A task lands on your desk that you could finish in forty minutes. Someone on your team would take half a day, do it less well, and need a conversation afterwards. The maths is obvious, so you do it yourself, again. That instinct is exactly how a capable manager becomes a permanent bottleneck and a team stops growing. Delegation looks like a time-management trick. Its bigger payoff is that it is the most reliable way you develop people, but only if you treat the handover as a teaching decision, not a dumping decision.

The quick version

  • Delegation is development, not offloading. The work you give away is the curriculum someone learns from, most managerial growth comes from real assignments, not courses.
  • Match the stretch to the person. How much you hand over should track how ready they are for this task, not how senior they are in general.
  • Pick a level deliberately. Delegation isn't all-or-nothing; it runs on a dial from "do exactly this" to "you own the outcome." Name the level out loud.
  • Protect the right to struggle. The learning lives in the bit you're tempted to rescue. Let people own the outcome, including the wobble.

The idea in depth

Three well-established ideas explain why delegation develops people, and where each one stops being useful.

People learn most from the work itself

The Center for Creative Leadership's long-running Lessons of Experience research is usually summarised as the 70-20-10 model: leaders grow roughly 70% from challenging on-the-job experiences and stretch assignments, 20% from other people (feedback, coaching, mentoring), and only 10% from formal training (CCL, "The 70-20-10 Rule for Leadership Development"). The exact ratio is a rule of thumb rather than a measured constant, treat the numbers as directional, not gospel. But the direction is the point: the steepest learning comes from doing real work with real stakes. As a manager, the supply of those experiences is almost entirely in your gift. Every meaty task you keep is a stretch assignment you've quietly cancelled.

So the move is: when a genuinely developmental piece of work appears, running the review, owning the vendor relationship, presenting to the leadership team, your first question isn't "can I find time to do this?" It's "who would grow from this, and how do I set it up so they can?" Delegate the assignments people learn from, not just the chores you'd rather not do.

How much you hand over should match readiness, for this task

Hersey and Blanchard's Situational Leadership model maps your style to a person's development level on a specific task. At the top sits S4, Delegating: low direction, low hands-on support, used when someone is both able and committed, so you trust them to own it. The trap is treating "delegating" as a fixed compliment you pay to senior people. Readiness is task-specific: a strong engineer running their first hiring loop is a near-beginner at hiring, even if they're an expert at code. Hand them an S4-level "you own it" brief for something they're actually at S1 or S2 on, and you've set them up to flounder. The model's evidence base is thinner than its popularity suggests, researchers have long noted its empirical support is weak, so use it as a lens for the conversation, not a law that picks the answer for you.

So the move is: before you delegate, ask two quick questions about this task, not this person, can they do it, and are they up for it? If yes to both, hand over the outcome and get out of the way. If not, you still delegate, but with more direction, a check-in, or a smaller first slice. The aim is to move them toward "you own it," not to demand it on day one.

Delegation runs on a dial, not a switch

The most useful idea here is the oldest. In 1958, Robert Tannenbaum and Warren Schmidt published "How to Choose a Leadership Pattern" in Harvard Business Review, a continuum of seven positions running from the boss deciding alone to the team deciding within agreed limits. The lasting lesson is that handing over isn't binary. Between "do exactly what I say" and "I never want to hear about it again" there are several deliberate settings: recommend an option for me to approve, decide and tell me before you act, decide and tell me after, decide, you own it. Most botched delegation is really a mismatched dial, the manager thought they'd handed over the decision; the report thought they were gathering options. The work then either stalls or surprises everyone.

So the move is: say the level out loud, in one sentence. "I'd like you to pick the agency and just tell me who you chose" is a different instruction from "shortlist three and we'll decide together," and the difference is the whole game. Naming the level removes the most common cause of delegation going wrong, and lets you turn the dial up next time as the person grows.

flowchart TD
    A(["A task appears"]) --> B{"Would someone
grow from this?"}
    B -->|"No, pure chore"| C(["Delegate to free your time"])
    B -->|"Yes, developmental"| D{"Can they do it,
and are they up for it?"}
    D -->|"Both yes"| E(["Hand over the outcome
S4, you own it"])
    D -->|"Not yet"| F(["Delegate with more direction:
smaller slice, a check-in,
recommend-then-approve"])
    F --> G(["Turn the dial up
next time"])
Delegation as a teaching decision, not a dumping decision. Leaders Loop

