Coach The Bench

Promoting From Within Without Promoting the Wrong Person

Promoting the best line cook to shift lead is how you lose your best line cook and gain a mediocre shift lead. A five-test screen so you stop making that trade.

Kiyong (Eric) Lee·Operations Executive·August 2026·10 min read

The short version

The skills that make someone great at the doing are almost never the skills that make them great at the leading. Promoting the top individual contributor is the default mistake.
Run every internal promotion through a five-test screen: consistency, coaching, bad day, cross-team, and boredom.
A candidate who passes four of the five is ready. A candidate who passes only one, no matter how strong that one is, is not.
Tenure, being liked, and volume of output are the three loudest false positives. None of them predict leadership.
How you talk to the candidate you did not promote matters more than how you talk to the one you did.

I have promoted around 60 people from within in my career. Around 45 of those promotions worked. Around 15 did not, and of those 15, almost all followed the same pattern. I promoted the best individual contributor on the team, they stopped being a great individual contributor, they became a mediocre leader, and I lost both.

What follows is the screen I now run before I make an internal promotion, especially at the shift lead to assistant manager or assistant to general manager jump. It is five tests, in a specific order. It takes about two weeks of focused observation to run properly, and it has cut my bad-promotion rate to something close to zero over the last few years.

Why promoting from within goes wrong

Promoting from within is not the problem. It is the right move most of the time. Retention is higher, morale is higher, and the person already understands the culture, the brand, and the customers. What goes wrong is not the strategy. It is the selection method.

The default selection method in most field operations is: pick the person whose numbers are best or whose tenure is longest or who everyone likes most. All three of those criteria are correlated with being a good individual contributor. None of the three predicts whether the person can lead other people. Leadership is a different set of muscles, and the only way to know if someone has those muscles is to test for them specifically.

The line cook you love is often the shift lead you will fire in six months. Being great at the work and being great at multiplying the work are two different jobs. Screen for the second one.

The five-test screen

Five specific tests. Run all five before any promotion above the line-lead level. Pass four out of five and the person is ready. Pass three out of five and you have a development plan, not a promotion. Pass two or fewer and you have your answer.

Fig. 1 · Pass four of five and they are ready. Three and they need a plan. Two and you have your answer.

Test 1: The consistency test

Pull twelve weeks of their key numbers. Sales per labor hour, guest satisfaction score, waste per shift, whatever the role is judged on. Look at the shape of the line, not the average. What you want is boring. Consistent, slightly above average, low variance. What you do not want is a jagged sawtooth with two brilliant weeks and three below-average weeks.

Spikes tell you the person is riding a wave, usually a good crew or an easy season. Boring tells you they are building a system that produces the same result on a Monday and a Saturday. Field leadership is the ability to produce the same result under different conditions. That is what you are trying to buy in the promotion.

Test 2: The coaching test

Ask them to teach a piece of their job to a new hire while you watch. In a restaurant, this is usually a station handoff or a prep task. What you are watching for: can they break the work into steps a stranger can follow, can they wait patiently while the new hire fumbles, can they resist the urge to take the knife or the pan out of the new hire's hand.

The person who cannot let the new hire fumble cannot lead a team. They will do the work themselves forever. They will resent their team for being slower than they are. They will burn out inside twelve months. Coaching is the single most predictive test of the five, and it is the one that catches the most false positives.

Test 3: The bad-day test

Watch them on a shift where something is going seriously wrong. The dishwasher is down. Two servers called out. A big table is unhappy. Do not intervene. Watch.

You are looking for one specific tell. Do they get quieter and more surgical, or louder and more blaming? A future leader gets quieter, moves closer to the problem, starts making decisions with less noise around them. A non-leader gets louder, faster, more visibly frustrated, and starts assigning blame in front of the team. Both look like effort. Only one predicts leadership.

Test 4: The cross-team test

Ask two departments the candidate does not report into what it is like to work with them. In a restaurant that usually means: ask the kitchen if the candidate is a good front-of-house partner, ask front-of-house if the candidate is a good kitchen partner, ask maintenance if they are treated well when they show up.

Cross-team trust is the single best predictor I have found of how someone will lead across functions later, and it costs almost nothing to check. The candidate who is beloved by their own team but disliked by every adjacent team is going to build a silo, not a bench. Every promotion I have regretted failed this test and I found out after, not before, because I did not bother to ask.

Test 5: The boredom test

The last test is the simplest. Are they already doing 20 to 30 percent of the next role, unpaid and unasked, because they are bored in the current one? The shift lead who is already writing the next week's schedule for their GM without being asked. The prep cook who is already teaching the new hires how to break down a case of chicken. The server who is already covering the reservation book when the host steps away.

The boredom signal is real. If someone is bored in their current role and voluntarily reaching for the next role's work, promoting them is confirming what they have already told you. If they are doing exactly and only their job description, they are either happy where they are, which is fine, or they are not curious about the next role, which is also information. Either way, they are not the promotion.

A prep lead who almost got it

Third year at Zareen's. There was a prep lead at the Mountain View location, call him Miguel. Best knife on the floor. Passed every kitchen skills test we had. Native Spanish, functional English, the kitchen team followed him around like ducklings. Everyone assumed he was the next assistant kitchen manager.

I ran the five-test screen on him one December when the slot opened up. Consistency: perfect, 14 weeks of top-quartile output with almost no variance. Bad day: excellent, when the walk-in compressor died on a Saturday morning he moved fast and quietly and had a plan inside 20 minutes. Cross-team: strong, the front-of-house team had asked me if we could put him on the private events team because of how easy he was to work with. Boredom: also strong, he was already showing new hires how to break down a whole octopus without being asked.

