You hired someone, told them "you own sales now," and finally exhaled. A month later you're up at midnight rewriting their proposals because clients keep leaving, and they honestly don't understand what they did wrong.
How do you delegate effectively? Hand over the outcome, the right to decide how to get there, and the duty to fix what goes wrong. A to-do list is optional after that. Delegation works when four conditions are met:
Miss any one of them and within a month the work lands back on your desk, now with a salary attached.
Over fourteen years of working with entrepreneurs I've done more than 2,000 interviews about what actually helped them grow, and delegation is one of the most common places people get stuck. Business owners know how to hand out tasks. What they keep is the fixing.
Delegating responsibility means giving someone ownership of the end result, the freedom to choose how to reach it, and the obligation to fix whatever goes wrong. A task describes an action ("call this list"). Responsibility describes an outcome ("ten paid orders by Friday, and if they're not there, you know why and what you're doing next").
I explain it with pizza. If I ask a cook at a restaurant to make me a pizza, I'm delegating responsibility for that pizza being ready and being good. If he knows I'm going to double-check it, fix it and redo it anyway, he gets sloppy, because Igor has his back. The whole difference between a task and responsibility lives in the doer's head, in the thought "if this fails, I'm the one fixing it."
That gives me a simple test I use with clients. Saying "you're responsible for this" doesn't make anyone responsible. Responsibility belongs to whoever cleans up the mess afterwards. If your kid didn't clean their room and you ended up doing it, the only thing you delegated was your grumbling. Same with employees: if you quietly sit down and fix their mistake, the job is still yours, except now you're also paying a salary for it.
Now the main objection I hear from managers, and the one every leadership blog repeats: "you can't delegate accountability, the owner always answers to clients, the bank and the tax office." That's true, and I don't argue with it. Responsibility comes in two layers. The outer layer, answering to the market, stays with the owner forever; you took it on the day you started the business. The inner layer, fixing the mistake and understanding its cause, goes to the person doing the work. When I say "delegate responsibility," I mean the inner layer.
In 1974 William Oncken and Donald Wass described the same mechanics in the Harvard Business Review article "Management Time: Who's Got the Monkey?" (HBR reprinted it as a classic in 1999). The "monkey" is the next move on a problem. An employee comes with a question, the manager says "let me think about it," and the monkey jumps from the subordinate's back onto the boss's. Their rule is simple: after any conversation, the next move stays with the employee, and they come back with it at an agreed time. The manager helps, but never takes the monkey home.
Almost everyone finds delegation hard. Gallup assessed 143 CEOs from the Inc. 500, the fastest-growing private companies in the US, and found strong delegator talent in only one in four. Their companies grew 112 percentage points more over three years, and their average 2013 revenue was about $8 million versus $6 million for the rest. It's a correlation inside an already elite sample, and "talent" here is Gallup's psychometric assessment. Still, even among the best, the revenue gap is about a third.
The second obstacle is guilt. Akinola, Martin and Phillips (Academy of Management Journal, 2018) showed across five studies that women leaders associate delegation with guilt more often and delegate less, and that framing delegation as a contribution to people's growth narrows the gap. That's the study's gender finding. My own observation is separate: I see the guilt in people of any gender, especially in coaches and therapists. "It feels awkward to ask." "He's already swamped." "Easier to do it myself."
The third obstacle I call the "nobody does it better than me" glitch. It feels good, because it gives you permission to be proud of your exhaustion. I wrote about it in detail in my piece on the fear of hiring your first employee. I had clients from Uzhhorod, Ukraine, who ran a visa agency. When they came to me in November, they were making $7,000 a month.
The owner did everything himself. I asked him to set up a simple sales report, and it turned out all the sales were made by one person, him, while the managers did almost nothing. I said: "Forward the line, switch off your phone, let the managers handle the leads." It took me three weeks to talk him into simply switching off his phone; he felt he'd lose control. Today the agency makes about €44,000 a month. That growth came from more than one change, and I won't pretend a switched-off phone did it all. But as long as every call went to the owner, the managers had nothing to do.
And the fourth, which people talk about least: money in the owner's head. If you don't pay yourself a proper salary for the role you're actually filling, your brain keeps thinking "I'm saving money by doing it myself." If someone you hired for $1,800 a month were doing that work, you'd see that you can give that money to them and sell your own hours for more.
