticker dots (a different color or size from the heat-dot stickers). The Decider reviews the heat map, re-reads the speed critique sticky notes, and places their super-vote dots on the concepts the team will move forward with. In most sprints, the Decider places all three dots on a single sketch, signaling a clear winner. If the Decider splits their dots across two sketches that represent fundamentally different approaches, the team enters a Rumble: both concepts get prototyped and tested head-to-head on Day 5.
The Decider does not need to justify their choice at length, but a brief sentence explaining their reasoning helps the team understand the direction. Once the super-vote lands, the decision is final. There is no negotiation, no appeals process, no 'but what about' discussions.
Tip: If the Decider is agonizing between two very different concepts, lean toward the Rumble. A head-to-head user test on Day 5 resolves the question with real data instead of opinion. But limit Rumbles to two concepts maximum, because prototyping three directions in one day is almost always too thin.
Step 8: Document the Winning Concept and Capture Learnings
Photograph every sketch on the wall at high resolution, including the dot clusters and sticky notes. For the winning sketch (and the runner-up if doing a Rumble), create a clean reference document that includes the three-panel sketch, the speed critique notes, the heat-map pattern, and the Decider's stated reasoning. This documentation becomes the primary input for storyboarding and is critical for anyone who joins the sprint mid-stream or needs to understand why this direction was chosen. Store photos in the team's shared workspace within 30 minutes of the session ending, because physical sticky notes fall off walls and marker drawings smudge.
Also capture any strong elements from losing sketches that the team might incorporate, these often surface during storyboarding as useful additions.
Tip: Label each photo with the sketch number from the wall, the creator's name (now that anonymity is no longer needed), and the dot count. This metadata saves enormous time when the storyboarding team needs to reference specific elements two hours later.
Examples
Example: B2B SaaS Onboarding Redesign (5-Person Team)
A project management SaaS company is running a design sprint to fix their trial-to-paid conversion. The team of five (product manager, two designers, an engineer, and a customer success lead) has defined the sprint challenge as 'How might we help new trial users complete their first project within 30 minutes?' The Decider is the product manager.
During Crazy 8s, the customer success lead's early panels show variations of guided wizards, but her sixth and seventh panels shift to a pre-populated sample project that users can modify rather than build from scratch. This concept was not on anyone's radar before the session. In the solution sketch phase, three of the five participants independently incorporate elements of pre-populated content, but with different interaction models: one shows a choose-your-template flow, another shows an AI-generated starter project, and the third shows a clone-from-gallery approach. During the art museum, dots cluster heavily on the AI-generated starter project's second panel, which shows the system asking three questions and producing a customized project.
The speed critique reveals that the heat came from the personalization element, not the AI aspect specifically. The Decider places all three super-vote dots on the AI-generated concept but notes on a sticky that the storyboard should explore whether the three questions could be replaced with a single industry selection, reducing friction. The team moves into storyboarding with a clear concept and a specific iteration to test.
Example: Healthcare App for Patient Check-In (8-Person Team with Rumble)
A hospital system is sprinting on a digital check-in experience to replace paper forms. The eight-person team includes clinicians, administrators, IT staff, and a UX designer. The sprint challenge is 'How might we reduce patient check-in time from 12 minutes to under 3 minutes?' The Decider is the Chief Medical Officer, who is present for Day 2 and 3 but was remote for Day 1.
Because the team has eight people, the facilitator pairs them into four sketch teams of two, crossing clinical and administrative roles. This produces four solution sketches instead of eight, keeping the art museum manageable. Two concepts dominate the dot voting: Concept A, a pre-visit mobile form sent 24 hours before the appointment, and Concept B, a kiosk with photo ID scanning that auto-fills demographic data. The dot patterns are nearly equal, but critically, the dots on Concept A cluster on the 'pre-visit' timing panel while dots on Concept B cluster on the 'auto-fill' interaction panel.
During speed critique, the team realizes these are not competing ideas but complementary ones. However, the Decider sees them as fundamentally different user experiences (one requires patient effort at home, the other requires hardware investment) and declares a Rumble. Both concepts will be prototyped and tested with real patients on Day 5. The facilitator documents both winning sketches with their specific dot patterns so the storyboarding team can build two parallel flows.
The Day 5 user tests ultimately show that patients over 65 strongly prefer the kiosk while patients under 40 prefer the pre-visit mobile form, leading to a dual-track implementation.
Example: E-Commerce Checkout Optimization (Remote Sprint, 6-Person Team)
A direct-to-consumer brand runs a remote design sprint using a digital whiteboard tool (Miro or FigJam). The sprint challenge is 'How might we reduce cart abandonment at the shipping options step from 34% to under 20%?' The team of six includes two product managers, two designers, a data analyst, and a frontend engineer. The Decider is the VP of E-Commerce, attending asynchronously due to timezone differences.
