Dot voting

What it is

Dot voting doesn’t have a single canonical origin. The practice spread through mid-twentieth-century quality-management and design-thinking communities as a lightweight vote-with-your-feet convergence step, and became ubiquitous in modern facilitation through IDEO’s design-thinking toolkit, Stanford d.school’s workshops and the broader Liberating Structures community. There is no foundational paper — the technique is old, shared and explicitly uncopyrighted.

The mechanics are simple. After a generative exercise (brainstorm, fishbone, retrospective, SWOT) the board has more items than the team can act on. The facilitator hands each participant the same number of dots — stickers on a physical board, a feature in Miro/Mural, or a show-of-fingers in a small group — and asks people to place them on the items they think matter most. Counts are tallied. The highest-dotted items become the focus of the next exercise.

Why it works isn’t the voting; it’s the quietness of it. A spoken-rank discussion tends to surface the loudest voice’s priorities first. Dot voting runs silently, in parallel, and reveals what the room actually thinks — which is often different from what the room has been saying.

When to use it

Use dot voting when:

• You have more items than you can act on and need to pick the top few. Dot voting converges a wide-open board into a short action list in two or three minutes.
• The team just ran a silent-generation exercise — dots are the natural convergence for 1-2-4-All, a fishbone or a retrospective board. Generation goes wide; dots bring it back.
• You want a low-stakes, visible signal of priority without forcing consensus. Dots show distribution — unanimous versus split — which itself is useful information.
• The group is mixed seniority or the discussion is at risk of being anchored by a loud voice. Dot voting routes around that.

Don’t use dot voting when:

• You need a ranked ordering across the full list. Dots only reveal the top items; everything below the cutoff is just noise. If you need a stack-rank, use force-ranking or a 2×2 prioritization matrix.
• The items aren’t comparable. Dots assume every item on the board is a candidate for the same kind of decision. If the board mixes action candidates with observations or questions, sort first, then vote within the action candidates.
• You need a weighted decision (effort × value, risk × impact). Dot voting is one-dimensional. For weighted decisions, use a 2×2 matrix.
• The group is three or fewer people. Three votes spread across three voters is statistical noise. In small groups, discuss directly.

How to run it

Total time: 2–4 minutes. Faster in small groups; slightly longer when people want to read the full board before voting.

Set the budget (15 seconds). Three dots per person is the default. Use four or five if there are many items (20+) on the board. Fewer than three tends to force first-impression voting without room for second choices.

State the prompt (15 seconds). Make the question explicit. “Place your dots on the causes you think are highest-leverage to fix.” Or “the failure reasons you think most likely to actually hit us.” The prompt shapes which items get attention; vague prompts produce vague results.

Silent voting (1–2 minutes). People walk up and place dots — or drag them on a virtual board. No talking during voting. Multi-dotting (putting more than one dot on the same item) is allowed by default unless you rule otherwise.

Tally and read out (30 seconds). The facilitator counts dots visibly and reads the top three or four items. That shortlist becomes the input for the next exercise (5 Whys, A3, action drafting, whatever comes next).

Don’t debate the result. The vote is the vote. If the result feels wrong to someone, that’s a signal to investigate in the next exercise — not to re-run the vote. Re-running a dot vote teaches the room that their dots don’t count.

Worked example

A team finishes a fishbone diagram on recurring deployment failures. The board has 14 candidate causes spread across six category branches. Each of the six participants gets three dots. The facilitator’s prompt: “Place your dots on the causes you think are highest-leverage to address this sprint.”

After 90 seconds of silent voting, the board looks like this — each card is a cause from the fishbone, with a red dot for every vote it received:

Eighteen dots cast across fourteen candidates (six participants × three dots). The starred top pair — staging drift and no rollback rehearsal — tied for the shortlist. The third-place item goes on the watch list. Everything under two dots is context.

The facilitator circles the top two. Those become the action candidates for the sprint. The 3-dot item is flagged for the watch-list. Everything else is context, not action.

What the distribution tells you. Two items tied for the top with four dots each, out of 18 total dots cast. That’s not a runaway winner — it’s a split room, with two equally-weighted concerns. Acting on both is the right call. Had one item had eight dots and the rest scattered, the call would have been to focus exclusively on that one.

Common failure modes

• Facilitator votes. The facilitator’s dots carry social weight they shouldn’t. Unless the facilitator is genuinely an equal participant (a playing coach, not a scorekeeper), don’t vote. If you feel strongly about an item, add it to the conversation through the prompt, not the tally.
• Dots before the board is understood. People vote based on what they remember of the discussion, not what’s actually on the board. Fix: give the room 30–60 seconds of silent read-through before distributing dots.
• Strategic voting. Someone stacks three dots on a pet item they know nobody else will support, to guarantee it makes the shortlist. Usually harmless in small teams, but corrosive if it becomes a pattern. Fix: cap stacking at two dots per item when the group is larger than eight.
• Voting on vague items. “Communication” gets four dots. What does that even mean? Fix: before the vote, rephrase vague items into something specific enough to act on. If it can’t be rephrased, it isn’t an action candidate.
• Treating the dots as the conclusion. The dots tell you where to focus next, not what to do. The next exercise — root-cause analysis, A3, pre-mortem — does the real work. Teams that stop at the dots produce dot-voted shortlists and no actions.

References

In the playbook

• Fishbone diagram — the fishbone’s “mark the vital few” step is a dot vote.
• Pre-mortem — the cluster-and-rank step uses dot voting to pick which failure reasons get mitigated.
• SWOT analysis — dot voting is the natural convergence when the SWOT quadrants fill with too many items to discuss.
• Pareto analysis — dot voting’s qualitative cousin; Pareto works from data, dots work from judgment. Use Pareto when you have numbers, dots when you don’t.

External references

• IDEO, Design Kit: Methods — dot voting appears throughout the IDEO design-thinking toolkit under names like “Determine What to Prototype” and “Bundle Ideas.”
• Stanford d.school, Design Thinking Bootleg — the d.school’s facilitator deck formalizes dot voting as one of its convergence tools.
• James Macanufo, Sunni Brown & Dave Gray, Gamestorming (O’Reilly, 2010) — includes dot voting among the core facilitation games with practical variants.
• SessionLab, Dot voting — workshop-ready facilitator notes with timing and material suggestions.
• Atlassian Team Playbook, Prioritization — modern remote-friendly variants using digital whiteboards.