You gave a task away. You remember the conversation, you remember feeling relieved, and here it is again on your desk just a few weeks later.
Most owners read this as evidence about their people. It's usually evidence about the handover. Work returns for a small number of reasons, all of them fixable, and none of them are that your team is incapable.
Start with what's actually happening. Nobody in your company sat down and decided to escalate everything to you. They learned it. Every time something important eventually landed on your desk anyway, the whole team took a quiet lesson: the real decision happens over there. After enough repetitions, escalating stops being a choice and becomes the culture. People will tell you they want more autonomy and then check with you anyway, and both of those things are true at once.
So the question isn't whether your team is willing. It's what you have been teaching them.
Reason one: you handed over the work but not the standard
You know what good looks like here because you have done it several hundred times. They have not. When the work comes back at seventy percent, the missing thirty is almost never effort. It's a specification nobody ever wrote down, which lived entirely in your head, and against which their work is now being judged.
The fix is unglamorous and takes about fifteen minutes. Before handing something over, give them three things: an example of the finished article you were happy with, the two or three elements it must contain to be acceptable, and a plain statement of what "done" means. Done can mean sent to the client. It can mean on your desk for review. It cannot mean whatever they assume it means.
This is the cheapest intervention available to you and the one most often skipped, because explaining the standard feels like a delay when you are busy and the work is already late.
Reason two: nobody drew the decision rights
People escalate because they don't know where their authority stops. In the absence of a line, checking with you is the safe move, and safe moves are what people default to.
Draw the line crudely and out loud. Three tiers is enough to cover 99% of decisions being made:
- Decisions they make alone, and you find out later or not at all.
- Decisions they make and then tell you, so you're informed but not consulted.
- Decisions they bring to you before acting.
Write it once per role. Put a number on it where you can, because "small purchases" means something different to each of you and a figure removes the argument. Then hold to it, including on the days when you would have decided differently. The first time you overrule a decision that was legitimately theirs, the line moves back to where it was.
Worth revisiting the tiers every few months. If someone has been making tier-three calls well, they should be tier two by now. A line that never moves teaches people that growing their authority isn't on offer.
Reason three: you took it back
This is the expensive one, and the one owners rarely see themselves doing.
The work comes back at seventy percent. You're busy, the client is waiting, and you don't want to demoralize someone who tried. So you fix it yourself, quietly, and say nothing. It feels like the generous choice. From where they sit, the message is that the work was fine, and then some invisible thing happened, and now they're not sure whether it was theirs at all. Next time, they hand it over earlier and less finished, because that's what your behavior taught them to do.
If the work wasn't good enough, say so and send it back with the specific gap named. That conversation takes ten uncomfortable minutes and buys you a permanent improvement. The silent redo takes an hour and buys you nothing.
There's a harder version of this, and it's the one that changes companies. For most of what you've delegated, eighty-five percent of your standard is genuinely fine. Not for everything. Not for the work your reputation rests on. But for most of it, the difference between your version and theirs is invisible to the client and expensive to you. Deciding which work needs your standard and which does not is one of the few decisions that actually gives you your week back.
The complication nobody mentions
In a lot of owner-run businesses, letting go carries real commercial risk, and the advice above can sound naive if you're the one holding the relationships.
You may be the reason the biggest account stayed. The client may expect to see you, specifically, and may say so. Handing that over is not a matter of trusting your team more. It's a genuine risk to revenue, and pretending otherwise helps nobody.
Two things are worth saying about it. The first is that "I can't release this account" is a much narrower statement than "I can't release anything," and the second sentence is the one most owners are living by. Look at what actually sits on your desk. The share of it that is genuinely relationship-critical is usually smaller than it feels.
The second is that even the relationship-critical work can be handed over in stages, over quarters rather than weeks. They sit in on the call. They run one part of it while you're there. They run the whole call with you present and silent, which is harder for you than for them. Then they run it. The client meets your successor gradually, in your company, with your endorsement visible, which is a very different experience from being handed to a stranger after you've burned out.
What to do this week
One thing. Not a system.
Pick something that came back to your desk after you had given it away. Work out which of the three reasons it was: the standard was never explicit, the decision rights were never drawn, or you took it back. Then fix that one instance properly, with the person, in a conversation.
You'll usually find the same reason accounts for most of the pile.
This is one spoke of the leadership bench, our model for building a leadership pipeline in a small company. See also five ways to give someone leadership reps before you give them the title and making the bench a system, not a favor.