Recurring meeting efficiency trends

Recurring meeting efficiency trends measure how the useful output of a standing meeting changes over time, relative to the time and headcount it consumes. A weekly leadership sync with 8 people that produces 4 decisions and 6 action items in an hour is doing real work. The same meeting six months later, producing one decision and two actions for the same 8 people, is decaying, and the trend is what makes that decay visible.

The metric matters because recurring meetings are the largest hidden cost in most organisations and the hardest to question. A meeting that was useful when it was set up keeps running long after the reason for it has faded, because nobody is accountable for whether it still earns its place. Looking at the trend rather than a single reading separates a meeting that is naturally quiet this week from one that has been declining for a quarter.

Efficiency here is output per unit of attention, not how smoothly the meeting runs. A calm, well-chaired meeting that reaches no decisions is inefficient. The output that counts is decisions made and owned actions created, because those are the only things a meeting produces that the rest of the week then acts on. A meeting that creates plenty of actions but where the action item completion rate stays low is producing the appearance of output, not the substance.

The base efficiency score divides the useful output of one sitting by the attention it cost, measured as attendees multiplied by meeting hours. The trend is then the direction of that score across consecutive sittings of the same recurring meeting. The inputs below are what you need to settle before the number is comparable from week to week.

1. 1

   Decisions made

   Distinct decisions that were reached and recorded in the meeting. A decision is something that was open before the meeting and closed by it. Reaffirming a prior decision does not count, or every meeting scores well for doing nothing.

2. 2

   Action items created

   Actions agreed in the meeting that have a single owner and a due date. Vague intentions with no owner are not output, because nothing happens after the meeting ends.

3. 3

   Attendees

   The number of people who actually attended, not who was invited. This is the denominator that punishes overlarge meetings, where adding bodies rarely adds output.

4. 4

   Meeting hours

   The scheduled length in hours. Hold this consistent, because a meeting that runs long to reach the same output is becoming less efficient even if the output count is unchanged.

A worked example. A weekly operations meeting has 6 attendees and runs for 1 hour. In one sitting it produces 3 decisions and 5 action items, giving an output of 8 over an attention cost of 6, an efficiency of 1.33. Eight weeks later the same meeting has crept to 9 attendees and 90 minutes, producing 2 decisions and 3 actions. The score is 5 divided by 13.5, which is 0.37. The output fell while the cost rose, and the trend has dropped to roughly a quarter of where it started. That decline is the signal, not either reading on its own.

A metric tree decomposes meeting efficiency into the things that drive output up and the things that drive attention cost up, then traces each down to a specific cause. This turns a falling score into a diagnosis of why the meeting is decaying.

The first level splits efficiency into output produced and attention consumed. Output breaks into decisions reached and owned actions created, each of which depends on the right people being present and the agenda being decision-shaped rather than update-shaped. Attention cost breaks into headcount and duration, each of which tends to creep upward over time without anyone deciding it should. When the score falls, the tree shows whether output dried up, cost crept up, or both at once, which are three different problems with three different fixes.

This is the gap between a dashboard and a decision. A dashboard shows the meeting score dropping. The tree shows that output held steady while attendee count grew from 6 to 11 as people were added defensively, so the fix is trimming the invite list, not changing the agenda.

There is no universal target for a meeting efficiency score, because the right level depends on the meeting type. A decision-making meeting should score far higher than an information-sharing one. The honest benchmark is your own trend, but the bands below give a practical sense of whether a recurring meeting is producing output worth its cost, using decisions plus owned actions over attendees times hours.

The number to watch is the direction over a quarter, not the level in any one week. A meeting that has fallen from 0.9 to 0.3 over two months is decaying even though 0.3 is not catastrophic on its own. A stable 0.4 for a status meeting may be perfectly acceptable. The benchmark frames the conversation, the trend tells you whether to act.

Improving the trend means fixing the branch that is dragging it down, not running better meetings in general. The metric tree points at whether output fell or cost rose, and each has a concrete fix.

The decomposition decides the intervention. If output fell, fix the agenda and who is in the room. If cost rose, prune attendees and shorten the slot. Treating a declining score as a sign people need to focus harder, when the real cause is that the meeting has doubled in size, wastes the effort.

KPI Tree lets you model this by connecting the efficiency trend to the meetings and owners behind it. Each recurring meeting carries RACI ownership, so a single accountable person owns whether it still earns its place rather than it running on autopilot. When the trend falls past a threshold, the metric pushes to that owner so the decay is surfaced while it can still be fixed, instead of being noticed a year later. The verified impact loop then checks whether the decisions a meeting produces actually move the metrics they were meant to, so you learn which meetings are doing real work and which are theatre.

1. 1

   Counting attendance as output

   A full room is a cost, not a result. Measuring how many people showed up rewards exactly the behaviour, overlarge meetings, that the efficiency score is meant to discourage.

2. 2

   Scoring updates as decisions

   Reaffirming what was already known is not output. If status updates count as decisions, every meeting scores well for changing nothing, and the metric loses its meaning.

3. 3

   Reading one week in isolation

   A single quiet sitting is normal. The signal is the direction over a quarter. Reacting to one low reading, or relaxing because of one high one, misses the decay that the trend exists to catch.

4. 4

   Ignoring duration creep

   Holding output constant while the meeting runs longer still means the meeting is getting less efficient. Tracking output without tracking the time it took hides half the picture.

5. 5

   Never acting on the number

   A meeting efficiency trend that nobody owns changes nothing. The point of the metric is to give someone the standing to shrink, shorten, or cancel a meeting, so the score has to land with an accountable owner.