On a Monday morning, a marketing manager asks an agent for a launch plan for a new product line. By lunchtime she has a positioning document, three creative routes, a channel plan and draft copy for each channel, all good enough to argue about. By Tuesday she has tightened it into something she'd put her name to. Then she checks when the brand steering group next meets. It is the first Thursday of the month. That was last week.
So the plan, which took a day and a half, waits twenty-four days for a sixty-minute slot, in which it will be the fourth item of five and get eleven minutes. Someone will ask to see a fourth route. The next slot is a month after that.
Making got fast. Deciding didn't.
Most office work has two parts: making the thing, and deciding what to do with it. For most of working history, making dominated. A campaign plan took weeks to put together, so a monthly review cadence was perfectly sensible. By the time the work was ready, the meeting had come round.
Agents have collapsed the making. Drafts, analyses, prototypes and first versions arrive in hours. The deciding hasn't changed at all. It still happens in rooms, at fixed intervals, among people whose calendars were full in August. The result is a strange inversion: the visible work is quick, and almost all of the elapsed time is waiting.
That shift is easy to miss because nobody measures waiting. Project trackers record when a task was started and when it was finished. They rarely record the three weeks it sat in "for review", and they never record the decision nobody made because the right four people weren't free on the same afternoon.
Meetings exist for reasons, and few of them are about information
The tempting response is to declare war on meetings. Plenty of companies have tried: no-meeting Wednesdays, stand-ups where nobody may sit, bans on recurring invites. The meetings come back, usually within a quarter. Chesterton's fence applies. Before you tear something down, find out why it was put up.
A steering committee does several things that have nothing to do with sharing information. It spreads accountability, so that if the launch flops nobody stands alone. It gives senior people a reason to be seen and a place to be consulted. It lets disagreements surface in front of witnesses, which is more honest than settling them in corridors. And it creates a rhythm people trust: things get decided on Thursdays.
Membership matters too. A seat on the steering group is a mark of rank, and the people in it notice when decisions start being made elsewhere. Moving an approval out of the meeting can feel, to the person who used to give it, like a quiet demotion. That reaction is human, and it explains why so many attempts to speed things up stall at the first committee that would have to vote on its own reduced role.
None of that is foolish. But all of it was designed for a world where work arrived slowly. Agents haven't touched any of those needs. They've simply made the queue in front of them much longer.
More options, longer meetings
You might expect faster production to make decisions easier, since the work arrives more polished. Often it does the opposite.
When a buyer spent a fortnight on a supplier comparison, they brought three options to the procurement board, because three was what they had time to research properly. An agent can compare twelve suppliers by lunchtime, with a scoring matrix and a risk summary for each. So the board now gets twelve. Twelve options rarely produce a faster decision. What usually comes back is a request to "return with a shortlist", which means another month.
Parkinson's law holds that work expands to fill the time available. The committee version is that deliberation expands to fill the options available. Cheap production creates abundance, and a committee faced with abundance does what committees do: it asks for more analysis. Since analysis is now cheap too, it gets it, and the cycle repeats.
The calendar is a queue nobody manages
Seen from the work's point of view, a recurring meeting is a service window. Items queue up, the window opens for an hour, a few get processed and the rest roll over. An operations manager would recognise it instantly as a batch process with a monthly cycle time, and would never design a production line that way.
Batching made sense when work arrived in batches. Legal reviewed contracts at a weekly slot because redlines took days to prepare anyway. Now a commercial lawyer can get a first redline from an agent in an hour, check it properly in another, and then wait five days for the weekly review where a partner initials it. The partner's actual contribution takes ten minutes. The wait takes the rest of the week.
Nobody owns this queue. The committee chair owns the agenda but not the throughput. The person waiting owns the work but not the calendar. So the delay belongs to everyone and therefore to no one, which is the most stable kind of problem an organisation can have.
Engineering teams met this earlier than most. Code that took an afternoon to write could wait a week for a change board, and the better teams responded by deciding which changes were routine enough to ship without one. Everyone else is now arriving at the same fork, only with marketing plans and contracts instead of code.
Decide which decisions need a room
The fix isn't fewer meetings. It is deciding in advance which decisions need a meeting at all, and building a faster path for the rest. A few habits help.
Write down decision rights. For each recurring committee, list the kinds of decision it makes and set thresholds below which one named person can decide alone. A campaign under a set budget, or a contract on standard terms, shouldn't wait for a quorum.
Make written approval the default. Circulate the work with a named decider and a deadline. If there are no objections by Thursday at five, it goes ahead. People who care will read it. People who don't were never going to add much in the room either.
Cap the options. Ask the agent for twelve if you like, but bring two or three to the people deciding, with a recommendation attached. Someone has to exercise judgment before the meeting, and that someone should be the person closest to the work.
Keep the meeting for what only a meeting can do: real disagreement between people who both have a point, decisions that change direction for a whole team, moments where trust needs repairing face to face. Those deserve time in a room, and they get far better attention when the room isn't clogged with sign-offs.
A twenty-minute audit
Open your calendar and look at the recurring meetings you chair or attend. For each one, write down the last three decisions it actually made. Not the updates it heard. The decisions.
Some meetings will struggle to name one. Those are rituals. They may be worth keeping for other reasons, but they shouldn't be gates that work has to pass through. Others will show a list of approvals that one person could have given by email in an afternoon. Move those out first. What's left is the work that genuinely needs people in a room, and it will finally get the room's full attention.
The marketing manager's launch plan needed one decision from one person: does this positioning fit the brand? The brand director sat on the same floor the whole time. The plan spent three weeks waiting for a meeting she could have skipped.
