First, some context, because it helps to understand why this problem feels so much worse than it did a generation ago.
David Allen put his finger on this a quarter century ago in Getting Things Done: work used to have edges. A field was plowed or it wasn’t. The boxes were packed or they weren’t. You could see the work, and you could see when it was done. When your work is visible and finite, you don’t need much of a system. The work itself tells you what’s next.
Most of us don’t do that kind of work anymore. Peter Drucker called what replaced it “knowledge work,” and its defining feature is that it never announces what’s next. Knowledge work has no edges. Every project could be improved a little more. Every inbox refills itself. And the inputs never stop: emails, messages, meetings, requests from every direction, each one carrying a new task or three. You could work every waking hour and not reach the bottom of the list, because there is no bottom.
The cost of sending those inputs collapsed, too, and that changed things a lot as well. When delegating meant writing a memo or walking down the hall, the sender paid a real price to hand you something, enough that they’d pause to consider whether it was worth it, whether it was even your job. That friction helped keep everyone’s responsibilities in their lanes. Now a request costs the sender a few seconds and a keystroke, and if they picked the wrong person, the mistake is nearly free to them. So the requests multiply by a hundred small delegated tasks nobody thought twice about sending.
Here’s the part that matters for our purposes: when the work stopped being visible, the job of deciding what to do next stopped being obvious. That decision used to be free, the work told you. Now it’s a real cognitive task, it repeats dozens of times a day, more often than it used to and in a more dynamic environment than your parents ever navigated, and almost nobody has a good way to handle it. Every failure mode in this chapter is, at its root, a bad answer to that one question.