Then the coaching test. I asked him to teach a new hire how to butcher a chicken while I stood at the back of the kitchen. Miguel got two minutes into the demo, saw the new hire hesitate on the first cut, and quietly took the knife back and finished the demo alone. He did not do it meanly. He did it because he could not stand watching a bird get cut wrong. Fifteen minutes later he was still on the same bird and the new hire was still standing next to him watching. Perfect chicken. No teaching had happened.

Four out of five. I did not promote him. I told him exactly why, in Spanish, at the back of the kitchen, that afternoon. I told him he had passed the four hardest tests. I told him the one he had not passed was the difference between being the best cook in the room and being the person who makes the whole kitchen better. I gave him a six-month plan: two new hires a month, coach them through the same three tasks, hands off the knife, come tell me at the end of each hire how it went.

Seven months later, promotion. He is now the executive sous chef at that location. Every new prep hire in the region goes through him. The delay cost us nothing. Promoting him early would have cost us a great cook and produced a frustrated lead who could not let go of the knife.

The candidate you slow down and coach becomes the leader you keep. The candidate you promote too early is the one you write a farewell email for eleven months later.

The three false positives

Three signals that look like promotion readiness and are not. If you catch yourself using any of them as your primary reason, stop and re-run the five-test screen instead.

Tenure. Someone has been at the company eight years. That is not a reason. That is a fact. Tenure predicts loyalty, not leadership. Reward it with pay and recognition, not with a promotion that will burn them out.
Being liked. The person everyone likes is often the person who cannot say a hard no. The role you are promoting them into is going to require them to say hard nos every week, in front of the people who currently like them. If they cannot do that, being liked will turn into being resented inside a quarter.
Volume of output. The top producer on the shift is the best individual contributor on the shift. Those are the same statement. Leadership is not about your own output. It is about multiplying everyone else's. High volume is not high leverage.

How to talk to the person you did not promote

The five-test screen creates a specific problem. You are going to have a strong internal candidate who does not pass, and you are going to have to tell them so. If you handle this conversation badly, you lose the person. If you handle it well, you get another year of great work from them and, in many cases, a real promotion twelve months later.

The conversation has four parts. First, tell them yourself before they hear about it in the break room. Second, be specific about what they passed. The five-test screen gives you the language. Third, be specific about what they did not pass, with one concrete example. Fourth, hand them a written six-month development plan with two things they need to demonstrate, and put a monthly check-in on the calendar.

Skip any of those four steps and you will lose them. All four, and you will keep most of them, and about a third of them will pass the screen next time.

The point

Promoting from within is the right default. It builds culture, it builds retention, it builds the bench. But it works only when the selection method actually selects for leadership. The default of promoting the top individual contributor is not that method. It is a habit disguised as a decision.

Run the five tests. Consistency, coaching, bad day, cross-team, boredom. Pass four of five. Take the two weeks to observe properly. And when the strongest candidate does not pass, have the four-part conversation before you make anyone else the offer. Your bench will build slower and hold longer. That is the trade you want.

Building a promotion pipeline?

I work with a small number of multi-unit operators each year on bench strength, promotion screening, and the shift lead to assistant manager to general manager jump. If you have made a bad promotion recently, or you are about to, I would trade notes.

Get in touch

Questions people ask

What is the biggest mistake in promoting from within?

Promoting the best individual contributor into the leadership role and losing both. You lose the individual contributor because they no longer do the work they were great at, and you get a mediocre leader because the skills that made them great at the doing are not the skills that make anyone great at the leading. Promoting from within works only when you have screened for leadership signals, not performance signals.

How do you know if someone is ready for a promotion?

Run the five-test screen. Consistency over 12 weeks. Ability to teach the job. Behavior on a bad day. Trust across other departments. And whether they are already doing 20 to 30 percent of the next role without being asked. A candidate who passes four of the five is ready. A candidate who passes only one, no matter how strong that one is, is not.

Should you promote your best line cook or your best server to a lead role?

Only if they pass the coaching test. Great line cooks and great servers get to be great by refining a personal craft. Great leads get there by making other people faster. Those are different skills. If your best line cook cannot teach a new hire how to butcher a chicken without taking the knife out of their hands, they are not going to lead a bench either.

What are the false positives in promotion decisions?

Three big ones. Tenure. Being liked. And volume of output. None of those predict leadership. Someone can be at the company eight years, be everyone's favorite, and be the top producer in the shift, and still be a bad manager because they cannot delegate, cannot coach, and cannot hold a hard conversation with a friend.

How long should someone hold a role before you promote them?

Long enough to see them through one full cycle of the role. In restaurant field ops that usually means a full year, because you need to see them through a holiday rush, a slow season, a full inventory cycle, and at least one difficult HR event. Promoting on nine months of visible performance and skipping the hard quarter is where most bad promotions get made.

What do you do when the strongest internal candidate is not ready?

Say so, out loud, with specifics. Tell them what they passed on the screen, what they did not pass on the screen, and what has to change for a re-look in six months. Then either hire externally or promote a weaker candidate who passes the screen. Do not stretch a role for the wrong person because you are afraid of a hard conversation. That is how you lose the person and the role.

How do you protect the person who did not get promoted?

You tell them yourself, before they hear it in the parking lot. You tell them what specifically got the other person the role. You give them a written six-month development plan with the one or two specific things they need to demonstrate. And you follow up on it every month. If you skip any of those steps, you are going to lose them, and losing them costs more than the promotion decision.