At some point in the last year, I had to stop doing a lot of things. Not all of them at once, but enough that my role shifted. I went from making things directly to coordinating things indirectly. From builder to something else — overseer feels too removed, coordinator too passive. What I actually do now is delegate. And stay sane while I do it.
The hard part isn’t the letting go.
The illusion I had to unlearn
I used to think delegation was clean. You identify something you’re doing. You find someone capable. You hand it over. You step back. That’s trust.
I’d read enough leadership books to know that micromanaging kills motivation. That teams need autonomy. That your job as you grow is to get out of the way.
So when I started delegating real work — systems that mattered, decisions that had weight — I tried to practice what I’d read. I’d hand something off with clear context, then I’d disappear.
I’d wait for updates.
The updates would come, but something about them always felt like I was reading a map of a place I used to know. Things had shifted. Details I thought were fixed had moved. The direction felt right in broad strokes but slightly off in ways I couldn’t name without being back in the room.
I’d ask a question. Get an answer. Realize I needed three more pieces of context to make sense of it.
The person I’d delegated to would get tired of questions. I’d get tired of not understanding. Both of us would start operating with less transparency, more assumption.
That’s not delegation. That’s friction dressed up as trust.
What stepping back actually costs
The belief that you should completely disengage from delegated work is seductive because it sounds mature. It sounds like confidence. It sounds like you’ve moved on to bigger things.
What it actually costs is your ability to stay calibrated.
I had one situation where I handed off a project to someone I trusted completely. Clear brief. Strong track record. I figured my job was done.
Six weeks in, I checked in. Something had shifted. Not in the person’s competence — they were executing well. But somewhere in the translation from my intent to their implementation, there was a divergence. Small enough that catching it early was trivial. Significant enough that another six weeks would have baked it into the foundation.
I asked how it happened. The answer was simple: they hadn’t heard from me, so they’d made a reasonable assumption about what I cared about. It was reasonable. It was wrong for what I was actually trying to do. But how would they know?
The problem isn’t that people drift when you’re not watching. The problem is that small misalignments compound silently until you can’t fix them without rework.
And the worst part? The person being delegated to usually assumes that no news means you’ve moved on. So they stop signaling uncertainty. They stop asking whether their read matches yours. They’re trying to prove they don’t need you.
Which means the exact moment you most need visibility — when there’s drift forming — is when you’re least likely to get it.
The false choice
For a long time I thought there were only two ways to delegate:
Option one: You let go completely. You trust. You don’t ask questions. You accept that the work is no longer yours and neither is the outcome if it goes sideways. You prove you’re not a control freak.
Option two: You stay involved. You ask regular questions. You review work. You jump in when something doesn’t look right. You prove you care about the outcome.
But option one leaves you blind and option two leaves you exhausted. Both feel like they have to be true.
I’ve found there’s a third thing that actually works: small, regular calibration. Not surveillance. Not disengagement. Navigation.
What calibration looks like
The first thing I had to unlearn was that asking questions meant I didn’t trust the person.
That’s backwards.
What actually erodes trust is silence. Silence makes people paranoid. Silence is when they start guessing at what you want instead of knowing it. Silence is when small misalignments become big ones without anyone noticing.
What builds trust is informed presence.
So I started a pattern that sounds almost embarrassingly simple: one specific question, every week or two, to the person who has the delegated work.
Not a status update. Not “how’s it going?” — that’s too vague and too easy to answer with empty reassurance.
A specific question about the thing itself:
- “How is this going versus what you expected?” (signals: drift, if it exists)
- “What part of the original intent is hitting the most friction?” (signals: which assumptions were wrong)
- “What would make you uncertain about the direction right now?” (signals: their doubts before they become problems)
That’s it. One question. Five-minute conversation. Then I’m gone again.
What surprised me is that it completely changes the dynamic. The person being delegated to knows I’m informed. They know I’m thinking about it. They know I care enough to check in. And because the question is specific, they can’t pad the answer with reassurance — they have to be honest about where the real pressure points are.
And I get what I actually need: early warning on divergence, not after-the-fact course correction.
Why this works
The reason I had to invent this is that I couldn’t make the traditional two options work. Either I was too far away and didn’t understand what was happening, or I was too close and the person felt like they couldn’t make a move without checking with me first.
Regular small calibration sits in the middle. It’s not surveillance, it’s navigation. You’re not looking over their shoulder. You’re getting a bearing check every week.
The other thing it does is prevent a failure mode I’d watched happen in other organizations: the delegated person gets comfortable, assumes they’ve got the full picture, and makes decisions that don’t account for what’s changed in the broader context since they took over the work.
I can usually see that from the outside. They can’t. But I can’t see it without asking.
The uncomfortable part
Sometimes the calibration question surfaces something harder to sit with: the original intent was wrong.
Not wrong in a way that’s anyone’s fault. Wrong because the thing changed, or the constraints shifted, or the person doing the work understands the territory better than I do from my more remote perch.
That’s the moment where staying connected actually matters.
Because if I’m not there — if I’ve disappeared completely — I never hear that feedback. The person just tries to make their interpretation work. They spend energy fitting the work into my original intent rather than telling me my intent doesn’t fit anymore.
That’s waste. That’s the cost of complete delegation without calibration.
But it’s also the moment where I have to be honest about what delegation actually means. It doesn’t mean the work stays exactly as I envisioned it. It means I’ve chosen someone who can make good decisions in the territory, and my job is to stay informed enough to recognize when those decisions should shift the direction, not to enforce the original plan.
What I’m doing now
I still believe in delegation. I still step back. I still want people to have autonomy and ownership.
But I do it with one small addition to the handoff: a standing calibration question.
Not forever. Just until the pattern is stable enough that I’m confident in the drift detection. Then usually I can ease back to quarterly or even less frequent check-ins.
But in the early stages, when the work is getting its shape, when assumptions are still testing themselves against reality — that’s when the small regular question is worth more than all the trust speeches I could make.
The hardest part of delegation isn’t letting go.
It’s being informed enough while you’re let go that you can recognize when the direction is changing before it’s too late to change it back.