On this page
Key takeaways
- When your best manager quits, the largest cost is not recruitment but everything they knew that was never written down, which leaves the same day they do.
- There are three costs: recruitment, which everyone counts; the ramp, which some people count; and the knowledge, which nobody counts.
- A handover, usually two weeks of notice and a walkthrough, captures keys, codes, suppliers and rotas, but almost none of the judgement, because the departing manager does not know what they know.
- Keep a continuous record of what was noticed, what was decided and what happened next, so a departure removes a person rather than a history.
- A manager whose knowledge is captured can take a holiday or be promoted without leaving a hole, which is a better job, not a smaller one.
A good bar manager holds a large amount of the business in their head. Which supplier short delivers and needs checking. Which regulars will complain and which will simply leave. How Friday staffs differently from Saturday. What went wrong the last time a large group booked late. Which staff member is close to leaving.
None of that is in any system. It is in a person, and on the day they resign the business loses it in a single step.
In hospitality that day comes often. In July 2026 the monthly quits rate in accommodation and food services was 3.7 percent, against 1.9 percent across all nonfarm jobs, according to the Bureau of Labor Statistics.
The three costs, in order of size
Recruitment, which everyone counts
Advertising, interviewing, the time of whoever runs the process. This is real and it is the number that reaches a spreadsheet, which is exactly why it dominates the conversation.
The ramp, which some people count
A new manager takes months to reach the standard of the one who left, and during that period the venue runs at reduced quality. Some businesses account for this. Most account for it as a fixed cost of turnover rather than as something that varies with how well the handover was done.
The knowledge, which nobody counts
This is the one that hurts, because it does not appear anywhere and it does not recover on its own. The new manager will eventually accumulate their own version of it, over a year or so, by making some of the same mistakes.
Worse, some of the knowledge is not recoverable at all. Nobody else knows what the previous manager noticed about a supplier eighteen months ago, so the business relearns it by being let down again.
Why does the handover not fix it?
Most venues do run a handover. Two weeks of notice, a walkthrough, a document. It captures the procedural material well: keys, codes, suppliers, rotas.
It captures almost none of the judgement, for a simple reason. The departing manager does not know what they know, which is the point the philosopher Michael Polanyi made in The Tacit Dimension (1966): we can know more than we can tell. Knowledge that has become automatic is invisible to the person holding it, and asking someone to write down everything important produces the things they can retrieve on demand, which is the shallow layer.
What actually helps?
Two things, neither of which can be done in the notice period.
Move the recurring judgements into a record while the person is still there. Not a document written at the end, but the ordinary weekly output of running the place: what was noticed, what was decided, what happened next. If a venue keeps that continuously, a departure removes a person rather than a history.
This is the argument for closing the loop on actions rather than just generating reports. A report says what happened. A record of what was decided and whether it worked is the thing a successor can actually read.
Reduce the dependency before it becomes acute. A venue where one person holds everything is fragile regardless of how good that person is. Distributing the standard across roles rather than individuals costs some efficiency in the short term and is the only thing that makes a departure survivable.
The part that is uncomfortable
Strong managers frequently prefer the arrangement where they hold everything. It makes them essential, it is how they demonstrate value, and asking them to write it down can read as a threat.
The framing that works is the one that is also true: a manager whose knowledge is captured can take a holiday without the place slipping, can be promoted without leaving a hole, and is not the reason the business cannot grow. That is a better job, not a smaller one.
If you want to see what a continuous record of decisions and outcomes looks like in practice, that is part of what CoreTAP does: the action, who it went to, and whether it changed anything.
Frequently asked questions
Why doesn't a two week handover protect us?
It captures the procedural material well: keys, codes, suppliers, rotas. It captures almost none of the judgement, because knowledge that has become automatic is invisible to the person holding it.
What knowledge do we actually lose when a manager leaves?
Things like which supplier short delivers, which regulars will complain and which will simply leave, how Friday staffs differently from Saturday, and which staff member is close to leaving. Some of it is not recoverable at all, so the business relearns it by being let down again.
How do I reduce the risk before anyone resigns?
Move recurring judgements into a record as part of the ordinary weekly running of the place, and distribute the standard across roles rather than individuals. Neither can be done in the notice period, and the second costs some efficiency in the short term.
My manager does not want to write things down. How do I raise it?
Strong managers often prefer holding everything, because it makes them essential. The framing that works is also true: a manager whose knowledge is captured can take a holiday without the place slipping and be promoted without leaving a hole.