You hired people to take work off your plate, and you still check, change or approve what they do before it goes out. In our view, the fix is to hand over decisions, not just tasks: who decides what, up to a written limit, with you reviewing the result after instead of approving the work before. By the end of this article you will have the six steps to hand decisions over, a one-line decision rule with a limit that you can copy, and a clear answer to what you check instead of the work.
Six steps hand over the decision, not just the task
These six steps work with one person and three decisions at a time. Do them in order, because each step hands the next one what it needs. Our rule of thumb is to start with one person, so you learn the method before you change how the whole team works.
Log the decisions you check before you hand anything over
Do this. Pick the person whose work you check more than anyone else's, and over their next ten pieces of work write down every time you check, change or approve something, and which decision sat behind it: a price, a refund, a spend, the wording, the schedule or a supplier.
It worked when you have ten logged checks, each tied to a named decision rather than to a task.
Why. In our view, a list of tasks hides the real issue: the work has moved to someone else, but the decisions inside it still come back to you. One owner put it plainly: "If I’m hovering over your shoulder to make sure things get done, I’m basically not delegating anything....". We read that as the same point from the owner's side: if you check every step, the decision never left your desk.
Hand over the reversible decisions first and keep the one-way doors
Do this. Mark each decision on your list as a two-way door (you can undo it at small cost) or a one-way door (hard or impossible to undo), then pick up to three two-way doors that came up more than the others in your log.
It worked when you have three reversible decisions to hand over and a short list of hard-to-undo decisions that stay with you for now.
Why. In his 2015 letter to Amazon shareholders, Jeff Bezos wrote that some decisions are "consequential and irreversible or nearly irreversible", and called them one-way doors 1. Bezos wrote that one-way-door decisions "must be made methodically, carefully, slowly, with great deliberation and consultation" 1. Bezos wrote that most decisions are "changeable, reversible", and called them two-way doors 1. Bezos wrote that two-way-door decisions "can and should be made quickly by high judgment individuals or small groups" 1. We use the same split to choose what goes first: in our view, a reversible decision is the cheapest place for someone to learn your judgement.
Hand over next
Low risk, but too few cases yet to learn from.Hand over first
Can be undone and comes up often: your three lines start here.Keep
Stays your call for now.Keep the call, ask for a recommendation
They bring the answer they would give; you decide.Write each decision as one line with a limit
Do this. For each of the three decisions, write one line that names who decides, up to what limit (a sum, a discount or a date), what comes back to you above that limit, and that they tell you after; then add one sentence on why the decision matters to the customer and to the business.
It worked when someone who reads the line can say, without asking you, whether a given case is theirs to decide.
Why. Each line names one person as the decider, not a group. Bain & Company's RAPID tool names five roles in a decision: Recommend, Agree, Perform, Input and Decide 2. Bain says the Decide role "makes the final decision and commits the organization to action", and that "Ideally, there should only be one decider for each decision" 2. In our view, the limit is what turns 'ask me first' into 'decide, then tell me', so you can let go without losing sight of the risk. In our view, a person who doesn't know why a decision matters can play it safe or come back to ask, so write the why down.
Example. Horst Schulze, co-founder of The Ritz-Carlton, confirmed in a 2019 Knowledge at Wharton interview that "every employee was empowered" to spend up to $2,000 to correct a mistake 3. Schulze said this came with "teaching every employee problem resolution and saying, 'You can spend up to $2,000'" 3. Schulze said he knew "a repeat customer lifelong was worth about $200,000 to the company" 3. Schulze said: "Nobody ever spent $2,000, but they bought breakfast or sent a food basket or some cookies" 3. Schulze said he was sued over the policy, and that "It was not a decision to throw away money" 3. We read that as the point of a high limit: it settled who decides, and the spending stayed small.
Match each handover to how much the person has done that decision before
Do this. Hand the three lines over in one short talk, agree for each decision whether the person finds out and reports, decides and keeps you informed, or decides and tells you after, and ask how much check-in they want and in what form.
It worked when the person can repeat back, for each decision, what they decide, the limit, the stage and when they tell you.
Why. Andy Grove's High Output Management has a section on what Grove calls task-relevant maturity, as Ben Horowitz notes in his foreword to the book 4. Horowitz sums up the rule this way: "If the employee is immature in the task, then hands-on training is essential" 4. Horowitz continues: "If the employee is more mature, then a delegate approach is warranted" 4. We apply that to each decision, not to the person: the same employee can be new to refunds and experienced with the schedule. Our rule of thumb is that less experience with a decision means more check-ins, not more say from you over how the work gets done. Netflix says it expects its managers to practice "context not control", giving their teams "the context and clarity needed to make good decisions instead of trying to control everything themselves" 5. Netflix adds that context not control "should not be confused with hands-off management" 5. In our view, the stages are how you stay close to the work without taking the decision back.
