Monthly platformhealth review

When to use

Hold this review in the first week of every month, covering the month before. Plan one session for each group of platforms that shares a Product Manager.

Roles

• Product Manager: leads. You present each platform's scorecard and decide what needs action.
• Program Manager: brings the cross-platform view and the open action list, and records decisions.
• Home Team Product Owners and lead engineers: explain the data and agree the actions their teams will take.

Before you start

• Pull the month's data for the six measures: availability, error budget, incidents and MTTR, performance, security and currency, and cost.
• Fill in the scorecard for each platform, and mark each measure green (on target), amber (close to target) or red (missed).
• List the open actions from earlier reviews and from post-incident reviews.

Steps

1. Review last month's actions. Go through each open action. Close the ones that are done, re-date the ones still moving, and escalate the ones that are stuck.
2. Walk each scorecard. For each platform, spend most of the time on red and amber measures. Skip the greens unless something surprising happened.
3. Apply the error budget rule. If a platform used up its error budget, confirm that reliability work goes ahead of Discretionary work in its Home Team backlog until it recovers.
4. Look for patterns. The Program Manager highlights repeats across platforms, such as the same contributing factor in several incidents or the same dependency failing.
5. Agree actions. For each red measure, agree one action with an owner and a due date. Keep them small enough to finish within a sprint or two.
6. Check the Run allowance. If a platform is red for two months running, the Program Manager records a recommendation to raise its Run allowance at the next capacity planning.
7. Publish the scorecard and the actions within 2 working days.

Outputs

• A published scorecard for each platform, with green, amber and red ratings.
• Actions with owners and due dates, in the Home Team backlogs.
• Run allowance recommendations for the next capacity planning.

Pitfalls

• Reading the whole scorecard aloud. Focus on red and amber. Everyone can read the greens themselves.
• Actions without capacity. Reliability actions come out of the Run allowance. If that allowance is always used up, raise it rather than letting actions slip.
• Ignoring cost and security because nothing is broken yet. Those are the risks that build up out of sight.