First a distinction, without which nothing else works. Delegating is handing someone work you used to do yourself. Assigning is putting someone on a task that nobody did before. Delegating takes load off you; assigning adds a new function you now have to oversee. Hence my golden rule: don't add new assignments until you've delegated. Want to start a YouTube channel or build a funnel? First take something off your plate. Otherwise you get a familiar cycle: you hire a specialist, you don't understand the area, you can't check the work, a month later there are no results, you fire them and decide "I tried, it doesn't work."
The second trap is delegating chaos. If the process isn't assembled in your own head, if you don't know its stages or what "good" looks like, then what you're handing over is the right to decide how everything should work. To a person or to an AI agent, it doesn't matter. When it starts glitching, you can't tell where it broke, because you didn't understand the system before you handed it off. It's the same rule I described in why AI implementation fails in small business.
What should you hand off first? Write down everything you did this week and find the repeating area that eats the most hours; hire for that, and the person pays off fastest. Sales come with a caveat. If you're an expert and your sales aren't working yet, delegate them last: until you've learned to sell your own product, there's nobody to pass it to, and for an expert the technical side of the product goes first. If sales are already flowing and depend only on you, like at the visa agency, that's exactly when to hand them off, or you become the ceiling of your own company.
What should a leader never delegate, at least until the decision rule is written down? The decisions that set the company's direction: its goals, its values, who becomes a key person, where the big money goes. Goals for a single function can go to that function's lead; goals for the company stay with the owner. You can prepare these decisions with your team and gather data, but the final word is yours until you've described how you make them (more on that below). And one more thing: give access to money and the client base with limits. That's insurance, and it has nothing to do with distrust.
Back in 2019 I described on my channel how I move a business onto autopilot. I list every function of the company (marketing, sales, ideology, hiring, purchasing and so on) in a table and give each one a label. That's the scale that shows what you've really delegated and what only looks delegated.
| Level | What it means | Who fixes the mistakes |
|---|---|---|
| Nobody does it | The function is needed but doesn't exist | Nobody, so the business loses money quietly |
| I do it | The function lives in your head and hands | You |
| I hand off execution | Someone follows instructions, you make the decisions | Mostly you |
| I delegate management | They run the function and its people, you see results in numbers | They do, you read the report |
| I delegate development | They improve the function and set its goals | They do, you join on strategy |
Truth be told, the third row isn't delegation in the sense of this article yet, it's handing over hands. Responsibility moves to the person from the "management" level up. The owner's job is to gradually move functions into the bottom two rows. Not necessarily all the same way: some areas you can hand off all the way to development, and in others you might keep development for yourself and delegate only management.
Inside the scale there's a distinction that answers half of the "why isn't this working for me" questions. Functions are either practical (what to do with your hands) or mental (how to make decisions). In one of my first businesses I got tired of training every new hire from scratch, so I wrote the most meticulous instruction manual I could. It had everything: when a client comes downstairs, you stand up from your desk and smile; if you recognize them, you say "Welcome back!"; if you don't, you ask "Is this your first time with us?" It was so detailed that nobody could come to me with a question.
But it wasn't enough, because the manual had no moment of decision. So after the practical functions I move on to the mental ones and describe how and when I make decisions: the algorithm I use to hire someone, how I decide whether to work with a partner. In effect, I write down how I think. Only once the moment of decision is delegated can I take my focus off the function and know it works. Sometimes better than with me.
Quality gets lost in two places: the person doesn't know what "good" looks like, and you don't give them time to grow into it. The first is fixed with a standard. Show a finished example and name the acceptance criteria in plain words: what a proposal must contain, how many hours we take to reply to a client, what return rate is normal. Then ask the person to retell the task in their own words. If their version doesn't match the one in your head, the problem is in how you set the task, and the doer has nothing to do with it.
The second is fixed with a step-by-step handoff. First I do it, you watch. Then you do it, I watch next to you. Then you do it alone and I only check the result. At the start, quality almost always dips below your level, so agree with yourself in advance on what result you'll accept in the first month. If you expect your own level right away, you'll take the function back in week two.
And if the person still can't handle it after a handoff like that, the order is straightforward: check the instructions and the standard first, then the training, and only then the person. I wrote more about choosing people in the piece on the fear of hiring, and about building a first team in the article on first hires and delegation.
