Delegation rarely fails because the wrong person was chosen. It fails because the task moved while everything the task depended on stayed where it was.
The owner gets an hour back. The business gets a slower version of the same work, plus a checking step that did not exist before. After that happens twice, most owners conclude that delegation does not work here, which is the wrong lesson drawn from accurate evidence.
The work that cannot be handed over does not grow
The reflex to solve this by hiring is covered separately in why hiring more people stops fixing the problem. This post is about the work itself.
There is a category of work only the owner can do. Deciding what the business will not pursue. Holding a relationship that exists because of who they are. Setting the standard nobody else has the standing to set.
That category is surprisingly small. More to the point, it stays roughly the same size as the business grows, because it is bounded by what one person is uniquely placed to do rather than by how much there is to be done.
What grows is everything else.
The gap in that second bar is the interesting part. It is not lost time. It is capacity the business has already paid for in salaries it is already meeting, sitting out of reach.
The three conditions a task needs before it can move
A task is not a self-contained object. It arrives with three things attached. Handing over the task alone leaves all three behind.
- A written route through. Not a procedure document nobody opens. The half-dozen decisions the task actually involves, written where the person doing it will look. Undocumented work takes longer to explain than to do, which is why it rarely gets handed over.
- Authority to finish. If the person taking it has to come back for a decision at any point, the task has not moved. It has acquired a second participant. Authority means a named limit inside which they do not ask.
- A system configured for the current size. Access, permissions, a licence, a queue. Work routed through one person is often routed there because the software was set up when that was reasonable.
Move the task without those and the bottleneck travels one desk sideways. Move those first and the task often moves on its own, because somebody else can already see how.
This is also where the research lands. Summarising the World Management Survey in the Journal of Economic Perspectives, Bloom and Van Reenen report that firms whose chief executive founded the business score poorly on management practice. Their explanation is plain: a mature firm needs to move beyond informal rules.
The order most businesses get wrong
The instinct is to find the person, hand over the work, then fix the problems that surface. That order guarantees the problems surface in front of a new person who has no context for them, which is how a competent hire acquires a reputation for being slow.
Reversing it is unglamorous. Write down the route. Set the limit. Fix the access. Then hand it over, to somebody who now has a fair chance.
It also explains the most common objection, which is that documenting it takes longer than doing it. That is true once. It is false by the fourth repetition, on work that is usually running weekly.
What this is really costing
The hour is the smallest part of it.
What the owner is not doing instead is the first cost. An hour on invoicing is not worth an invoicing clerk's hourly rate. It is worth whatever the business gives up by not having its owner on the work only they can do.
Underneath that sits a ceiling on everybody else, because people stop bringing work forward once they learn where it will end up anyway.
Then there is what it does to the value of the business. An operation that depends on one person for routine decisions is worth less than one that does not. A buyer is paying for a business that runs on its own. What they would inherit is one that runs through you.
How to measure your own version of it
Two steps. Your own diary gives you the number. Four systems tell you whether to believe it.
Step one: the two-week audit
Take your last two weeks. Tag every block of an hour or more into one of three categories.
Your number is Category 2 hours ÷ total hours, the delegable share of your week.
Step two: four readings that can contradict you
These come from systems with no opinion about it. Where they disagree with your audit, they are right.
| Where to look | What to pull | The calculation | What it tells you |
|---|---|---|---|
| Job or project system | Items whose next action sits with you | Items on you ÷ open items = concentration | Category 2 with a face on it. Names the work you did not think to tag |
| Approval history in your finance system | Requests waiting on one signature | Median days to decision = decision lag | What the queue behind you costs everybody else |
| Accounting in Xero or MYOB | Gross profit against average headcount, three years | Gross profit ÷ average FTE = profit per head | Whether added people are adding output rather than arriving into the same constraint |
| Payroll | Overtime by team, against that team's volume | Overtime hours ÷ team hours = strain | Where capacity ran out already. Often where your category 2 work was quietly sent |
A high delegable share with healthy figures below it is a personal habit. The same share alongside rising decision lag and rising overtime is structural. It costs considerably more.
The one number to hold
- Category 2 hours a month, multiplied by what an hour of your time is worth to the business. Not your salary. What the business gives up when you spend an hour on invoicing
- Recalculate it quarterly. It moves when the business changes shape, which is exactly when nobody is looking at it
The control worth introducing
- A written authority limit for each person who takes work from you. What they may decide, up to what value, without asking
- Most businesses have a spending limit and nothing at all for decisions, which is why work comes back
- Before handing anything over, write the half-dozen decisions it involves. If you cannot, that is the finding
Where your number will mislead you
- A fortnight is a sample, not a measurement. Month-end, quarter-end and holidays all distort it
- Category 2 is systematically underestimated, because work you have always done stops looking delegable
- Profit per head flatters a business that has just moved work to contractors, since the cost leaves payroll for cost of sale
Which of your hours would survive a handover?
Take the three categories and be honest about the middle one. Not what could move in principle. What could move on Monday, to somebody already employed, without coming back to you.
If that list is short, the constraint is rarely the people. It is that the route, the authority or the access was never set up. Noticing any of them is rarely anybody's job.
The Free Health Check covers how decisions and work actually flow, in about three minutes. If you would rather see the whole operating picture evidenced rather than sampled, that is what a diagnostic is built to produce.