Team Health Check
Popularised by Spotify, the Team Health Check is a regular ritual where teams self-assess across key dimensions of healthy team functioning (e.g. 'Suitable Process', 'Teamwork', 'Delivering Value', 'Morale', 'Learning'). Each dimension is rated as healthy (green), some concerns (yellow), or unhealthy (red). Trends over time surface systemic issues before they become crises.
Déroulé d'animation
- 1
Set the frame: explain that results stay with the team and exist for self-improvement, not management reporting. Show the green/yellow/red meanings.
5 min - 2
Lay out the 8–12 dimension cards and walk through them once so everyone knows what is coming. Adapt or drop cards that do not fit your context.
5 min - 3
For each dimension: read the 'healthy' and 'unhealthy' descriptions aloud, have everyone vote simultaneously with cards or fingers, then capture the result and a one-line 'why' from a yellow or red voter.
20 min - 4
Review the full board together and, if you have previous results, compare against them — call out dimensions that moved rather than dwelling on absolute colours.
5 min - 5
Pick the top 1–2 dimensions needing attention and agree one concrete improvement action with an owner for each.
8 min - 6
Record the results for trend tracking and put the next health check (4–8 weeks out) in the calendar before people leave the room.
2 min
Conseils
The power is in the trend, not the snapshot — run it every 4–8 weeks and track changes.
Don't use results to evaluate teams externally.
The goal is team self-improvement, not management reporting.
Pièges courants
Sharing results upward as a management report — teams learn to vote green to look good and the check becomes theatre within two or three runs
Letting people vote one after another instead of simultaneously, so later voters anchor on the first colours shown
Deep-diving every dimension as it is voted on — twelve discussions blow the timebox and exhaust the room before the priorities are chosen
Running the check once and never again — a single snapshot invites debate about absolute scores, while the real signal is the trend
Closing without selecting 1–2 focus areas and owners, which turns the session into a colour ritual that changes nothing
Variantes
Adapt the dimensions to your context — the generic Spotify dimensions may not fit every team. Add a 'comments' round after voting for each dimension to surface the 'why' behind the colours.
Contextes d'utilisation
Quand l'utiliser
You want a recurring ritual (every 4–8 weeks) that tracks team health as a trend rather than a one-off snapshot
The team senses something is off but lacks the vocabulary to name it — the dimension cards give structure to a vague unease
A newly formed or recently re-organised team needs a baseline to measure its development against
Standard retrospectives have gone stale and the team keeps rehashing the same sprint-level topics
You want to surface systemic issues — process, morale, delivery friction — before they escalate into crises
Quand ne pas l'utiliser
Management wants to compare or evaluate teams — the moment scores feed a report upward, teams vote green to look good and the signal dies
The team is in an acute interpersonal conflict; a facilitated conflict conversation or Working Agreements session addresses that more directly
You only care about the last sprint's events — a Sailboat Retrospective is better suited to iteration-scoped reflection
Psychological safety is low and public colour votes will not be honest; run an anonymous survey first and build toward open voting
You cannot commit to repeating the check — a single run without trend data delivers a fraction of the method's value
Méthodes associées
Pour aller plus loin
Questions fréquemment posées
What are the best tips for running a team health check?▾
Vote simultaneously so nobody anchors on the loudest voice, run the check on a regular 4–8 week cadence and track trends rather than single snapshots, and keep results inside the team instead of reporting them to management. End every session by picking just 1–2 dimensions to improve, each with a named owner. Adapting the dimensions to your context matters more than copying Spotify's exact list.
What dimensions should a team health check cover?▾
Use 8–12 dimensions. Spotify's original set includes Delivering Value, Teamwork, Learning, Mission, Speed, Suitable Process, Support, Fun, Easy to Release, Health of Codebase, Awesome Codebase, and Pawns or Players. Swap in dimensions that fit your team — a generic list a team does not recognise produces vague votes.
How often should we run a team health check?▾
Every 4–8 weeks. The power of the method is in the trend, not the snapshot: a yellow that has been yellow for six months is a very different signal from a green that just turned yellow. Less frequent than eight weeks and the trend line is too sparse to act on.
How long does a team health check take?▾
Plan 30–60 minutes for a team of 3–15 people covering 8–12 dimensions: a short framing, roughly two minutes of reading and voting per dimension, then time to pick the top 1–2 areas and agree improvement actions. Cutting the action step to save time removes most of the value.
Should team health check results be shared with management?▾
Not as an evaluation or cross-team comparison. The check works because teams answer honestly, and honesty disappears once scores feed performance reporting. If leadership wants visibility, share the improvement actions the team chose rather than the raw colours.
Planifiez votre prochain atelier avec l'IA
Workshop Weaver vous aide à combiner des méthodes comme Team Health Check en un agenda complet et minuté en quelques minutes.
Essayer gratuitementMethod descriptions on Workshop Weaver are original content written by our team, based on established facilitation practices.