A long row of gold wall pegs where one peg sags under a heap of bone-colored canvas backpacks while every other peg hangs empty, under a small gold sun, in gold and bone woodcut style on a deep umber ground.

K7 Insights

Who's Got the Monkey? Why Every Problem Climbs to You

Part of San Diego Business Coaching

Key takeaways

  • A monkey is the next move on a problem. Say "let me think about it" and the next move is yours. In an owner-run company there's no middle layer to stop it, so every level's monkeys end up on you.
  • Only three things should climb to the owner: losing money past a limit you've set, losing a person, and a breach. Everything else belongs to whoever owns the work, and you see it in the weekly numbers.
  • Turn "losing money" into numbers with a decision-rights table: every decision that climbs gets an owner, a dollar or hour limit, and a rule for when it comes to you.
  • How you react to the first wrong call decides whether anyone makes a second one without you.

“Who’s got the monkey?” asks whose next move a problem is, and it comes from a 1974 Harvard Business Review article by William Oncken Jr. and Donald L. Wass. They showed how managers end up holding their team’s next moves every time they say “let me think about it.” In a big company those monkeys climb one level. In a company of 25 or 60 people they keep climbing until they reach the owner. This version is for the owner, which is who I coach: why every problem lands on you, the escalation rule I give owners, and a decision-rights table with dollar limits you can copy. (If you got here while sizing up business coaching in San Diego, this is a fair sample of how I work.)

What does “who’s got the monkey” mean?

A monkey is the next move on a problem, and whoever has to make that move has the monkey. That’s the core of Oncken and Wass’s “Management Time: Who’s Got the Monkey?”, first published in the November-December 1974 issue of Harvard Business Review and reissued as an HBR Classic in 1999 with a commentary by Stephen R. Covey. The editor’s note on that 1999 reissue says the article “has been one of the publication’s two best-selling reprints ever.”

Here’s the scene they open with, in my words. A manager is walking down the hall and an employee stops him: we’ve got a problem. He knows enough to get involved and not enough to decide on the spot, so he says he’ll think about it and get back to her. Before that conversation the next move was hers. After it, the next move is his. The monkey jumped. The authors’ sharpest point is what that does to the relationship. The boss now has a deadline he didn’t set, and the employee gets to stop by later and ask how it’s coming. He works for her now.

Then they do the math. A manager with four direct reports who each hand him three monkeys a day is holding 60 by Friday. He can’t feed them, because his own boss and his peers already own most of his week, so the monkeys sit. His team waits on him while he takes their work home.

The fix has two parts. First, every problem leaves the manager’s office with a level of initiative attached. They name five:

  1. Wait until told.
  2. Ask what to do.
  3. Recommend, then act on the answer.
  4. Act, then tell the boss right away.
  5. Act, then report on the normal schedule.

For a manager’s own people, the authors take levels 1 and 2 off the table entirely. Second, every monkey gets a next check-in on the calendar. Their five rules for the care and feeding of monkeys come down to this: feed it or kill it, never carry more than you have time to feed, feed by appointment only, do it face to face or by phone (a memo puts the next move back on you), and give every monkey a next feeding time and an initiative level.

It’s a great article, and you should read it. It also assumes a manager in the middle of a big org chart, with a boss of his own to push back on. If you own the company, there’s nobody above you to push back on.

Why do all the monkeys end up on the owner?

Work on the business, not in it. Fine. Doing what, on Tuesday at nine?

Your biggest customer calls with a problem. Whose phone rings?

Result on screen. Map by email.

Because in an owner-run company every path up the org chart ends at the same desk, and the person sitting there is usually the best problem-solver in the building. In Oncken’s corporation a monkey climbs one rung and stops at a manager who answers to someone else. In your company it keeps going.

Here’s a Monday morning at a 30-person commercial cleaning company in Mission Valley. By 8 a.m. three monkeys are waiting on the owner:

  1. A medical office is upset because last night’s crew left streaks on the floors the day before an inspection. The crew supervisor called the operations manager, who wasn’t sure he could approve overtime for a re-clean, so he texted the owner.
  2. The property manager for that same building has the owner’s cell from the original walkthrough, so she skipped everybody and called him directly.
  3. The estimator wants to knock 8% off a bid for a three-building office park. Nobody ever told him where the floor is, so the bid is sitting in the owner’s inbox.

Nobody designed it this way. Each monkey traveled up until it found someone allowed to decide, and in most companies this size there’s exactly one of those.