The facilitator adapts the physical exercises for remote work. Crazy 8s uses a pre-built Miro template with eight frames per participant, and a shared timer runs in the video call. Participants turn off their cameras during sketching to reduce social pressure. For solution sketches, each person works in a private Miro section hidden from others, then the facilitator copies all sketches to a shared 'gallery' board with names removed.
The dot voting uses Miro's built-in voting plugin, limited to 25 dots per person. ' The heat map shows 40% of all dots on this single element across two sketches that included it. The asynchronous Decider reviews a Loom video walkthrough of the speed critique, the photographed dot patterns, and the sticky note summaries, then places their super-vote via a time-stamped Miro annotation before the next morning. The concept moves to storyboarding with the specific instruction to test the cost-per-day reframing as the core hypothesis.
Example: Internal Tool Redesign for Large Enterprise (12-Person Sprint)
A financial services firm is redesigning their internal compliance review tool. The 12-person sprint team spans compliance officers, legal, engineering, and UX. The sprint challenge is 'How might we reduce the time compliance officers spend on routine document reviews from 4 hours to under 1 hour?' The Decider is the Head of Compliance.
With 12 participants, the facilitator creates four cross-functional sketch teams of three, each producing one solution sketch. Before Crazy 8s, the facilitator spends a full 20 minutes on the Day 1 recap because three team members missed the morning session. During Crazy 8s, the facilitator notices one team of engineers burning all eight minutes on a single detailed panel showing a system architecture diagram. The facilitator intervenes at minute three, reminding them that Crazy 8s is about user-facing concepts, not technical architecture, and resets their timer for five remaining panels.
This real-time correction saves the exercise. The four solution sketches range from an AI pre-screening dashboard to a collaborative annotation tool. Heat-dot voting reveals unexpected clustering: dots concentrate not on the AI concepts (which the facilitator expected to win) but on a low-tech concept from the compliance-and-legal team showing a checklist-based triage system that routes only exception cases to human review. The speed critique surfaces the insight: compliance officers do not trust AI to make judgment calls, but they trust a rules-based triage that they can audit.
The Decider, who initially favored the AI approach, changes course based on the dot data and the speed critique discussion, placing super-votes on the triage concept. This outcome, where the structured process surfaced a non-obvious winner that contradicted the Decider's initial preference, exemplifies exactly why the anonymous sketching and silent voting sequence exists.
Best Practices
Enforce thick markers for all sketching exercises. Sharpies or equivalent force participants to draw concepts at an appropriate level of abstraction. Fine-tip pens let people draw detailed UI wireframes, which slows ideation, creates false precision, and shifts the evaluation toward visual polish rather than conceptual strength. If someone objects that they 'can't draw,' remind them the thick marker makes everyone equally rough.
Keep the creator anonymous until after the speed critique narration. The moment a team knows whose sketch is on the wall, evaluation shifts from 'is this idea strong?' to 'is this person's idea strong?' Seniority, design skill reputation, and interpersonal dynamics all contaminate the signal. Anonymity is what makes the heat-dot voting meaningful rather than political.
Time-box every phase with a visible timer. Crazy 8s without a timer becomes leisurely doodling. The art museum without a timer leads to clumping and accidental discussion. The speed critique without a timer turns into a debate.
Visible countdown timers (a phone propped up, a projected timer, or a physical clock) create gentle urgency that keeps the energy high and prevents any single phase from expanding to fill available time.
Give participants more dots than they think they need (20-30 per person). Scarcity of dots forces premature prioritization and creates a winner-take-all dynamic where only the most obviously good sketches get any signal. Abundant dots let participants mark individual elements they like, even on sketches that are not overall winners. These element-level signals are often more valuable than whole-sketch rankings because they can be recombined during storyboarding.
Separate the Decider's vote from the group's heat map by at least 10 minutes. Ideally, run the speed critique between the heat-dot phase and the super-vote. This gap lets the Decider process the team's signal alongside the critique discussion rather than reacting to dot clusters in the moment. A Decider who votes immediately after seeing the heat map will almost always just ratify the most-dotted sketch, which defeats the purpose of having a Decider in the first place.
Never skip the recap of Day 1 artifacts before sketching begins. Teams that jump directly into Crazy 8s produce sketches scattered across different interpretations of the problem. The 15-minute recap of the sprint challenge, target map, and How Might We notes synchronizes everyone's mental model. The observable consequence of skipping it is that during speed critique, you hear 'Oh, I thought we were solving a different part of the journey,' which wastes an entire sketch slot.
