Staff Turnover Costs More Than the Raise You Skipped

A line cook asks for a dollar more an hour. You say no, because a dollar an hour across a full-time schedule is real money and you have already watched food cost climb twice this year. Three weeks later he gives notice. You post the job the same night and tell yourself you'll find someone cheaper.

You probably will find someone. What you won't do is write down what the replacement cost you, because that number never arrives as an invoice. It arrives as a hundred small leaks spread across six weeks, and by the time it's done you've forgotten what started it.

So write it down. Here is how to cost a replacement honestly, using your own numbers instead of someone else's.

Add up the hours before the new person starts

Start with your time, priced at whatever your hour is actually worth. Not minimum wage — the rate at which you'd pay someone else to do the thing you're not doing while you hire.

Writing and posting the ad. Sorting the pile that comes back, most of which is wrong. Texting people to book interviews. The interviews that happen. The interviews that don't, because two of the four don't show. Reference calls, if you still make them. A trial shift that you have to supervise, which means you're on the floor watching instead of ordering, prepping, or going home.

For a line cook or a stylist or a service tech, this is rarely under eight hours of someone's attention, and usually it's the owner's attention, which is the most expensive kind you have. If you're hiring for a role that needs a licence or a specific skill, you're hunting longer and you're interviewing more people to find fewer candidates.

Now add the shifts between the last day and the start date. Somebody covers those. Either it's overtime, or it's you, or it's a thinner schedule and a worse night.

Add the training, and be honest that it's two wages for one job

For the first stretch, you are paying two people to do one person's work. The new hire is on payroll producing almost nothing. The trainer is on payroll producing less than usual because they're explaining instead of working. That's not one wage, it's roughly one and a half to two wages for however long the overlap runs.

How long does it run? Be truthful. Not "they picked it up fast." Count the days until the person could run the station alone on a Friday without anyone checking. In a kitchen that's usually weeks, not days. At a front desk it's the moment they can handle an angry customer without getting you. In a salon it's when the chair fills on its own.

Add the mistakes during ramp, because they land on customers

This is the part nobody writes down, and it's the biggest part.

A new cook sends the wrong plate. You comp it. A new server rings in a modifier wrong and the ticket goes out twice. You eat the food cost twice. A new receptionist books two clients into one slot and one of them leaves. A new tech quotes a job wrong and you honour the quote because you're not going to argue with a customer over your own staffing problem.

Each of these is small. They happen for weeks. And a few of them turn into a one-star review written by someone who doesn't care that it was a trainee's third shift — we wrote about handling the review that isn't fair, but the better move is having fewer of them to handle.

Tickets also go out slower while someone's learning, which means tables turn slower, which means fewer covers on a night you were counting on. If you've already read how table turnover is really a kitchen question, you know how fast that compounds.

Add the pressure on everyone who stayed

Here's the cost that makes turnover self-feeding. While you're short, your best people absorb it. They pick up the extra shift. They work the station down a body. They train the new hire on top of their own job, usually for no extra money and no acknowledgment.

Your best people are also the ones with the easiest time getting hired somewhere else. They absorb it once as a favour. The second time, they start checking what else is out there. That's how one departure becomes three over a season — not because anyone was disloyal, but because you spent their goodwill covering a hole you created.

Now total the whole thing: your hours, the covering shifts, the double wages during overlap, the comps and the slow service, the goodwill. Compare it to a dollar an hour. On almost any full-time role, the raise is cheaper, and it's cheaper by a lot, and it's not close.

But the raise usually isn't what they wanted

That's the twist. People name money because money is the easiest thing to say out loud in a tense conversation. It is rarely the whole reason, and matching the number often doesn't keep them.

What keeps people, over and over, is two things that cost you very little.

Schedule predictability. Posting the schedule late is one of the most expensive habits in a small business. A person who doesn't know their Saturday until Thursday can't book a dentist appointment, can't commit to a class, can't plan anything with their kids. They will leave for a job that pays the same and posts two weeks out. Pick a day. Post on that day. Every week, even when it's annoying. If you need to change a posted shift, ask instead of announcing.

The second half of that is not cutting people without warning. If Tuesday is dead and you're trimming the schedule, the person losing hours needs to hear it from you before they see it on the sheet. Cutting hours silently reads as "you're being pushed out," and people respond by finding something else. Our post on treating the schedule as a labour cost decision covers how to trim without doing that damage.

Being told things. Staff find out the menu changed when a customer asks about an item that no longer exists. They find out hours changed when they show up. They find out about the new POS the morning it's installed. Every one of those is a small message that says the people doing the work are the last to know.

Fix it with something dull and consistent. A group chat you actually use. Five minutes before service where you say what changed. A note by the clock-in. It doesn't have to be elegant, it has to be reliable.

The things that aren't money but look like it

• Tell people where the road goes. "Prep to line in six months if you want it" is a reason to stay. Nobody stays for a job that is exactly the same job in two years.
• Pay the trainer. A small bump for the shifts they're training on costs you almost nothing and turns a resented chore into a role people want.
• Make the tip math transparent. Most tip fights are fairness fights, and fairness fights end in resignations. A tip pool that doesn't start fights is retention work.
• Fix the thing they complain about. The broken fryer. The printer that jams every Friday. The booking system that double-books. Fixing one of those publicly does more for morale than a speech.
• Say thank you on the night it happened. Not at the Christmas party.

Take the busywork off them

Some of the worst parts of a job are the parts nobody should be doing. Retyping online orders into the POS. Answering the same phone call forty times a shift while the dining room waits. Chasing people for appointment confirmations.

Those tasks don't just cost you labour hours, they cost you goodwill, because they're tedious and your staff know they're tedious. When we build ordering and phone systems, the retention effect is usually the part owners don't expect: the shift gets less aggravating, and the person who was going to quit over the phone ringing stops thinking about quitting.

The exit interview you should actually do

When someone resigns, the money conversation is over and you'll finally get a straight answer. Ask three questions and shut up afterward:

1. What made you start looking?
2. What would have made you stay, other than money?
3. What should I tell the next person on day one that I didn't tell you?

Write down what they say. After three departures, the same answer will show up twice. That answer is your actual turnover cost, and it's almost never a dollar an hour.