Confidence Voting
Gauge team confidence on any decision. Anonymous votes, instant results, four built-in scales. Find out what your team really thinks before you commit.
4 built-in confidence scales
Fist of Five (0-5)
Standard confidence scale. 0 = completely disagree or no confidence. 5 = fully confident, all in.
Traffic Light
Red (stop/block), Yellow (concerns), Green (go). Visual and fast.
Thumbs
Up, Down, or Neutral. The fastest possible team check.
Custom Scale
Define your own numeric scale with any range and labels.
What makes it work
Agreement and confidence are not the same thing
A team can agree on a plan and still expect it to fail. Everyone nods, the sprint gets committed, and three people walk out of the room privately certain that the integration work will eat the whole second week. Nobody lied. They were asked whether they agreed, and they did agree that the plan was the plan. Nobody asked them how confident they felt about it.
A confidence vote asks that second question directly, and it asks it anonymously so the answer is not filtered through the room. It takes about ninety seconds and it consistently surfaces the concern that would otherwise show up as a missed sprint goal two weeks later.
Choosing a scale
Fist of Five is the workhorse. Zero means you think this is a mistake, five means you would bet your weekend on it. The middle is where the information lives: a three is somebody saying they will go along with it and would rather not. Traffic Light trades resolution for speed and reads well at a glance, which makes it good for a room that is nearly out of time. Thumbs is the fastest check available and works when you genuinely only need to know whether anybody objects.
Whatever scale you choose, say out loud what the ends mean before you vote. Teams that skip this get a wide spread that means nothing, because half the room read four as good and the other half read four as barely acceptable.
What a divided vote is telling you
The instinct after a split vote is to talk the low voters round. Resist it. The low votes are the reason you ran the vote at all, and a team that learns their twos get argued away will stop voting twos. Ask the lowest voter what would have to be true for them to move up one point. That question is specific, it is answerable, and the answer is almost always an actual risk with an actual owner: an unclear dependency, an environment nobody has tested, a person who is on holiday during the second week.
Then re-vote. Confidence that moves from divided to near after a five minute conversation is worth more than confidence that started high because nobody wanted to be the one holding things up. Confidence that does not move is a signal to change the plan rather than the vote.
Where it fits alongside estimation
Confidence voting pairs naturally with Planning Poker, and you can run both in the same ScrumMastr room without anybody rejoining. Estimate the work first, then ask how confident the team is in the estimate. A sprint full of stories everyone sized comfortably but nobody feels confident about usually means the risk is not in the individual stories, it is in the sequence, the handoffs, or something outside the team entirely.
It works just as well as a standalone check before a release, after an incident review, or when somebody proposes an approach that the room has been politely quiet about. Anonymous, instant, and no account needed for anyone joining.
Keep reading
Try Confidence Voting with your team. Free, no signup.
Create a room