Photograph everything before removing anything from the wall. Physical artifacts are fragile. Sticky notes lose adhesive, marker bleeds, and someone will accidentally lean against a sketch during a break. Capturing the full wall with dot patterns intact preserves the team's collective intelligence in a form that survives the session.
Do this immediately after the Decider vote, not at the end of the day.
Common Mistakes
Allowing verbal pitching of solution sketches before voting
Correction
When creators pitch their own sketches, charisma and verbal fluency replace idea quality as the evaluation criteria. The most articulate person wins, not the best concept. This happens when facilitators feel uncomfortable with the silence of the art museum phase and try to 'help' by asking creators to explain their work. Watch for body language cues: if someone starts gesturing at their sketch during the silent walkthrough, gently redirect them.
The fix is strict adherence to the facilitator-narrates-first protocol during speed critique, with the creator speaking only at the end for 60 seconds of clarification.
Treating Crazy 8s as polished mini-sketches instead of rapid variations
Correction
Some participants spend the entire eight minutes perfecting two or three panels instead of filling all eight. This defeats the exercise because the divergent power of Crazy 8s comes from panels five through eight, where obvious ideas have been exhausted. You can spot this happening when participants are still drawing on panel two at the four-minute mark. ') and explicitly telling the group that unfinished panels are better than polished ones.
A scribbled concept that surfaces a new direction is more valuable than a beautiful rendering of the obvious approach.
The Decider delegates their vote to the group or defers to the dot majority
Correction
The Decider role exists precisely because democratic voting optimizes for consensus, which often means the safest, least controversial idea wins. When the Decider says 'I'll just go with whatever the team picked,' the entire convergence mechanism breaks down. This typically happens when the Decider does not feel confident in their understanding of the sketches. To prevent it, ensure the Decider is actively engaged during the art museum and speed critique.
Seat them with a clear view of the wall. ' That reframing usually unlocks a decision.
Placing dots on whole sketches rather than specific elements
Correction
When participants dot 'the sketch' rather than specific panels, annotations, or interaction moments, the voting data loses granularity. A sketch with 15 dots spread across the entire page tells you 'people liked this' but not what they liked about it. This matters because during storyboarding, the team needs to know which specific elements to preserve. The facilitator should explicitly instruct: 'Place dots on specific parts that excite you, a particular screen, a specific interaction, a clever annotation, not on the sketch as a whole.' Model the behavior by demonstrating on a sample sketch before the exercise starts.
Running design sprint exercises without a completed sprint challenge and target map
Correction
Teams sometimes jump into sketching because it feels productive, skipping the problem definition work from Day 1. The result is a collection of beautifully sketched solutions to different problems. During speed critique, you will hear variations of 'but that's not what I thought we were solving,' and the voting data becomes meaningless because participants are evaluating ideas against different mental models of the problem. If the sprint challenge is not clearly defined, stop the sketch phase and go back to mapping and defining the sprint challenge.
Two hours of alignment work saves an entire day of misaligned sketching and voting.
Generating too many solution sketches by having large groups sketch individually without sub-teaming
Correction
With more than seven or eight participants each producing a three-panel sketch, the art museum and speed critique phases balloon past their useful time window. The team fatigues, dot placement becomes careless by sketch twelve, and the speed critique loses its crispness. For groups larger than seven, pair participants into sketch teams of two or three. Each team produces a single solution sketch, reducing the total number of concepts while still capturing diverse perspectives.
The pairing should cross functional lines (pair an engineer with a marketer, for example) to preserve cognitive diversity.
Other Skills in This Method
Mapping Problems and Defining the Sprint Challenge on Day 1
How to run the Understand phase by creating a problem map, conducting expert interviews, identifying assumptions, and selecting a focused sprint target for the week.
Conducting User Tests and Synthesizing Feedback on Day 5
How to recruit the right test participants, run five moderated usability interviews in one day, capture observations, and identify patterns that validate or invalidate your sprint hypothesis.
Storyboarding the User Journey for Sprint Prototyping
How to create a step-by-step storyboard that translates the winning sketch into a coherent user flow, serving as the blueprint the team follows during prototype day.
Building a Realistic Prototype in One Day
How to rapidly create a high-fidelity, testable prototype using tools like Figma or Keynote that feels real enough to generate authentic user feedback without writing production code.
Planning and Customizing Your Design Sprint Agenda
How to structure the full multi-day sprint agenda, adapt the classic 5-day format into shorter 4-day or Design Sprint 2.0 variations, and prepare all necessary materials and logistics.
