Dot voting with a live poll fixes the wall, not the method
Fourteen people stand up, shuffle to a wall of sticky notes, and spend ten minutes deciding where to put three dots each. The VP goes first. The remote half of the workshop watches a webcam pointed at paper. Then someone recounts because two dots fell off. Dot voting is a good method with a bad medium. Dot voting with a live poll keeps the method and fixes the medium: same shortlist, same budget of picks per person, counted in seconds, visible to the people on the call.
We run prioritisation votes this way in our own planning sessions, and we are not precious about it. Sticky dots still win in a few rooms, and we say which ones below. For most workshops with a screen and a remote attendee, the poll is faster and fairer.
Why the wall distorts the vote
Early dots herd later dots
The first three dots on the wall are information for everyone who votes after them. Muchnik, Aral and Taylor ran a randomised experiment on a social news site. A single early positive vote created herding that inflated later ratings, while the crowd corrected most early negative votes (Social Influence Bias: A Randomized Experiment, Science, 2013). A wall of dots is that experiment run in a conference room, with the most senior person voting first. Nielsen Norman Group's facilitation guide lists the same failure modes by name: persuaded voting, groupthink, and the HIPPO effect (Dot Voting: A Simple Decision-Making and Prioritizing Technique in UX). HIPPO is the highest-paid person's opinion, and it drags the dots.
Remote attendees cannot reach the wall
A hybrid workshop with a physical wall has two classes of participant. The people in the room vote. The people on the call describe where they would like their dot placed, and someone else sticks it on for them, usually after the in-room votes are already visible. We have watched remote attendees give up and agree with whatever cluster was forming. That is not a vote.
The count is slow and arguable
Dots fall off. Dots get moved. Two people put four dots on one note and swear they only had three. The facilitator counts, someone disagrees, the facilitator counts again. Ten minutes of a ninety-minute workshop go to arithmetic. A poll counts itself.
The multi-select live poll version, step by step
- Turn the shortlist into one poll. Each idea becomes an option. Options have a 100-character limit and the question has 250, which forces the kind of crisp phrasing a sticky note never gets.
- Allow multiple selections. Voters get checkboxes instead of a single choice. Announce the budget out loud: "Pick three." Nielsen Norman Group suggests roughly a quarter of the options as the number of votes per person, so twelve ideas means three picks. We announce the budget and trust the room rather than policing it. The bar chart still works if someone picks four.
- Decide named or anonymous before you open it. If the HIPPO is in the room, uncheck "Ask name." Every voter gets a random ID, nobody has to campaign, and the person who disagrees with the VP can say so with a tap. The reasoning is the same as in our piece on why anonymous polls are not a nice-to-have.
- Keep the projector on the deck until the poll closes. The live bar chart updates as votes arrive, which is exactly the herding channel we just removed from the wall. So we do not project it while voting is open. The host watches the chart on the laptop, the room sees the question slide, and the chart goes on the big screen only after the close.
- Close with a timer or a manual stop. Set a countdown of 30 to 45 seconds, or stop it by hand once the votes stop arriving. The timer minimum is 15 seconds, which is already enough for three taps.
- Read the spread, not just the winner. Multi-select percentages do not add up to 100, because each percentage is the share of voters who picked that option. An option at 71 percent was picked by seven in ten voters. Two options at 58 and 44 are a real second tier. One option at 20 is a candidate for the parking lot.
The whole sequence takes thirty seconds of voting and about a minute of reading. Nobody leaves their chair. The person dialling in from another time zone voted at the same moment as everyone else.
When sticky dots still win
We would be lying if we said the poll always wins. Keep the physical dots when the group is small and co-located, under eight people around one table. Keep them when the options are spatial, like a journey map or a diagram where the dot's position carries meaning. Keep them when the energy of standing up is the point of the exercise, which happens more often after lunch than facilitation guides admit. Everywhere else, the wall is costing you time and silencing the quiet half of the room.
The budget matters too. Our free tier covers 100 participants per session, which is more than any workshop we have ever had to prioritise with. A facilitator does not need a paid seat to replace a pack of stickers.
Questions facilitators ask us
How many picks should each person get?
About a quarter of the number of options, rounded to a whole number. Twelve options, three picks. Twenty options, five. Fewer than that and the vote stops discriminating. More and everything gets a dot.
Can we run it with remote attendees?
Yes, and that is the main reason to switch. Everyone votes from a phone browser via PIN or QR at the same moment. There is no second-class vote relayed through a colleague.
What do we do about ties?
The same thing the facilitation guides recommend for the wall: narrow and revote. Build a second poll with the tied options only, give everyone one pick, and close it. Another thirty seconds.
Keep the dots, lose the wall
Dot voting with a live poll is the same prioritisation exercise with the herding removed and the remote half of the room let back in. Set up one multi-select question in the live poll, announce the budget, keep the chart off the projector until the close, and read the spread. If your brainstorms also die on the whiteboard, our piece on fixing broken brainstorms covers the step before the vote.