An operations review meeting agenda built around decisions, not dashboards

Start from the decisions

Performance-measurement systems often fail because they collect too many measures, are not linked to strategy, and are not reviewed or updated. Measurement also costs something: data collection, attention and distorted behaviour. More measurement is not automatically better, and the right number of measures depends on context.

Financial measures alone are lagging indicators and can distort behaviour. Adding customer, internal-process and learning perspectives lets managers see the drivers of future performance, not only its financial result. Scorecards can still turn bureaucratic and be gamed, so ownership and a regular audit matter.

Read the data in time order

A run chart plots a measure in time order and uses the median, runs and shifts to separate signal from noise. It is deliberately simpler than a control chart and makes the effect of a change visible without waiting for an aggregate before-and-after comparison. It is less sensitive than a control chart and does not establish cause on its own.

A short, regular huddle can help, with caveats

A scoping review of 158 empirical studies of frontline huddles found that 67.7% reported positive team-process outcomes and 44.3% reported positive clinical-care outcomes. The evidence is heterogeneous, mostly non-randomised, from healthcare, and positive findings may be over-represented, so it supports trying a short regular huddle, not a particular format or frequency.

Run it as a decision forum

People do not optimise over every alternative; they simplify and search until an option meets an aspiration level. A review that expects perfect analysis will stall, so frame each item as a decision with an owner and a date. Different situations also call for different approaches: simple ones suit established practice, complicated ones need analysis, complex ones call for small safe-to-fail tests. Applying one mode everywhere produces predictable failure.

A working agenda

This agenda is Portolan's, built from the playbook steps; it is not a standard. One: decisions due before the next review. Two: the few chosen measures, each as a run chart with its owner. Three: what moved, and whether it is signal or noise. Four: decisions made, with an owner and a date for each. Five: one measure to fix or retire, because a measure that is never reviewed or retired is how a system fills with noise.

The path, step by step

These steps come from the Measurement and operating cadence playbook. Each has a checkpoint you can observe.

1. 01

   Start from decisions, not dashboards

   List the recurring decisions the operation needs to make and who makes them. For each, name the measure that would change the decision and the cadence at which it needs to arrive.

   CheckpointA decision-to-measure map, with an owner and cadence for each decision.

   Common failureAdding every available metric to a dashboard and calling it visibility.

2. 02

   Choose a small balanced set with owners

   Select a handful of measures across financial, customer, process and capability outcomes. Give each a definition, a baseline, a target and one owner.

   CheckpointA one-page scorecard with definitions, baselines, targets and named owners.

   Common failureMeasuring twelve things and owning none of them.

3. 04

   Review time-ordered data in a short cadence

   Run a short, frequent review of the few measures that change weekly. Plot them over time, annotate interventions, and discuss the trend rather than a single snapshot.

   CheckpointA weekly chart with annotations, and a short review that starts from the trend.

   Common failureHolding a monthly meeting to review annual data.

4. 05

   Run the review as a decision forum

   Use a fixed agenda: what changed, what it means, what we will decide, who owns it and when. Separate information sharing from decision-making, and record the decision.

   CheckpointA decision log with owner, date and the measure that will show whether it worked.

   Common failureEnding the meeting with 'let's keep an eye on it' and no owner.

What this guide does not settle

Most sources are methods papers, frameworks and a scoping review of mostly observational healthcare studies. None shows that a particular agenda, length or frequency improves results in an operation like yours, so treat the agenda as something to test and revise after a few cycles.