The weekly support operating review

Most weekly support meetings are status theater. A dashboard is read aloud, a large backlog number creates anxiety, and the group leaves with the same problems renamed as observations. An operating review has a harder standard: it must change ownership, capacity, a customer promise, or a system.

The meeting is not a tour of every metric. It is a control loop for the few signals that reveal whether the operation is becoming safer or more fragile. Occupancy is a useful example: professionals commonly plan around a practical ceiling near

, because running continuously hotter removes the slack needed for variation, coaching, and recovery.

Bring one page, not twelve dashboards

Use a single decision sheet with six blocks:

1. Demand: arrivals by reason and channel, versus plan.
2. Access: age bands, service level, abandonment, and oldest unresolved work.
3. Resolution: repeat contact, reopen, transfers, and exceptions.
4. Quality: calibrated failure themes, not just an average score.
5. People: occupancy, absence, schedule pressure, and coaching capacity.
6. Improvement: the three changes expected to remove demand or reduce failure.

Every measure needs a definition and an owner. If two attendees cannot explain how a number is calculated, move the decision to the metric dictionary before debating the trend.

Run the 45-minute agenda

Minutes 0–5: promises at risk. Name any customer, regulatory, security, or service promise likely to fail this week. Start with consequence, not averages.

Minutes 5–15: demand and capacity. Compare arrivals with the forecast by interval or day. Separate a temporary spike from a persistent mix change. Decide whether to rebalance work, protect a queue, adjust a promise, or investigate the driver.

Minutes 15–25: failure demand. Review repeat contacts, reopens, escalations, and the oldest cases. Choose one contact driver whose cause sits outside support and assign a cross-functional owner.

Minutes 25–35: quality and people. Bring two or three calibrated examples. Ask what environment produced the behavior: unclear policy, missing knowledge, broken tooling, authority gaps, or a coaching need. Do not turn every system defect into an agent action.

Minutes 35–42: improvement work. Review only initiatives with a named outcome and delivery owner. Stop work that has no evidence of movement.

Minutes 42–45: read back decisions. State the decision, owner, due date, affected promise, and how the group will know it worked.

Keep three lists

Maintain a decision log, a risk register, and an experiment list. The decision log prevents the same trade-off being reopened every week. The risk register makes deferred work explicit. The experiment list distinguishes a temporary test from policy.

Avoid a giant action tracker. If the meeting creates more than five new actions each week, it is probably distributing anxiety rather than making choices.