Deadlines stimulate progress. But only if they are commitments. To hit a deadline means achieving the objective with absolutely A-level work, absolutely complete, absolutely on time, absolutely without complaint, absolutely. If you establish deadlines that everyone knows will slip, then you have no deadlines.

In a culture of discipline, there are only two acceptable ways to miss a deadline. First, the person to whom you have committed initiates a change in the deadline, without your having to ask (explicit, unsolicited absolution). Or second, you’re truly incapacitated by something that has happened to you or your loved ones (disease, accident, tragedy), and it would be inhumane to hold you to the deadline.

Paradoxically, then, the more your high-priority interrupts involve catching your people doing things wrong (so you can fix them), the less productive each person will become in the short term, and the less growth you’ll see from your team members in the long run. Finding itself in negative-criticism territory, the human brain stiffens, tenses, and—in meaningful ways—resists improvement. Machines and processes don’t do that. You can fix a machine, you can fix a process, but you can’t fix a person in the same way—people aren’t toasters. So, when it comes to your people, what should be your high-priority interrupt? If what you want is improvement, then it should be whenever someone on your team does something that really works. The goal is to consciously spend your days alert for those times when someone on your team does something so easily and effectively that it rocks you, just a little, and then to find a way of telling that person what you just saw.

There is a crucial yet hard-to-understand concept here. Most people grasp the need to set priorities; they put the biggest problems at the top, with smaller problems beneath them. There are simply too many small problems to consider them all. So they draw a horizontal line beneath which they will not tread, directing all their energies to those above the line. I believe there is another approach: If we allow more people to solve problems without permission, and if we tolerate (and don’t vilify) their mistakes, then we enable a much larger set of problems to be addressed. When a random problem pops up in this scenario, it causes no panic, because the threat of failure has been defanged. The individual or the organization responds with its best thinking, because the organization is not frozen, fearful, waiting for approval. Mistakes will still be made, but in my experience, they are fewer and farther between and are caught at an earlier stage.

Anytime somebody goes two days without reporting a constraint, you can bet there’s a bigger problem lurking. Busy, productive people who are doing anything of consequence get stuck pretty regularly.

Paradoxically, then, the more your high-priority interrupts involve catching your people doing things wrong (so you can fix them), the less productive each person will become in the short term, and the less growth you'll see from your team members in the long run. Finding itself in negative-criticism territory, the human brain stiffens, tenses, and—in meaningful ways—resists improvement. Machines and processes don't do that. You can fix a machine, you can fix a process, but you can't fix a person in the same way—people aren't toasters. So, when it comes to your people, what should be your high-priority interrupt? If what you want is improvement, then it should be whenever someone on your team does something that really works. The goal is to consciously spend your days alert for those times when someone on your team does something so easily and effectively that it rocks you, just a little, and then to find a way of telling that person what you just saw.