There’s a second reason, and it’s harder to hear. Before you owned the company you were probably the best estimator or the best fixer in it. I’ve called that being an “accidental escalation resource.” You became the owner and never stopped being the person every hard problem goes to. And if we’re being straight with each other, part of you likes it. Fixing in four minutes what your manager would sweat over for a day is the best part of a lot of owners’ weeks. That’s exactly why it’s so hard to stop.

How many monkeys are on your back right now?

More than you’d guess, and you can count them in a week. For five working days, keep a running note. Every time something lands on you, write down who brought it, what it was, and whose move it should have been. Don’t change anything yet. Just count.

On Friday, sort the list into three piles. The first is the stuff that really needed you: real money, losing a person, a breach. There are usually only a handful.

The second pile is decisions someone else could have made if they’d known how far they were allowed to go. This pile is usually the biggest, and every item in it is a limit you never wrote down.

The third pile is habit. Someone could have decided today and asked anyway, because asking you is faster than deciding. You trained that one, and you can untrain it.

A week of notes shows you what climbed to you this week. The 30-day test shows which decisions are climbing to you today, from the job that’s still broken after the first person tried to fix it to the call from your biggest customer, and where each one goes when you’re away. If “Problems climb to you” comes back near the top of your result, the rest of this article is your to-do list.

What should come to the owner, and what shouldn’t?

Three things should climb to you: losing money past a limit you’ve set, losing a person, and a breach. Everything else belongs to whoever owns the work, and you see how it went in the weekly numbers.

I gave a version of this to an owner I coach this spring. He was working out how much his leaders should be telling him, and whether asking for a daily update made him a babysitter. My answer was to write the boundary down: “Here’s what I would like to know about and here’s what I would not like to or I don’t need to know about. That takes away ambiguity.” Then the rule itself: “you don’t have to read my mind. If it doesn’t involve losing money, losing an employee or some type of a technology breach or cybersecurity thing that I know [the owner]‘s going to have a coronary on.” And the part after it: “do your thing. You’ll report to me KPIs that will tell the rest of the story.”

Here’s each line, translated for a company that isn’t in the security business.

Losing money needs a number, or every decision qualifies. A $40 refund is losing money. So is a $40,000 job that goes sideways. The table in the next section is how you turn that line into something your team can use.

Losing an employee means a resignation, a firing, or anything that could end in one. You hear about those the same day. The manager still runs the rehire. You just never find out last.

A breach is whatever your trade’s version of “we got hacked” is. Customer data out the door counts for everyone. For a contractor, add an injury on a job site. For a clinic, anything that touches a patient’s safety or a license. Write your version down.

The second half of the rule is the part owners skip. You still see every monkey you hand back. You see it on a schedule, in the numbers, and the hallway gets a lot quieter. Oncken and Wass called that feeding by appointment, and it’s the whole job of a weekly leadership meeting that actually produces decisions.

How do you set decision rights with dollar limits?

Take the decisions that climb to you most often, give each one an owner and a dollar or hour limit, and agree in advance on when it comes to you. On a coaching call last year I put it this way: “Now you never want to say, I have to quote it because you’ll never have time to do that. So it becomes just an approval. These 10 things I have to approve. These other things you don’t. Typically, you can also set dollar boundaries.”

Here’s a decision-rights table for that Mission Valley cleaning company: about 30 people, night crews with supervisors, an account manager, an operations manager and an estimator. The numbers are illustrative. Yours depend on your margins and on how much a mistake would hurt.

DecisionWho owns itThe limitWhen it comes to you
Credit or refund for an unhappy customerAccount managerUp to $500 per month on that accountOver $500, or the same customer twice in a quarter
Discount or price exception on a quoteEstimatorUp to 5% off, as long as the job stays above your margin floorMore than 5%, anything below the floor, or any multi-year contract
A re-clean, overtime or an overrun on a job that went wrongOperations managerUp to $5,000 in labor or 40 crew hours. Past $2,500 or 20 hours, they act and tell you that dayOver $5,000 in labor or 40 hours, or anything involving an injury or property damage
The biggest customer calls with a problemThat account’s named first contactAnything inside the credit and re-clean limitsPast those limits, or when the customer says they’re leaving. You’re the backup call
Supplies, equipment and softwareWhoever runs that budgetInside the monthly budget, up to $1,500 a purchaseOver $1,500 a purchase, outside the budget, or any new monthly contract
Backfilling someone who quitThat person’s managerInside the approved role and pay rangeA brand-new role, or pay above the range
A resignation, a firing, a complaint about a managerThe manager, with youNone. These always come upSame day
A data breach, a legal threat, an injuryWhoever finds itNoneImmediately

How to set the numbers without guessing:

Start with last month. Pull every approval you gave and write down why you said yes or no. Most were yeses you’d give again without a second thought. Set the limit just above those.

