Somebody sent a message on Tuesday morning. A short one — a question they needed answered before they could carry on with anything.

The answer arrived Thursday afternoon.

Nothing went wrong there. The person who answered was not ignoring anyone; they were in workshops for two days, it was not marked urgent, and they replied properly as soon as they had a moment to think about it. Two days is a civil response time. I have waited considerably longer for less.

Two days is also two days. It appears in no system, on no report, in nobody’s velocity, and if you asked either person about it a week later neither would remember it happened. Now multiply it by everyone who is currently waiting on somebody else, which is very nearly everyone, and you have what I suspect is the largest single category of delay in most organizations — and the only one with no number attached to it anywhere.

This is the chapter where the structural argument runs out. You cannot draw a boundary that makes people answer each other faster. What closes that gap is somebody deciding another person’s problem is worth ten minutes of their afternoon, and there is no org chart that produces that.

That is the whole observation, and it is smaller and more stubborn than most things in this paper.

Waiting on another person is the most common form of friction in any organization and the only one with no instrument pointed at it. A queue in a system can be measured. A ticket has an age. But the question sitting in somebody’s inbox behind four other things has no age, no owner, and no record that it was ever asked — and when it is finally answered, both parties experience the exchange as having gone fine.

Multiply it properly and the number stops being comfortable. If everyone in a two-hundred-person organization is waiting on somebody for an average of a day a week — which is a conservative guess and I would take a higher one — that is forty people’s worth of time, permanently, spent on nothing, in a place where every individual instance is unremarkable and nobody is doing anything wrong.

In the vocabulary this paper uses elsewhere, every one of those waits is a queue — just one with no name, no owner and no visible position in it. Which is why it is the hardest kind to shorten: you cannot staff a queue nobody has admitted exists.

The reason it stays invisible is that the waiting is never anybody’s whole day. People move on to the next thing, which is exactly the right professional response and exactly what removes the evidence. Nobody sits idle. The work just arrives later than it would have, and there is nothing to point at.

What moves this is not a system, and I have watched several organizations discover that expensively. It is closer to a norm — that a question from someone blocked gets ten minutes today rather than a proper answer next week, and that unblocking somebody counts as work rather than as an interruption to it. That is cultural rather than structural, which makes it slower to establish and much harder to lose once it holds.

The largest queue in most organizations is in other people’s inboxes, and it has never appeared on a single report.

Cultivating this: Ten Minutes Today, in Part III.