A five-minute approval can consume far more than five minutes of people's time. One person-minute is one minute of one person's active work. A teammate frames the question, waits for a gap in your day, reopens the work and may still need a correction.
The delegation debt calculator below counts active minutes you can name and keeps elapsed waiting in its own column.
I use delegation debt here for recurring work that still sits with you or keeps coming back after someone else was meant to own it. It is a planning label for this article, not an industry standard or a management score.
Choose one responsibility that keeps coming back
Start with one recurring responsibility that has a recognisable beginning and end. If you still need to find the dependency, use the founder bottleneck audit first and bring one routine row back here. The calculator is for a known handover candidate, not a diagnosis of your whole job.
Use a specific task such as 'prepare and send the weekly client status pack' or 'approve pricing exceptions inside the standard range'. 'Marketing', 'operations' and 'help the founder' are too broad to measure. Pull the last three to five normal occurrences if you have them, and leave a launch or incident week out of the baseline.
Put today's work and the handover in separate columns
In the first column, add the minutes you spend, the minutes other people spend doing the work, and any time spent correcting it or waiting without being able to move on. In the handover column, estimate the review you will keep, the new owner's working time and the rework you still expect.
Count the new owner's 35 minutes alongside your 45. Time returned to you and time saved across the team are separate outputs.
A survey of 215 unit managers in professional-services firms found that managers gave people more authority when those people had knowledge specific to the work, tasks involved more exceptions and monitoring cost less. The study on delegation and autonomy helps explain why your estimate has to match this work and this owner. Use it to estimate the review you should keep after handover.
If you cannot estimate the review you will keep because nobody knows which decisions still belong to you, set the decision boundary before you calculate. A transfer with hidden approvals will look better on paper than it feels in the week.
Count the minutes people actually spend
Active time includes doing the work, explaining it and correcting it. A 30-minute meeting with four people consumes 120 person-minutes. Count each person's contribution once, including each side of an approval conversation in the right person's column.
Elapsed time answers a different question. The Kanban Guide defines cycle time as elapsed time from an item's start to finish. Keep that clock time beside active work instead of adding it to labour. Record paid minutes in the active-time column.
Add the one-off time for documentation, training and supervised runs after the recurring work is clear. Past effort is already spent. Add any new ongoing tool or contractor cost to the handover column.
Find the payback point
Use the same period for every input. Monthly is usually easy for recurring work. Before you calculate, record how often the work runs; each person's active minutes now and after handover; rework now and after; one-off handover hours; and elapsed time now and after. Write every assumption beside the result so another person can challenge the number without reverse-engineering your spreadsheet.
| Output | Formula | What the result means |
|---|---|---|
| Current monthly load | monthly current hours = runs per month x (leader minutes now + owner execution minutes now + team rework minutes now) / 60 | Active person-hours consumed by the current route. |
| Future monthly load | monthly future hours = runs per month x (leader minutes after + owner execution minutes after + team rework minutes after) / 60 | Expected active person-hours after the handover has settled. |
| Leader capacity released | runs per month x (leader minutes now - leader minutes after) / 60 | Time returned to the founder or manager, whether or not total team time falls. |
| Active hours recovered | monthly current hours - monthly future hours | The change in whole-team active time. A negative result means the proposed route uses more person-hours. |
| Payback point | payback months = one-off transfer person-hours / monthly active hours recovered | How long the setup takes to repay in person-hours, only when active hours recovered is positive. |
| Elapsed flow change | elapsed time now - elapsed time after | Clock time removed per occurrence. Keep it outside the labour and payback totals. |
In a hypothetical weekly client-status example, the route runs four times a month. It currently uses 45 leader minutes, 35 owner-execution minutes and 20 rework minutes each time: 6 hours 40 minutes a month. After handover, it needs 10 leader minutes, 40 owner minutes and 10 rework minutes: 4 hours a month.
Under those assumptions, 6 person-hours of setup pays back in 2.25 months. The handover gives you back 2 hours 20 minutes and saves 2 hours 40 minutes across the team each month. Replace every estimate with actual time after the test.
Check the result before you trust it
When future monthly hours equal or exceed today's work, there is no time payback under those assumptions. The transfer can still return time to you, shorten a customer wait or develop another owner. Those may be good reasons to proceed, but the calculator has not proved a saving.
Money is optional. If you add it, use each person's actual hourly cost to the business and include only costs or savings that change because of the handover. Existing payroll moved to other work changes capacity, not cash. The OpenStax guide to relevant costs explains which future costs belong in the comparison, while ACCA's payback guide explains cash payback. Use that label only when both sides are actual cash movements. Company revenue is not your hourly rate.
Run the calculation three times: optimistic, expected and cautious. Change one uncertain estimate at a time. Start with how often the work repeats, how much review you keep or how long the new owner will need.
Test one handover and replace estimates with actuals
Choose work that repeats, has an obvious next owner and can be tried without creating a hard-to-reverse risk. A large theoretical gain may still be a poor first transfer when the work includes an irreversible customer promise or sensitive access.
If nobody can own the result, stop treating it as a delegation exercise and check whether the work is ownerless. If a viable transfer keeps getting reclaimed after meeting the agreed boundary, use the manager delegation diagnostic. Those patterns need a different conversation.
When the route is sound but no current person has the capacity or knowledge to carry it, choose the next role from the constraint. A hiring decision needs evidence beyond one attractive payback number.
Write down what a good result looks like, what you will still review and what would bring you back in. Run two to four normal cycles. Then replace the estimates with the minutes each person actually spent and decide whether to continue, narrow the boundary or stop.