Thursday, 9:30. Six people on a call — three from the agency, three from the client. Somebody says a version of so, where are we? and for the next eleven minutes nobody makes a decision, because nobody can yet. They are establishing facts.
No statement of work ever commissioned this meeting. It shows up around week two of almost every joint project, at the point where both sides quietly discover they cannot answer one simple question — what is the other company actually doing right now? — and once it exists, it never leaves. Cancelling it feels reckless. So it sits in the calendar until the project ends, collecting attendees.
The meeting isn’t the failure. The meeting is the workaround. That’s the part worth looking at.
What the first eleven minutes are for
Watch a joint status call with a stopwatch and it separates cleanly into two activities. One is judgment: choosing between options, accepting a trade-off, deciding who eats a delay. The other is archaeology: figuring out what happened since Tuesday.
Judgment needs people in a room. Archaeology does not. Every fact recovered in those eleven minutes already existed somewhere — in Atlas’s tracker, in a Meridian engineer’s inbox, in the head of a project manager who was on holiday last week. It was simply never in one place, so six people reassembled it out loud.
Do the arithmetic on your own project. Six attendees, half an hour, once a week: three person-hours a call. Over a twelve-week engagement that’s thirty-six hours — a working week, spent by senior people, and a good half of it buying information that was already written down.
The waste is annoying. The failure mode underneath it is expensive.
Two pictures of the same project
Everyone leaves the call aligned. That’s the point of it. What happens next is the interesting bit.
Friday afternoon, Atlas finishes the asset pack early and ships it. Monday, Meridian’s legal review slips a week because the reviewer is in a different time zone and nobody told her she was on the critical path. Neither event crosses the company line. Both sides update their own picture of the project and neither picture is now correct.
By Tuesday the gap has consequences. Meridian’s engineer believes Atlas is still waiting on the brand lead, so he doesn’t queue the sandbox credentials. Atlas believes Meridian has the credentials in flight, so nobody asks. Two competent people, both acting reasonably on the information they have, and between them a dependency is now sitting still.
Nobody lied. Nobody dropped anything. The state simply diverged, quietly, the way it always does.
This is the mechanism behind meeting sprawl on cross-company projects. Nobody wants more calls. But the only instrument anyone has for closing the gap is a synchronous conversation, so when the gap gets painful, the answer is always another conversation.
Why “more communication” isn’t a lever
The standard prescriptions all sound sensible and none of them touch the mechanism.
Send a weekly status report. A status report is a snapshot, taken by an interested party, at a moment of their choosing, describing their own performance. It arrives Friday and is wrong by Monday. It also costs the sending side real hours — hours the receiving side is often paying for.
Keep a shared spreadsheet. Now there are three pictures of the project: yours, theirs, and the one in the sheet that neither side fully trusts because updating it is nobody’s actual job.
Add a mid-week check-in. This works, in the sense that a bucket works on a leaking roof.
The client asking for all of this isn’t being paranoid, incidentally. They are working with the only instruments available. When you cannot see the work, you ask the people doing it to describe the work, and then you meet to discuss the description.
A recurring meeting is what an organisation builds when it can't build a database.
There’s a quick diagnostic here. Imagine cancelling the Thursday call for three weeks and changing nothing else. If both companies would genuinely lose track of where the project stands, then that meeting isn’t a communication ritual — it’s where the project’s state is stored. Thirty minutes a week of six people’s attention, functioning as a database with terrible availability and no history.
Meetings for judgment, not for state
Fixing this doesn’t require better meeting discipline. It requires the joint work to have a state that exists whether or not anyone is talking.
The call doesn’t disappear, and it shouldn’t. Two companies deciding to move a launch date, or working out who absorbs a cost overrun, need to hear each other’s voices. What changes is what the call is for. Archaeology comes off the agenda, and the half hour goes to the two things that genuinely need humans present: judgment, and repair.
The good version of that meeting is short, faintly boring, and entirely about the future.
You’ll know it worked when someone cancels a Thursday and nothing bad happens.