Skip to content
Atlassian

Atlassian Team Playbook

Run tomorrow · Primary + examples

Atlassian Working Agreements

A practical team-norm setting play that only works if the agreements stay behavioral, specific, and reviewable.

Best opened when

Open it when recurring collaboration issues keep slowing delivery, handoffs, or workshop quality.

What to steal first

Borrow the rule that every agreement must be behavioral and reviewable within a short horizon.

What not to copy

It is not a culture memo.

Core operating sequence

Play anatomy

Working Agreements gives a team a facilitated way to turn recurring collaboration complaints into explicit, behavioral contracts. The format starts from real friction — not idealized team values — and translates those frictions into agreements someone could observe in practice. The session is only useful if it produces agreements specific enough to inspect.

The transferable core is the behavioral specificity test and the review date. An agreement is behavioral if someone could notice it happening or not happening in the next two weeks. The review date is what converts the output from a workshop artifact into an operating norm. Without a return to the agreements, they decay immediately.

What does not transfer: the Atlassian template format, the exact dimensions or prompts, and the expectation that the agreements will self-enforce. The session creates the contract; the review enforces it. Those are different moments and require different facilitation.

Run it in the room

A clean first pass you can run

Participants
The intact working team. One facilitator who can keep the room in behavioral specifics rather than values talk. 4–10 people.
Timing
60–90 minutes for the initial session. 20–30 minutes for each review session.
Prep
Collect 2–3 specific recent friction examples from team members before the session, without attribution. These become the raw material.
  1. 1Share friction examples without attribution. Name the behavior and its impact, not the person who caused it.
  2. 2Group similar frictions. Each cluster is a candidate agreement area. Aim for 4–6 clusters.
  3. 3Draft agreement language for the top clusters. Each agreement must be behavioral and specific: 'We will share blockers in the daily standup before they are 24 hours old' is an agreement. 'We value transparency' is not.
  4. 4Test each agreement against this question: can you observe whether it is being followed? If not, rewrite it until you can.
  5. 5Agree when and how the team will review the agreements. Without a review date, agreements become wallpaper within a month.

You leave with

3–5 written behavioral agreements with a review date and a named owner for the review process.

First failure point: The agreements describe values rather than observable behaviors. Nobody can tell whether they are being followed, so they stop influencing anything within weeks.

What good looks like

If this is working, these are the signals you should be able to point to

  • Each agreement is phrased as a behavior the team can observe, not a value everyone already endorses.
  • The team can name at least one agreement it expects to be difficult to keep.
  • A revisit date is set and at least one person has committed to bringing back evidence of whether it held.
  • No agreement is on the list because it was safe to say — each one was named because it was the real problem.

How it worked there

The conditions that made it hold

Atlassian designed Working Agreements for cross-functional product teams with recurring collaboration friction — missed handoffs, meetings that duplicated other conversations, and norms that had never been explicitly stated. The play worked because Atlassian's operating culture expected team leads to revisit and enforce the output.

The play requires a room that is safe enough to surface the real friction. If the actual problem is that a senior person's behavior is creating the difficulty and that person is in the room with authority to veto the conversation, the agreements produced will paper over the issue rather than name it. The candor the play depends on is not built by the play itself — it has to already exist or be established by a facilitator who can hold the room.

What not to copy · Failure modes

What goes wrong when this is copied

The agreements sound admirable and cannot be observed. Teams write 'we will communicate openly' and 'we will respect each other's time' — phrases that feel meaningful in the session and are invisible in practice. An agreement that nobody can check is not an agreement; it is a value statement. The play only works when every item is specific enough that someone could notice it being kept or broken.

No revisit date is set, so the agreements become wall copy. The session produces a list. The list is photographed, sometimes posted, and never returned to. The agreements that were hardest to write — the ones about specific behaviors the team has been avoiding — are the first to go unenforceable. The review date is not a formality; it is the mechanism.

The play is used when the problem is structural, not behavioral. Working agreements cannot fix unclear decision rights, misaligned incentives, or delivery dependencies that are outside the team's control. Teams sometimes run it as a substitute for harder conversations about sponsorship, resourcing, or cross-team governance. The output looks collaborative but does not address what is actually causing the friction.

Weak signals to watch for

  • It is not a culture memo.
  • It is not a substitute for fixing broken resourcing or decision rights.
  • Do not frame agreements as broad values statements.
  • Do not close the session without a revisit date and a way to inspect whether the agreements held.

Closest Waypoint move

What to open next

Primary route

Align stakeholders / get on the same page

Use the situation route when working friction comes from deeper directional disagreement rather than norms alone.

Use this when you need explicit behavioral norms and can accept that the agreements will require active facilitation to survive contact with real friction.

technique

Stakeholder Mapping

Use it when the working issue is really about hidden influence and unclear expectations across groups.

Sources and confidence

Primary + examples

Reviewed by Discovery Waypoint Editorial Team · 2026-04-04