Why anonymous polling works (when public discussion doesn't)
Public discussion has a known failure mode. To speak in front of a room, a person has to decide their thought is fully formed, decide it's worth interrupting the flow, decide it won't make them look stupid, and beat the four other people doing the same calculation. Confident people clear those filters in seconds. Everyone else stays quiet.
You end the workshop with what feels like consensus. It isn't — it's selection bias.
Anonymous submission removes all four filters at once. People write before they speak. Submissions are not attributed to faces. The group sees a list of ideas, not a list of people. The political layer — who agreed with whom, who pushed back on the boss — disappears.
What shows up in its place is the actual signal. The junior designer's quiet observation that the strategy is missing a piece. The remote participant's two-line summary that reframes the problem. The opinion the room had but no one would have said out loud.
The pattern is older than the software. Facilitators in the 1980s called it brainwriting. What changes with software is the friction: instead of paper or sticky notes, the audience answers from their phones, and you see results live.