Delegation: levels of delegation and how to hand over work well

Most managers know they should delegate more. Many still don’t, or they hand work over in a way that brings it straight back: a vague request, no clear authority, then a surprise when the result isn’t what they pictured. The fix is to treat delegation as a set of choices you make deliberately. What exactly is being handed over, how much decision-making power goes with it, and how will you both know it’s on track?

This guide sets out a simple scale of delegation levels, a briefing template, and a way to follow up without hovering.

What you hand over, and what you keep

Wikipedia’s article on delegation separates three ideas that are easy to blur. Responsibility is the work itself. Authority is the power to make decisions and use resources to do it. Accountability is answering for the result. The article makes two points worth remembering: responsibility without matching authority is meaningless, and the manager remains answerable for the outcome even after handing the work over.

So when you delegate, you are not getting rid of a problem. You are sharing the work and some of the power, while staying responsible for making sure it succeeds.

The levels of delegation

Delegation isn’t on or off. Between “do exactly as I say” and “it’s your call, tell me if you need me” there are several steps. Management 3.0 sets out seven levels in its Delegation Poker practice:

The value of the scale is that it makes the level explicit. Much friction in teams comes from a manager thinking they are at level 3 while the team member assumes level 6. Management 3.0 suggests a Delegation Board: a grid with key decision areas down one side and the seven levels across the top, so everyone can see who decides what. Levels apply to decisions, not people. The same person might be at level 6 for booking travel and level 3 for hiring.

Matching the level to the person and the task

How do you choose a level? Situational leadership offers some context. Paul Hersey and Ken Blanchard first set out the idea in 1969, and as Wikipedia’s summary of situational leadership theory explains, it proposes four styles (telling, selling, participating and delegating) matched to how able and willing a person is for a specific task. Someone new to a task needs more direction. Someone skilled and committed needs little more than the handover.

The model is widely taught, but the same article notes that several studies don’t support all of its prescriptions. Use it as a prompt, not a formula. Practical questions work just as well:

• Has this person done something like this before, and how did it go?
• What is the cost if it goes wrong, and can it be undone?
• How much do they want the work, and do they have the time?
• Will this decision affect people outside the team?

Start one level lower than you think the person can handle for a high-risk task, and one level higher for a low-risk one. Then move the level up as trust builds.

A briefing template

Jesse Sostrin, writing in Harvard Business Review, recommends explaining why the work matters, checking that your expectations have landed, and asking the other person how much involvement they want from you. A short written brief covers all three. Use these headings:

1. Outcome: what result you need, in one or two sentences.
2. Why it matters: who it’s for and what it feeds into.
3. What good looks like: the standard, with an example if one exists.
4. Constraints: deadline, budget, brand or legal rules, things not to change.
5. Authority: the delegation level, and what they can decide or spend without asking.
6. Resources: people, documents, access and tools they can use.
7. Check-ins: when you’ll talk, and what you’ll look at.
8. Escalate if: the specific situations where they should come back to you straight away.

Finish the conversation by asking them to play the task back in their own words. If their version doesn’t match yours, fix the brief now rather than at the deadline.

Following up without taking over

Follow-up is part of delegating, not a sign of distrust. The trick is to agree it in advance so it doesn’t feel like checking up.

• Set check-ins by risk and level. At levels 3 and 4 you might meet weekly. At level 6 a short note at the end may be enough.
• Ask questions before giving answers. “What options have you looked at?” keeps the decision with them.
• Resist taking the work back when it gets hard. If they bring a problem, ask what they recommend.
• Review at the end: what went well, what you’d change, and whether the level can go up next time.
• Give credit publicly when it works.

An example (illustration only)

A marketing manager wants to hand the monthly customer newsletter to a junior colleague, Priya, who has written social posts but never a newsletter. The manager sets levels for each decision instead of one level for the whole job.

Discount offers stay low because they affect margin and other teams. Everything else moves up after three issues that hit their deadlines and open-rate targets. The brief names two escalation triggers: any legal or complaint issue, and anything that would delay the send by more than a day.

Common mistakes

• Delegating tasks but not authority, so every small choice still comes back to you.
• Leaving the level unstated and then overruling a decision you had implied was theirs.
• Handing over only the dull work, never anything that helps the person grow.
• Giving a deadline but no picture of what good looks like.
• Checking in so often that the person stops making decisions.
• Never moving the level up, even after the person has proved they can handle more.

Delegate when the work could be done well by someone else, when it would help them develop, or when it frees you for work only you can do. Keep the work, for now, when the risk is high and nobody on the team is ready yet. But set a date to revisit that, or nothing will ever change.