Review the result after the decision, not the work before it
Do this. For the next thirty days, approve nothing that falls under the limits; read what the person tells you after, and when you go through their decisions, ask one question: did it reach the result we agreed?
It worked when within thirty days you have approved nothing under the limits and can name the decisions they made without you.
Why. Judge each decision on its result: done, on time, at the quality agreed and the goal hit, not on whether it was made the way you would have made it. One person who says they have built many teams put it this way: "The third key is that other people will not do things the way that you do them, but did they accomplish the objective? If yes then why stress about the how, unless it’s an efficiency issue, then that’s coaching." Keep a short list of critical items that you check directly by asking to see the result, and expect the rest. If you disagree with a call that was inside the limit, say so and let it stand. In his 2016 letter to shareholders, Bezos told readers to "use the phrase 'disagree and commit'" 6. We read 'disagree and commit' as the habit that keeps a handed-over decision handed over.
Widen the limit on the decisions that held, and fix only the one that didn't
Do this. On day thirty, count how many of the three decisions came back to you and check their results; raise the limit or move to the next stage on each decision that held, and where one went wrong, change that one person's limit instead of taking the decision back from everyone.
It worked when at least one decision has a higher limit or a later stage, written on its line.
Why. In our view, trust in someone's judgement grows from watching their results, and the written line is where that trust gets recorded. In our view, a new rule for the whole team after one person's mistake takes the decision back from the people who got it right. If you and the person keep disagreeing about what a good result is, raise it at once instead of letting it build. Bezos also wrote: "recognize true misalignment issues early and escalate them immediately" 6.
We think micromanaging is the cost of keeping the decisions after handing over the work
One employer asked: "Why am I paying just to have to micromanage?" In our view, the answer is often that the work was handed over and the decisions were not. We think you stay the bottleneck when you hand over the tasks but keep the authority: the work moves, and the decisions still come back to you.
Our reading is that a business can only grow as far as the owner's own capacity while every decision runs through the owner. That is why, in our view, handing over decisions matters more than handing over tasks.
The owner hands over the tasks, keeps every decision and approves the work before it goes out.
The team decides within written limits and tells the owner after; the owner reviews the results.
A University of Birmingham study by Dr Daniel Wheatley (Work and Occupations, 2017) used two separate years of data for 20,000 employees from the Understanding Society survey 7. The study found that "employees with higher levels of autonomy in their work reported positive effects on their overall well-being" 7. The university says the research also highlighted that "in many cases managers remain unwilling to offer employees greater levels of autonomy", because their primary role remains one of "control and effort extraction" 7. Wheatley said: "Greater levels of control over work tasks and schedule have the potential to generate significant benefits" 7. In our view, an owner who writes the limits down is making the opposite choice.
We think telling an owner to trust their people names the goal, not the path. In our view, the path is a written limit that widens as the results come in.
On the Scorecard's decisions row, this article covers the move from "tasks are handed over, I approve everything" to "written limits exist for who decides what". In our view, the written limit is the whole difference between the two.
A submarine, a streaming service, a carmaker and the Army each gave decisions to the people doing the work
L. David Marquet commanded the nuclear-powered, fast-attack submarine USS Santa Fe from 1999 to 2001 8. According to the publisher of Marquet's book Turn the Ship Around!, the Santa Fe had poor morale, poor performance and "the worst retention rate in the fleet" 8. Marquet once gave an impossible order without knowing it, and his crew tried to follow it anyway, his publisher writes 8. When Marquet asked why, the answer was "Because you told me to" 8. We read 'Because you told me to' as what a crew says when every decision comes from the top.
Marquet's publisher says he moved to "giving control to his subordinates", and that the Santa Fe "skyrocketed from worst to first in the fleet" 8. The company built on Marquet's method sums it up in one line: "I intend to..." replaces permission-and-compliance 9. In our view, 'I intend to...' is the one-line handover in spoken form: the person states the decision, and you step in only if you must.
For every significant decision, Netflix names an "informed captain" who is "responsible for making a judgment call on the right way ahead" 5. Netflix says it prides itself "on how few, not how many, decisions senior leaders make" 5. Netflix says it avoids "decision-making by committee, which tends to slow companies down and undermine accountability" 5. We read that as the same rule at a larger scale: one named decider per significant decision, and fewer decisions for senior leaders.
Toyota's UK magazine says every Toyota member is viewed as an expert, and "each and every one is permitted to stop the production line" 10. Toyota says that activating the andon alert, "usually by a pull-cord or button", "automatically halts production" 10. We read that as a decision right given to every person on the line, with no approval needed first.
US Army doctrine ADP 6-0, as quoted on army.mil in 2015, defined mission command as "the exercise of authority and direction by the commander using mission orders to enable disciplined initiative within the commander's intent" 11. In 2015, army.mil reported that the Army had adopted six principles of mission command: "build cohesive teams through mutual trust, create shared understanding, provide clear commander's intent, exercise disciplined initiative, use mission orders, and accept prudent risk" 11. We read 'accept prudent risk' as the price of any limit: some calls under it will go differently than yours would.