Facilitating a Design Sprint as the Sprint Master
How to effectively facilitate each phase of a design sprint, manage group dynamics, enforce timeboxes, and guide teams through structured exercises from start to finish.
Running Design Sprints Remotely with Distributed Teams
How to adapt the design sprint framework for remote or hybrid teams using tools like Miro, FigJam, and Zoom, including async exercises and strategies to maintain energy and engagement.
Frequently Asked Questions
How do I run design sprint exercises with participants who say they can't draw?
This objection comes up in nearly every sprint. The answer is that these exercises do not require drawing skill. Thick markers (Sharpies) are mandatory because they prevent detailed illustration and level the playing field. Show examples of successful sprint sketches that use stick figures, boxes, arrows, and text annotations. Remind participants that the speed critique will be narrated by the facilitator, so clarity of concept matters far more than visual polish. If someone remains anxious, suggest they write out their concept in annotated bullet points across the three panels rather than drawing it. The ideas, not the aesthetics, are what get voted on.
Should I run sketching and voting before or after storyboarding in the sprint?
Always before. The sketch-and-vote sequence (Day 2 and early Day 3) produces the winning concept that becomes the input for [storyboarding](/skills/storyboarding-sprint-concepts) (mid Day 3). Storyboarding takes the selected sketch and expands it into a detailed, frame-by-frame user journey that guides prototype construction. If you storyboard before converging on a concept, you end up storyboarding a vague consensus idea rather than a specific, vetted solution. The sequence is fixed in the Google Design Sprint for this reason.
How long should the complete sketch-and-vote session take?
Plan for 2-3 hours total, broken down as follows: 15-20 minutes for the recap and setup, 8 minutes for Crazy 8s, 5-10 minutes for quiet reflection, 30-45 minutes for solution sketching, 10-15 minutes for the art museum silent walkthrough, 30-45 minutes for the speed critique (scaled to the number of sketches, roughly 3 minutes each), and 10-15 minutes for the Decider vote and documentation. Teams with fewer than five members can finish in about two hours. Larger teams or Rumble decisions push toward three hours. Build in a 10-minute break between solution sketching and the art museum.
What if the Decider's super-vote contradicts the heat-dot results entirely?
This is by design, not a failure. The heat-dot voting captures collective interest, but the Decider has organizational context the team may lack, such as budget constraints, strategic priorities, or regulatory concerns that make the most-dotted concept infeasible. The facilitator should briefly ask the Decider to share their reasoning (one or two sentences), then move on. If team members push back, remind them that the sprint process gives everyone an equal voice through sketching and voting, but the final call belongs to the Decider. The Day 5 user test will validate or challenge the Decider's choice with real data.
How do I adapt these design sprint exercises for a remote or distributed team?
Use a digital whiteboard tool like Miro or FigJam with pre-built templates for Crazy 8s panels and solution sketch sections. Give each participant a private working area during sketching phases, then have the facilitator copy all sketches to a shared gallery board with names removed. Use the tool's built-in voting feature for dot placement. The biggest remote risk is losing the silent evaluation discipline: participants will chat in Slack or unmute to comment during the art museum. Counter this by having participants turn off cameras during sketching and muting all communication channels during the dot-voting phase. For more detailed remote adaptation techniques, see [running remote design sprints](/skills/running-remote-design-sprints).
Why does my sketch-and-vote session keep producing safe, incremental ideas instead of bold ones?
This usually means the Crazy 8s phase is not pushing hard enough. When participants produce only four or five sketches instead of eight, they stay in their comfort zone. The breakthrough ideas live in panels six through eight, where the obvious approaches have been exhausted. Check three things: are you enforcing one-minute-per-panel transitions vocally? Are markers thick enough to prevent detailed wireframing (which slows people down)? Did the Lightning Demos before sketching include examples from outside your industry? If all three are yes and results are still incremental, try a variant called 'Crazy 8s with a twist' where panels five through eight must solve the problem for a wildly different user persona (a child, a CEO, someone with no internet access). This forced reframing often unlocks lateral thinking.
Can I use these exercises outside of a full five-day design sprint?
Yes, the sketch-and-vote sequence is one of the most portable design sprint exercises. Teams regularly extract it for standalone workshops when they need to converge on a direction for any design or product decision. The minimum viable version requires a clear problem statement (replacing the full Day 1 sprint challenge), individual sketching, anonymous display, silent dot voting, and a designated Decider. You can compress it into a 90-minute workshop if you reduce Crazy 8s to 'Crazy 4s' (four panels, four minutes), shorten the solution sketch to a single panel, and limit the speed critique to the top three most-dotted concepts. What you lose is the depth of exploration, so use the compressed version for smaller decisions and the full version for high-stakes strategic choices.