Most owners who say delegation doesn’t work have done it once, like this.
They hand something over, loosely. It comes back not quite right. They fix it themselves, because that’s faster. The next time, they don’t bother handing it over. And they conclude, reasonably, that it’s quicker to do it themselves.
It is quicker. For this week. Every week, forever.
Two ways to get it wrong
Delegation fails in opposite directions, and most people swing between them.
Abdication: handing something over with no brief, no check-in, and no support, then being disappointed with the result. It feels like trust. It’s actually abandonment.
Micromanagement: handing something over and then supervising every step, so closely that you might as well have done it. It feels like quality control. It teaches the person that their judgement isn’t wanted.
The answer isn’t somewhere in the middle. It’s matching the amount of oversight to the person and the task — and there’s a genuinely useful way to do that.
Task-relevant maturity
Andy Grove ran Intel, and in 1983 he wrote High Output Management, which is still one of the most practical management books available. One of its central ideas settles the micromanagement question.
Grove called it task-relevant maturity. How much supervision someone needs depends on how experienced they are with this particular task — not on how senior they are, how long they’ve worked for you, or how capable they are generally.
A brilliant, experienced person doing something new to them needs close guidance. A junior person who has done this exact thing twenty times needs very little.
That removes most of the awkwardness. Close supervision isn’t a verdict on someone’s ability; it’s appropriate to their experience with the work in front of them. And it’s temporary — as their experience grows, your involvement shrinks. The mistake is supervising everyone the same way, which over-manages your experienced people and under-supports your new ones.
Check early, when it’s cheap
Grove’s second useful principle comes from manufacturing: catch problems at the lowest-value stage possible. A fault found in raw materials costs almost nothing to fix. The same fault found in a finished product costs a great deal.
Delegated work is the same. If you only see the finished proposal, any problem means redoing the whole thing — which is exactly where owners take work back. If you see a one-page outline first, the same problem takes ten minutes to correct.
So build the early check in deliberately. Ask for the outline, the rough draft, the plan before the execution. Review hard at that stage. Then step back and let them finish.
This is the practical difference between checking the work and micromanaging it: one look early, rather than constant looks throughout.
What to hand over
Not just the task. Four things.
What good looks like. The standard you’re holding, stated explicitly. Most delegation fails because the expectation lived in the owner’s head and was never said out loud. Clarity does most of the work here.
The constraints. Budget, deadline, things that can’t be changed, the client who’s particular about something.
The decisions they can make. This is the one owners forget, and it’s the one that matters most. Handing over work without the authority to decide anything just produces someone who asks you more questions. Say explicitly what they can decide alone.
When you’ll check in. Agreed in advance, so neither of you is guessing.
Accept good enough
The hardest part, and the one that makes owners take work back.
Someone else doing your task will rarely do it exactly as you would. At first, it may genuinely be a bit worse. That’s the price of delegation, and it’s worth paying — because 80% of your quality, done by someone else, is infinitely more than 100% of your quality competing for time you don’t have.
The useful question isn’t “is this how I’d have done it?” It’s “is this good enough to meet the standard we agreed?” If yes, it’s done. Let the stylistic differences go.
If the answer is repeatedly no, that’s information: either the brief wasn’t clear, or the person isn’t the right fit for this task yet. Both are fixable. Taking the work back fixes neither.
What to keep
Not everything should be delegated. Keep what genuinely requires you: the relationship only you can hold, the decision that sets direction, the judgement call that defines the business.
And it’s fine to keep one thing you simply love doing, as long as you’re honest that it’s a choice rather than a necessity.
Everything else is a candidate. A useful test: if you were ill for a month, what would you want someone else to already know how to do? Start there.
The first one is the hardest
Your first real delegation will take longer than doing it yourself. Writing the brief, answering questions, reviewing the early draft — it’s more work than the task, the first time.
The second time is quicker. By the fifth, it’s saving you hours. That’s the entire transition from doing the work to running the business, and it only happens if you get through the first few attempts without taking the work back.