Size each limit to a mistake you can live with. If the estimator gets a 5% discount wrong, the company survives. If that row makes you nervous, read why San Diego owners cave on price first. A discount costs more than it looks, and the floor is easier to set once you’ve run that math.

Decide how they tell you. This is where the Oncken and Wass levels earn their keep. Under the limit, they act and report on the normal schedule (their level 5). Close to the limit, or anything unusual, they act and tell you that day (level 4). Over the limit, they bring you a recommendation and act on your answer (level 3). Nobody waits to be told, and nobody just asks what to do.

Review the calls in the weekly meeting, then move the limit. When someone keeps getting a decision right, raise their number. When someone gets one wrong, talk about why before you touch it.

And when a job runs over, make the cost visible. On that same call I told an owner: “Just take $40 an hour, multiply by how many hours the project is over, and say, guys, this came out of my pocket.” Nobody’s in trouble. “They need to feel that this sucks.” People who can see the cost make better calls with the limit you gave them.

What are the levels of delegation?

The levels of delegation describe how much someone decides before a problem comes to you, from “do exactly what I say” up to “handle it, I’ll see it in the numbers.” The version most people find first is Michael Hyatt’s five levels of delegation:

  1. Do as I say.
  2. Research and report.
  3. Research and recommend.
  4. Decide and inform.
  5. Act independently.

The top of that list tracks the Oncken and Wass levels from 1974 closely: recommend, then act on the answer; act, then tell. The difference worth noticing is level 5. Hyatt’s tells the person “No need to report back.” Theirs still reports on the normal schedule, and for an owner that’s the version to use, because you want the decision out of your hallway and still in your weekly numbers. If you’ve seen a list of seven, it’s probably Management 3.0’s, which runs from “tell” to “delegate” with more steps in between.

Here’s where owners go wrong with any of these lists. They give the level to a person. “You’re a level 4 now” sounds generous, and it falls apart the first time that person meets a decision they’ve never made. A level belongs to a decision. The table above already works this way: the estimator’s small discounts and a multi-year contract sit at different levels.

Move a decision up one level at a time, starting from level 3, and only after it’s gone right a few times at the level below. “Win $400,000 of new contracts this year, all above the margin floor” is the hand-off at the top. “Bid ten office parks” is a to-do list. Give them the number and let them choose the method, because people who get a say in how own the result. I wrote it on LinkedIn in January 2025: “When they weigh-in, they buy-in!”

How do you hand a monkey back without dropping it?

Hand back the next move, put the next check-in on the calendar before they leave the room, and judge the result when it comes.

In the article, the manager’s turnaround comes on a Monday morning. He calls each employee in, puts the monkey on the desk between them, and works out what the next move is and who makes it. Then he tells them, more or less, that the problem leaves his office the way it came in: on their back. He’ll help at a set time, and he won’t make a move alone.

The owner’s version is shorter. When someone brings you a problem, ask “What do you recommend?” If the answer’s good, say “Do it, and tell me Friday how it went.” If it isn’t, ask what they’d need to decide it without you next time, and get them that.

Then check the work, on a schedule. Hand it off and never look again, and you’ve abdicated, which is how a hand-off quietly rots for a year. Check every step, and you’re hovering. As I told him on that same call, “you walk a tightrope a little bit because you don’t want to abdicate, but you also don’t want to micromanage.” The way across is to check outcomes (the customer, the margin, the date) and leave the method to them. Better yet, set it up so you never have to ask. If the work’s being done right, you should be able to look at the result any day you like and see it.

Asking for a recommendation only works if they can see what you see. If the job margin, the customer’s history and the price floor all live in your head, nobody else can recommend anything, and the monkey comes right back. Give the people who own a decision the numbers that decision depends on.

Don’t overcorrect, either. I did. I was hired to run a services company, and we grew it from $2M to $20M while I sat in that seat. Around 2017 I’d pulled back so far that I went about six months without talking to a single front-line employee. When I told that owner the story this spring, I put it this way: “It got to a point where company culture got away from me.” I couldn’t tell you whether the decisions got better or worse in those months. The culture drifted, and so did my sense of who I was in that company.

So the advice I give now comes with a footnote: once the hand-offs are working, “randomly hop into the huddle probably once a month, twice a month.” Step out of the day-to-day decisions. Stay in growth and direction, and stay where your people can see you.

What happens when your team makes the wrong call?

