Check in on your team without putting anyone on the spot
Anonymous mood checks help you understand team energy and sentiment. Track trends over time. Spot issues before they become problems.
When to run a mood check
Start of stand-up
Quick energy check before diving into updates. Takes 30 seconds.
Before a retro
Set the tone by understanding how the team feels first.
After a tough sprint
Check in on the team when you sense things were hard.
Weekly pulse
Run a mood check every week to track trends over time.
What you get
Why ask at all
Teams rarely fall apart suddenly. They slide, and the slide is visible for weeks before anything shows up in the delivery numbers. People stop arguing in code review. The retro gets shorter every sprint. Somebody who used to volunteer stops volunteering. Every one of those is easy to miss individually and obvious in aggregate, which is exactly what a regular mood check gives you.
The other reason is that asking changes the norm. A team that gets asked how it is doing every week, in a format that takes half a minute, develops a much lower threshold for mentioning that something is wrong. That threshold is worth more than the data.
Where it fits in the week
The two placements that stick are the start of a standup and the opening of a retro. At standup, an energy check takes thirty seconds and gives the whole team the same information the manager has: whether everyone is running on empty this week or just one person is. At a retro it does something different, setting the emotional context before the conversation about process starts, so that a team having a hard time does not spend forty minutes politely discussing their board columns.
A weekly pulse in a channel works too and asks even less of people. Pick one placement and keep it for a couple of months. A mood check that moves around gets forgotten, and comparisons across rounds are only meaningful if the conditions stay roughly the same.
What to do with what you learn
Share the result with the team, always, and share the distribution rather than a single number. Then pick one thing. A mood check that produces a list of eight action items produces nothing, because eight items is a way of avoiding a choice. One change, named, owned, and revisited at the next check is what makes the exercise worth repeating.
Sometimes the right response is to name what you see and ask nothing. If the chart dips during a hard release week and everyone knows why, saying so out loud is a complete and sufficient response. Not every signal needs an intervention, but every signal needs acknowledgement.
What makes people stop answering honestly
Three things, reliably. Hunting for who submitted the low score, which teaches everyone that anonymity is theoretical. Collecting the data and never mentioning it again, which teaches them it is a formality. And running it so often that it becomes background noise, at which point everybody clicks the middle option without reading.
Anonymous voting handles the first problem mechanically: in ScrumMastr the host sees the distribution and the trend, never who voted what. The other two are habits rather than features. Ask less often than you think you need to, and respond in public every single time.
Keep reading
Happy teams build better products. Start checking in.
Create a room