Manager development for owner-operators who cannot be everywhere

Most owners do the opposite. They delegate the work and keep the judgment. The manager gets the ordering, the schedule, and the closing, but every real decision still routes through a text message to the owner at nine at night.

That arrangement feels safe. It is the reason you cannot take a Saturday off four years in.

It also has a price you can read on your own statement. The National Restaurant Association's Restaurant Operations Data Abstract, 2025 edition, collected financial and operating data from more than 900 restaurant operators nationwide. Salaries and wages including benefits ran a median of 36.5% of sales for full service restaurants in 2024, with operators reporting a pre-tax profit at 34.2% and operators reporting a loss at 42.9%. Limited service sits lower and shows the same shape, at 31.7% overall, 30.0% profitable and 34.1% losing.

The distance between a profitable full service restaurant and a losing one is 8.7 points of labor. Most of that distance is made of decisions taken during a shift by whoever was standing there. If the person standing there is not allowed to decide, the decision does not get made late. It does not get made at all.

Why does your manager keep calling you about things they could decide?

Because calling you has never once been the wrong answer, and deciding alone has occasionally been.

When a manager makes a judgment call you disagree with, you explain why it was wrong, calmly, and teach them exactly one thing: the call carries risk and the phone does not. When they call you first, nothing happens at all.

People do what is safe, and you decide what is safe. If the only unpunished path runs through your phone, that is the path.

The fix is not a speech about ownership. Define a category of decision that is theirs, state the boundary in dollars or hours, and do not overturn them inside it even when you would have chosen differently. A reversed decision inside someone's own territory costs more than the wrong call did.

What is worth handing over first?

Start with decisions that produce a record. You need to be able to review the judgment afterward without having been present, and that is only possible where the work leaves a trail.

Ordering and receiving is usually first. There is an invoice, a count, and a usage number, so a bad call is visible on paper a week later. It also teaches how money moves through the building, which no conversation does. That is why an inventory system sized to a small restaurant is a development tool and not only a cost control.

Second is a report they own. Not a report they run for you, a report they read and act on. Reading the POS for something other than the sales total gives a manager a weekly diagnostic they can act on before you see it, and the habit of arriving at a meeting with a cause rather than a number.

Third is a piece of the guest experience with a real budget attached. Comps, recovery, a small discretionary spend. Give a number and let them work inside it.

What develops nobody: keys, the alarm code, and the right to stay latest. Those are responsibilities, not decisions.

How do you tell whether a manager is developing or just surviving?

Watch what they bring you, not what they finish.

A surviving manager reports events. Slow Tuesday. Two callouts. The dish machine again. Everything they say happened to them.

A developing manager reports causes and choices. Tuesday was slow so I cut a server at seven instead of eight. We were short so I moved prep to the morning. The machine went down and here is what I want to do about it.

The tell is whether the sentence has a decision in it. Not a good decision necessarily. Any decision.

The second tell is what happens when they are wrong. A developing manager tells you about a call that did not work before you find it. A surviving one lets you find it, because in their model of the job, being found out is the same as being fired.

If nobody has ever brought you their own mistake, that is not luck. The cost of confession in your building is higher than the cost of concealment, and you set both prices.

What is one delegated decision actually worth?

More than the conversation about it, and here is the arithmetic so you can run it on your own boundary.

Take the simplest decision on the list: the cut. Your manager may send one person home when sales at a stated hour fall below a stated number, without calling you. That is one sentence and one number.

Say the trigger fires on three shifts a week and each cut is two hours early. Six labor hours a week. Use a loaded hourly cost of $21, meaning wages plus payroll taxes plus whatever benefits you carry, and substitute your own figure because this input varies more than any other.

Six hours at $21 is $126 a week. Across a year, $6,552.

Now convert it into the only unit that makes the point. The Association puts the pre-tax margin of a typical independent restaurant near 5%. At that margin, $6,552 of profit is what a restaurant doing roughly $131,000 in annual sales produces in a year. One delegated decision, one sentence long, is worth more than most of what you will be sold this year.

Run the same conversion on the whole picture and it gets harder to ignore. A restaurant doing $1.6M in net sales at the all respondent median of 36.5% pays $584,000 in labor. The same restaurant at the profitable median of 34.2% pays $547,200. That is $36,800 a year, and it is not a wage decision. It is a stack of small calls made or not made on ordinary nights.

The version of this that fails is the one where you delegate the cut and then ask about it the next morning in a tone. Do that twice and the trigger stops firing.

How do you know whether the delegation is real?

Not by counting how rarely they call you, which is the metric everyone reaches for and the easiest one in the building to fake.

There is a useful parallel from a different industry. In automated phone systems, the headline number vendors report is the containment rate, meaning the share of calls resolved without a human. Practitioners who work on these systems make the same point every time: containment is the metric most easily gamed by making escalation hard. Bury the transfer option and containment goes up while the service gets worse. The stronger signal, and the one no platform reports natively, is the callback within twenty-four hours rate. If a large share of the callers you contained call back the next day, the containment number was a lie.

It is also worth knowing that no credible independent containment benchmark could be verified from the major analyst firms. The numbers in circulation are vendor numbers.

Translate that straight across. Your containment rate is how often your manager decided without calling. Your callback rate is how often the decision came back to you within a week anyway, as a complaint, a reversal, a fire that had to be handled on Saturday, or a thing you discovered yourself.

So track two things for six weeks. How many decisions did they make inside the boundary, and how many of those returned to your desk. A manager making twelve calls a week with one coming back is developing. A manager making two calls a week with none coming back has not been given anything.

Low call volume is not evidence of a good manager. It is evidence of a narrow boundary or a frightened one.

Why does this get expensive at the second location?

Because at one location you are the system. You catch things by walking past them.

At two, half of your walking past goes away, and everything you never wrote down becomes an accident waiting for a Friday. Most of what changes when you go from one location to two is this, and there is no version of that transfer that happens in the month before you open.

The manager who will run the second store has to be running the first one, unsupervised, well before the lease is signed. If you are still the decision maker at store one on opening day at store two, you have two stores without a manager.

There is a smaller version of the same test. Something changes on the menu, and now someone has to make sure the printed menu, the POS, and the website all agree. Handing a manager the whole loop, including keeping seasonal menu changes reflected on the website, tells you quickly whether they finish things nobody is watching.

What to do this week

Write down every decision you made in the building this week. Just the list. Most owners are surprised by the length and by how many items are worth less than a hundred dollars.

Circle three a manager could make with a stated boundary. Ordering inside a dollar limit. Comps inside a nightly cap. Cutting the floor inside a sales trigger.

Tell that manager the boundary in one sentence, in writing, and tell them you will not overturn them inside it. Then do not overturn them inside it.

Set a fifteen minute weekly review where they walk you through the calls they made and why. Ask what they would do differently. Do not correct in the moment unless money is still on the table.

Do that for six weeks before you add a fourth decision.

Questions about handing off decisions?

Email me at eric@seod.com with the list of decisions that currently only you can make. Just the list. I will tell you which three I would hand over first, what boundary I would put on each, and which one I would keep no matter how good the manager is.

Sixteen years running multi unit restaurants and more than $54M in annual P&L means I promoted a lot of people, and got it wrong often enough to know that the failure was almost always mine. If part of the handoff is a report the manager has to read rather than run, an explanation of Google Analytics 4 for people who did not ask for it is the marketing version of the same problem.

Otherwise, keep going with more on running the operation.