The evidence that it pays, and the honest caveat

The business case is real but worth reading carefully. A 2015 Gallup study of 143 Inc. 500 CEOs found that those with high "Delegator" talent posted an average three-year growth rate 112 percentage points higher than peers with low delegator talent, and generated 33% more revenue (Badal & Ott, Gallup, 2015). That's a striking number, but it's a correlation among fast-growing founders, not proof that delegating causes growth, and CEOs aren't the same population as line managers. Read it as strong supporting evidence, not a guarantee. The harder problem is that managers know all this and still don't do it. A figure often quoted from London Business School's John Hunt, that only about 30% of managers think they delegate well, and barely one in ten does it well in their staff's eyes, is repeated everywhere (popularised by Martin Zwilling in Inc.) but, as far as we can trace, never tied to a primary study; treat it as illustrative, not a measured fact. The credible, current diagnosis comes from MIT Sloan's Elsbeth Johnson, writing in HBR (2025): managers stay "mired in the details" of their teams' work, with real costs to the organisation. The block is rarely skill. It's the forty-minute maths at the top of this piece.

A worked example

Illustrative scenario; figures are made up to show the mechanics.

Priya runs a six-person marketing team. Every quarter she builds the campaign report herself, three hours she resents, because she's faster at it than anyone. Dan, eighteen months in, is sharp and keen but has never owned anything that goes to the leadership team.

The offloading move, "Dan, can you pull the Q2 numbers into the deck by Thursday?", is a chore handover. Dan does data entry, learns almost nothing, and Priya still writes the story.

The development move starts with the dial. Dan can handle the mechanics (able) and wants the visibility (committed), but he's never framed a narrative for executives, so on that part he's a beginner. So Priya splits it: "You own the whole Q2 report, analysis, story, and recommendation. The one bit we'll do together is the three headline messages for the exec audience, because that's new for you. Bring me your draft messages Tuesday; everything else, just run it." She's set an S4 "you own it" level on the body of the work and a lower, supported level on the one slice that's a genuine stretch, and said both out loud.

Tuesday, Dan's headline messages are weak. The offloading instinct is to fix them. Instead Priya asks three questions: who's in the room, what decision do they need to make, what would make them act? Dan rewrites them himself. The report is rougher than Priya's would have been, and it's the first thing Dan has ever fully owned, so the next one needs half the scaffolding. That gap between "rougher than mine" and "his own" is the entire return on the investment.

flowchart LR
    A(["You do it
every time"]) --> B(["Recommend
options, you approve"])
    B --> C(["Decide, tell you
before acting"])
    C --> D(["Decide, tell you
after"])
    D --> E(["You own
the outcome"])
The delegation dial, after Tannenbaum & Schmidt, turn it up as readiness grows. Leaders Loop

Frequently asked questions

Isn't it faster and safer to just do it myself?

For one task, almost always. Across a quarter, no. Every time you take back the developmental work, you keep the team dependent on you and cap how big you can ever be, trading three hours saved this week for a report you'll still be writing in two years. Delegation is an investment with a slow payback, which is exactly why it loses to the urgent forty-minute task, and exactly why it matters.

What if they do it badly and it's client-facing?

Match the dial to the stakes, not just the person. High-stakes, irreversible work gets a lower setting, "bring me your draft before it goes out", while the person still owns the thinking. The choice was never control versus abandonment. It's where on the dial to sit, and you move it up as trust builds on that kind of task. This is where the situational lens earns its keep: more support where they're genuinely new, less where they've proven themselves.

How is this different from coaching or just giving feedback?

Delegation creates the experience; coaching and feedback help someone extract the lesson from it. In 70-20-10 terms, the assignment is the 70% and your conversations around it are the 20%, they compound. The handover without the follow-up is just abandonment with extra steps; the feedback without a real stretch assignment is advice with nowhere to land. You need both. (See Coaching skills for the asking-not-telling half of this.)

I delegated and it still went wrong. What happened?

Usually one of three things: the dial was never named, so they guessed the level; the stretch was too big with no support, so "you own it" meant "you're on your own"; or you rescued at the first wobble and taught them you'd always take it back. Name the level, size the stretch to readiness, and hold your nerve through the messy middle.

Who should get the stretch when everyone's already busy?

Give developmental assignments to the people whose growth you most need next, usually your potential successors and your quietly capable mid-team, not only your obvious stars. Spreading stretch deliberately is how you build bench strength instead of a single point of failure. (More in Developing future leaders & successors.)

Related in the Toolkit

Where to go next