Meeting duration analysis

Meeting duration analysis is the practice of measuring how long meetings run and comparing that length against the work each meeting needs to do. It looks at average duration, total duration across a team, and how that time is split by meeting type, so you can see where hours are going. A thirty-minute slot that always runs to forty-five minutes is a duration problem. A weekly review that books a full hour but resolves in twenty minutes is a different one.

The simplest way to read it is in two numbers. Average duration tells you how long a typical meeting runs. Total duration tells you how much collective time the team spends in meetings overall. If twelve people each sit in fifteen hours of meetings a month, that is 180 hours, the equivalent of more than a full working month of effort, spent on talking rather than doing. Both numbers matter, because a short average can still add up to a large total when meetings are frequent.

Duration analysis matters because meeting length is rarely chosen on purpose. Calendar tools default to thirty or sixty minutes, and meetings expand to fill whatever slot they were given. Nobody reviews whether the standing hour still needs an hour. Measuring duration turns that quiet drift into a number you can act on, the same way you would track cycle time for any other process.

The headline figure is average meeting duration, which is total meeting minutes divided by the number of meetings. The more useful figures come from breaking that total down by who attends, how often meetings recur, and what type of meeting it is. The inputs below are what you need to capture before the numbers mean anything.

1. 1

   Actual meeting length

   The real start and end times, not the scheduled block. A meeting booked for an hour that ends in forty minutes should count as forty minutes, and one that overruns to seventy should count as seventy. Calendar exports often only carry the scheduled length, so this is the input most teams get wrong.

2. 2

   Attendee count

   The number of people in each meeting. Duration alone understates cost. A thirty-minute meeting with two people costs one person-hour. The same meeting with twelve people costs six. Capturing headcount lets you weight duration by the people it consumes.

3. 3

   Meeting type

   Whether a meeting is a standup, a one-to-one, a review, a planning session, or an ad hoc call. Different types carry different reasonable lengths, so a single average across all of them hides where the overrun actually sits.

4. 4

   Recurrence

   Whether the meeting is a one-off or part of a recurring series. A recurring meeting multiplies its duration by how often it repeats, so a fifteen-minute daily standup costs more over a month than a one-off ninety-minute workshop.

Once you have these inputs, the more revealing measure is person-hours rather than raw duration. Multiply each meeting length by its attendee count, sum across the period, and you get the true labour cost of meetings. A team can have a healthy thirty-minute average duration and still lose hundreds of person-hours a month because the meetings are large and frequent. This is also where duration analysis connects to focus: every hour booked is an hour removed from the deep work that moves a sprint velocity or any other delivery metric.

A metric tree decomposes total meeting time into the drivers underneath it, so you can see whether the cost comes from long meetings, large meetings, or frequent ones. This turns a single uncomfortable number into a set of levers, each owned by a different part of the organisation.

The first level splits total meeting hours into average duration, attendees per meeting, and meeting frequency. Each of those decomposes further. Average duration is driven by the default slot length people book, how often meetings overrun, and how much of the agenda is genuinely needed live. Attendees per meeting is driven by whether the invite list is curated or copied forward, and whether optional people treat the meeting as optional. Frequency is driven by how many recurring series exist and how many one-off meetings get booked on top.

This structure lets you diagnose precisely. If total meeting hours are climbing, the tree tells you whether the cause is meetings getting longer, rooms getting fuller, or the calendar getting busier. Each of those points to a different fix owned by a different person.

There is no universal correct meeting length, because the right duration depends on the work. The useful benchmarks describe how much total meeting time a role should carry and what a healthy average looks like for each meeting type. The ranges below are typical patterns for knowledge teams, not hard limits.

On total time, a common rule of thumb is that individual contributors should keep meeting load under roughly 25 percent of their working week, leaving the rest for focused work. Managers naturally run higher because coordination is their job. When an individual contributor passes 40 percent meeting load, the duration analysis is no longer a curiosity. It is a direct explanation for why delivery is slow.

Improving meeting duration is not about banning meetings. It is about making each one carry only the time it needs. The levers fall into shortening meetings, shrinking attendee lists, and removing meetings that should never have been live in the first place.

The metric tree approach to improving meeting duration starts by finding which driver carries the largest gap against benchmark. If average duration is healthy but total hours are high, the problem is frequency or attendee count, and shortening individual meetings will barely move the number.

KPI Tree lets you model this by connecting each driver to the team and person who owns it. The default slot length and the agenda standard sit with operations or the team lead. The invite list discipline sits with each meeting organiser. The frequency of recurring series sits with whoever scheduled them. With RACI ownership on every node, the accountable owner is named, and when total meeting hours move in the wrong direction the platform pushes the change to that owner rather than leaving it buried in a calendar nobody reviews.

1. 1

   Measuring scheduled length instead of actual length

   Calendar blocks show what was booked, not what happened. Meetings that routinely overrun look fine on paper while quietly costing the extra time. Capture real end times or the whole analysis understates the cost.

2. 2

   Ignoring attendee count

   Duration on its own says nothing about cost. A short meeting with fifteen people consumes more time than a long meeting with two. Weighting duration by headcount is what turns the metric into a true labour figure.

3. 3

   Averaging across meeting types

   A single average that mixes standups, one-to-ones, and workshops hides where the overrun lives. Break duration down by type so you can see which category is actually stretching.

4. 4

   Treating recurring meetings as one-offs

   A fifteen-minute meeting feels cheap until you multiply it by every day it repeats. Recurring series are where total time accumulates, so they need to be costed across the whole period, not per occurrence.

5. 5

   Cutting time without cutting purpose

   Forcing every meeting shorter without removing any of the work just compresses the same agenda into less time and pushes the rest into follow-ups. Reduce duration by removing what does not need to be live, not by rushing what does.