They will, and how you react the first time decides whether anyone makes a second call without you. I put it bluntly to a leader I was coaching: “if every mistake that they make in trying to improve the company is met with, oh, my God, you’re so dumb, why didn’t you come talk to me before you made that decision? Your VPs are useless at that point.” They turn into ventriloquist dolls. Everybody in the meeting is talking, and it’s all your voice coming out.

Most owners don’t mean to do it. I wrote a LinkedIn post last year about the monologue nobody says out loud: “You don’t know what you’re doing.” “I know better.” “I’d teach you but I don’t have time.” Underneath it: “Our obsession with control is fueled by fear of failure. If we make the call, we believe we won’t fail. If we do fail, we trust ourselves to clean it up.”

So when the call goes wrong, let them clean it up, with you coaching from the side. Same post: “Build resilience. Coach the cleanup. Let them fail.” Then go back to the table and decide whether the limit was wrong or the information was. More often than not it’s the information.

What do owners say that puts the monkey on their own back?

Mostly polite, helpful-sounding things. Listen for these in yourself this week.

  • “Let me think about it.” The original hallway line. The next move is now yours. Ask “What do you recommend?” instead.
  • “Send me something on that.” Now you have to read it and reply. Oncken and Wass had the memo version of this in 1974. Try “Bring it Thursday with your recommendation.”
  • “I’ll call them.” The customer learns your number works better than your team’s. Have the account’s first contact make the call and tell you how it went.
  • “Yes,” to a customer your team already told no. You taught the customer to go around your people, and you taught your people that their answer doesn’t count. Back the call your team made, then fix the limit later if it was wrong.
  • “Loop me in.” Every thread you’re copied on is a thread waiting on you. Swap it for “Tell me Friday if anything crossed the limit.”

Which monkey should you hand back first?

The one that climbs to you most often. Your team can’t stop reading your mind until you write down something else for them to read.

So here’s the assignment, and it’s the question I asked in that LinkedIn post: “What is one decision you still refuse to delegate, and what safety net would make you hand it off this week?” Put the answer on one row of the table. Give it an owner and a limit. Walk through the first decision with them, then let them make the next one, and review it at the weekly meeting. That’s the whole job for this week, and it’s the one monkey nobody on your team can take off your back.

If you’re not sure which row to start with, the 30-day test will show you which decisions are climbing to you right now. And when you want someone holding you to the table once it’s written, that’s what my San Diego business coaching is built around.

What are the four rules of monkey management?
They come from The One Minute Manager Meets the Monkey (1989) by Ken Blanchard, William Oncken Jr. and Hal Burrows. Describe the next move, give the monkey one owner, insure against the risk by agreeing how much initiative that owner takes, and schedule the next feeding and check-up. The original 1974 HBR article lists five care-and-feeding rules instead.
What are the five levels of delegation?
The list most people find first is Michael Hyatt's: do as I say, research and report, research and recommend, decide and inform, and act independently. Its upper levels track the levels of initiative Oncken and Wass described in 1974, except that their top level still reports on a schedule. Set a level for each decision, and move a decision up once it keeps going right.
What is upward delegation?
Upward delegation, also called reverse delegation, is when a problem your employee owns ends up on your desk. It usually starts with you saying "let me look into it," or answering a question you should have bounced back to them. The work moves up, the employee waits on you, and you become the bottleneck on something that was never yours.
How do I get my employees to stop bringing every decision to me?
Write down what has to come to you (money past a set limit, losing a person, a breach or safety issue) and give every other common decision an owner and a dollar limit. When something climbs anyway, ask "what do you recommend?" and let them act on it, then review their calls at the weekly meeting.
What's the difference between delegation and abdication?
Delegation is handing someone the work and checking the result. Abdication is handing it off and never checking, which is how a hand-off gets worse for a year before anyone notices. Check outcomes on a schedule (the customer, the margin, the date) and leave the method to the person doing it.
Is Who's Got the Monkey the same as The One Minute Manager Meets the Monkey?
They're related. "Management Time: Who's Got the Monkey?" is the 1974 Harvard Business Review article by William Oncken Jr. and Donald L. Wass. The One Minute Manager Meets the Monkey is a 1989 book by Ken Blanchard, Oncken and Hal Burrows that retells the idea as a story and adds the four rules. Same monkey, different packaging.

Work on the business, not in it. Fine. Doing what, on Tuesday at nine?

Your biggest customer calls with a problem. Whose phone rings?

Result on screen. Map by email.

Let's see if we're a good fit.

Pulling up the calendar…

Nothing this week? Text me: (619) 249-8121. Or email kyle@k7leadership.com.  ·  Calendar trouble? Open it in a new tab.