Gallup surveyed 143 CEOs of Inc. 500 companies; those with high Delegator talent posted an average three-year growth rate of 1,751%, 112 percentage points more than CEOs with limited or low Delegator talent 12. Gallup found that the CEOs with high Delegator talent "generated 33% greater revenue in 2013": $8 million versus $6 million 12. Gallup reports that only one in four employer entrepreneurs have high levels of Delegator talent 12. Gallup writes that "Non-Delegators do not trust others to do things as well as they themselves can" 12. Gallup describes Non-Delegators as "centralizing and controlling decision-making at every turn" 12. We read these figures as a link between delegating and growth, not proof that one causes the other.
Handing over decisions has limits: people need training first, and some calls stay yours
One person added a warning: "Yes, delegate and don't micromanage, but you also have to train people and get them up to speed with how you work and what the expectations are. Expecting people to walk in and get it without any help or guidance is bad management too." In our view, not hovering does not mean disappearing, and handing over with no support is the other way delegation fails. One person drew the line like this: "If you’ve trained someone, and they are meeting your expectations and you’re still telling them how to do everything, then you’re micromanaging and likely driving them insane." In our view, that is the test to run before step one: if the person isn't trained in a decision yet, start them at the first stage, not at no stage.
Go back to the owner quoted in step one. The same owner named two possible causes: "I’ve realized micromanaging usually means one of two things: I hired the wrong person. I’m a control freak who hasn’t built proper systems." Our rule of thumb is to check, before you judge the hire, whether a written line and limit ever existed. If one person keeps failing after a clear handover, treat it as a performance question, which our guide to managing underperforming employees covers.
I’d also feel weird being in charge of someone or having to boss them around and tell them what to do.one owner who still does everything alone
In our view, the written line takes the bossing out of it: you agree the decision and the limit once, and after that the person decides and you read the result.
One owner with three locations described the setup this way: "Because of that, I rely heavily on staff and cameras to monitor what’s going on." One person answered that store owner like this: "yoo this is a management infrastructure problem not a people problem. you can't rely on cameras and enforcement when you're not there. hire a store manager and delegate authority." One person who runs a B2B shop passed on what two HVAC owners had told them about cameras in their trucks: "The guys felt like 'Big Brother' was breathing down their neck and that management didn't trust them anymore." We read that as the cost of watching in place of handing over: a camera checks the work, while a written limit hands over the decision.
Some decisions stay yours: the one-way doors from step two, and any decision you haven't yet trained someone for. In our view, not everything should be handed over, and the sorting in step two is how you keep the hard-to-undo calls with you.
Start the thirty-day handover with one person this week
the thirty-day handover
If you can do only one part this week, do the log, because it shows you which decisions you are still holding. Keep the log to one person, so it adds no meetings.
You now have the steps, the line and what to check instead
You started with the question of why you pay people and still check their work. The six steps move three decisions from your desk to one person in thirty days. The one-line decision rule names who decides, up to what limit, and that they tell you after. What you check instead is the result, after the decision, plus a short list of critical items you ask to see.
Read before: How to Replace Yourself in Your Business (Before You Can Afford a Team) and How to Build Systems in Your Business That Your Team Actually Uses. Read next: How to Manage Underperforming Employees (Before You Decide to Fire Them), then our coming guides on spending limits and on reviewing your team's decisions once a week.
Where are you on decisions? Take the Scorecard to see your row and level.
Sources
- 2015 Letter to Shareholders · Amazon.com, Inc. (Jeff Bezos), April 2016
- RAPID® Decision Making · Bain & Company, 13 October 2023
- Becoming the Best: Excellent Customer Service Always Wins · Knowledge at Wharton (interview with Horst Schulze), 2 May 2019
- Andy (foreword to High Output Management) · a16z (Ben Horowitz), 13 November 2015
- Netflix culture memo · Netflix, undated
- 2016 Letter to Shareholders · Amazon.com, Inc. (Jeff Bezos), April 2017
- Autonomy in the workplace has positive effects on well-being and job satisfaction, study finds · University of Birmingham, 25 April 2017
- Turn the Ship Around! · Portfolio / Penguin Random House (L. David Marquet), 16 May 2013
- Intent-Based Leadership · Intent-Based Leadership, undated
- Andon – Toyota Production System guide · Toyota UK Magazine, 31 May 2016
- Understanding mission command · army.mil (James D. Sharpe Jr. and Thomas E. Creviston), 30 April 2015
- Delegating: A Huge Management Challenge for Entrepreneurs · Gallup Business Journal (Sangeeta Bharadwaj Badal and Bryant Ott), 14 April 2015