Control and micromanagement get confused, which is why some managers check every step and others don't check at all. Healthy control rests on three things you write down for any role: the purpose of the role (why it exists), the result of the role (what the person must deliver in a month) and the metrics you use to measure it. If you don't measure an employee's result, you can't tell whether they pay off, and they can't see whether things got better or worse. I explained why you can only manage what you measure in a separate article.
The most dangerous mistake here is mundane: set a task, promise to check, and don't check. You said "have it done by Wednesday," and on Wednesday you didn't ask. After a couple of Wednesdays like that, people learn that deadlines mean nothing here, and discipline falls apart. The cure comes from the monkey article: the person doing the work brings the status. On Wednesday they send the report themselves, you just read it, and if there's no report, that's your reason to talk.
One more point for those who work themselves into the ground "because I'm responsible for everything." Being responsible for the whole ship and answering the phone every minute are different things. The outer layer of responsibility stays with you, but how you carry it is your choice. If your method is to personally plug every hole, the problem is in your management method. Sometimes one message that sets the team in motion replaces ten hours of your running around.
A telling case from this fall happened to a member of our club whom I once taught from scratch. She runs a small bath bomb production and needed to be away for a few days, the first time she'd tried managing the company remotely. She left the manager rules, and the staff were warned in person: everyone sends a message when they arrive at work, at the end we count how much was made, and pay is based on that. Pay was per shift.
She came back, and production had barely run while she was gone. A classic dilemma. Say "you're not getting paid," and tomorrow nobody shows up. Pay as if nothing happened, and everyone learns that this is fine. Usually the owner picks one of two evils and then cleans up the consequences herself.
We worked out a third way. First, confirm the fact: "Were you warned about the rules?" They were. Then call things by their name, calmly and face to face: in fact, you ignored my request, and we both know it. Then a decision in which the consequences are theirs. Pay for the disputed days is postponed until the end of the month. If everyone works honestly and logs every shift for the whole month, she accepts that as proof that people worked on those days too and pays everything. If not, the money for the disputed days is lost. Now it pays to work properly, and the condition was stated openly and accepted.
A caveat you can't skip when applying this case: the scheme fits piece-rate or per-shift pay tied to output. If your people are on regular employment contracts, withholding pay for time worked is illegal in most countries, and consequences should be built through the bonus part. Check the mechanics with a lawyer before you announce anything.
The manager's job in stories like this is always the same: figure out how to make the people you delegated responsibility to deal with the consequences. If you always cover for them, they calmly shift responsibility back onto you: "Igor's behind me, he'll fix it." When it's clear nobody will rescue them, they have to grow up. That's why most people who say "I delegated responsibility" actually delegated a task without responsibility, then clean everything up themselves, get irritated and fire people.
If more than half of the functions in your table sit in the "I do it" row, that's about your stage of business; lazy employees are secondary. I wrote about how that stage changes as you grow in the article on the stages of business growth.
A task is an action you hand over; responsibility is the result, plus the right to decide how to get it and the duty to fix mistakes. The test is simple: if you're the one who fixes things after an error, the responsibility stayed with you. Answering to clients and the market always remains with the owner.
Most often it's the feeling that nobody does it better, guilt about loading people up, and fear of losing control. Gallup found strong delegator talent in only one in four CEOs of fast-growing companies. The fear eases when the function has instructions, a measurable result and regular check-ins.
The decisions that set direction: company goals, values, key people and big spending. You prepare them with your team, but the final word stays with the owner until the decision rule is written down. And never delegate what you don't understand yourself, because that's handing over chaos.
Control the result, not each step. Write down the role's purpose, monthly result and metrics, and have the person bring their status on agreed days. If you promised to check on Wednesday, check on Wednesday, or people quickly learn that deadlines mean nothing.
At the start, that's normal. Show an example and acceptance criteria, hand off in steps, and decide in advance what level you'll accept in the first month. If quality still doesn't improve, check your instructions and training first, then the person.
In my practice there are five: nobody does it, I do it, I hand off execution, I delegate management, I delegate development. Real responsibility moves to the person from the management level; below that you're still the one fixing mistakes.
A 5-day game for experts — from zero to your first $1,500
Start free →5 days — how a practitioner and entrepreneur puts AI to work
Learn more →Build a producer team around an expert in 30 days
Learn more →A private community of entrepreneurs — growth without the operational grind
Learn more →In the free version of the Alliance you take a level test, and the system builds a route for your stage: what to hand off first, how to set goals for a role and where to start your team.
Take the test