Study · what the numbers say

Nine points. That's what separates the restaurants that make money from the ones that lose it.

I didn't do this study. I read it for you: the surveys from the National Restaurant Association and Restaurants Canada, an experiment run in 28 stores by three universities, the biggest meta-analysis ever done on employee engagement, and an old Cornell calculation that hasn't aged a day. The nine points in the title are nine points of labour cost, as a share of sales: that's the gap between the restaurants that make money and the ones that lose it, not a list of nine items. Put end to end, these numbers say one thing: what decides your profitability isn't your menu, it's how you manage your team, day after day. Every number has its source at the bottom of the page.

1. The finding: labour is the line that separates the two camps

That's the National Restaurant Association's annual survey, more than 900 American restaurateurs, 2025 edition¹. Between a profitable restaurant and one running at a loss, the labour gap is nearly nine points of sales. The ones losing money pay more than six points above the median; the profitable ones, more than two points below. In quick service, same story, tighter: 30.0% for those who win, 34.1% for those who lose¹.

Why is nine points huge? Because the margin is only five. Before the pandemic, food and labour each took about 33 cents of every sales dollar, other expenses 29 cents, and roughly 5% was left before tax². A restaurant that lets its labour slide from 34 to 43% isn't trimming its margin: it erases it, then digs.

2. And it's tightening: nobody has a cushion anymore

To keep the same 5% margin as in 2019, a typical American restaurant would have to sell 36% more². Nobody sells 36% more. So the margin comes from somewhere else: the hours.

In Canada, it's the same slope. In November 2025, 44% of restaurants were operating at a loss or at break even, and 89% of operators named labour as their number one concern, ahead of food³. In 2026, 36% are still at a loss or at break even, three times the 2019 level, and 71% report falling profitability⁴. This isn't a bad year. This is the new ground.

3. In your place, in dollars

One point of labour is one percent of your sales. Pick what looks like you; I do the rest.

4. Why it slips, even with good managers

Nobody decides to pay nine points too many. It happens through four small doors, always the same ones.

The schedule is made ahead of time, without looking at sales. Last week gets copied over, one person gets added "just in case", and the rainy Saturday looks like the sunny one. Nobody cuts at 2 p.m. The rush is over, three people are still on the floor for six customers, and everyone waits for the manager to say it. The number comes too late. The accountant's monthly report says there was a problem five weeks ago; nobody remembers which day. The target doesn't exist. Without a percentage on display and a colour every morning, "a lot of people" stays an impression, not a number.

The common thread: the information exists (sales are in the register, hours are on the time clock), but it doesn't reach the right person on the right morning. It's a rhythm problem, not a skill problem.

5. What the science says about the schedule itself

The Stable Scheduling Study⁵ is the most serious experiment ever run on scheduling floor employees: 28 stores of a large clothing chain, 19 of which changed the way they scheduled and 9 served as controls, in San Francisco and Chicago, from November 2015 to August 2016, led by researchers from the University of California, the University of Chicago and the University of North Carolina. What they changed: the schedule published two weeks ahead, the end of unannounced on-call shifts, shifts kept stable from one week to the next, shift swaps through an app, and guaranteed hours for those who wanted them. Result: sales went up 7%, productivity 5%, without one more dollar in marketing. It's retail, not restaurants, and that's why I cite it with its real label; but a floor is a floor, and the schedule two weeks ahead with confirmation is exactly what I do.

6. A team that stays, a team that cares

Cornell's calculation⁷ adds up what gets forgotten: the posting, the interviews, the paperwork, the training, the mistakes of the first weeks, and the productivity lost while the position sits empty. A restaurant of twelve employees that replaces six a year, which is ordinary in this trade, spends the equivalent of one full-time salary just keeping the door revolving. And Gallup⁶, across 183,806 business units and 736 studies, measures the other end: teams in the top quartile of engagement make 23% more profit than those in the bottom quartile, with fewer absences, less turnover, less theft, fewer accidents. Engagement, in a restaurant, starts with dull things: knowing your schedule, being paid right, having a file in order, a written procedure when you don't know, and someone who notices when you do well.

7. What that means, concretely, in the morning

Everything above fits in one sentence: a restaurant's profitability is decided in the daily management of its team and its hours, and it gets lost for want of a number at the right time. That's why I wasn't built as a report, but as a morning colleague.

The number comes the next day, not the next month. Your manager enters yesterday in two minutes, I give a green, yellow or red verdict, with labour as a percentage against your target. The cut becomes a decision, not an awkward moment. Every morning I compute the sales goal that makes the day green; below it, you cut, and I say how many hours. The schedule goes out two weeks ahead, confirmed by everyone: I track it in the management timeline and flag it when it slips; and with someone on call for service and for the kitchen, two different people, or I refuse to publish. Employee files are in order: hiring in nine steps, signed handbook, tip agreement, vacations, recognitions and disciplinary follow-up, in order. And the hours you cut, I show them to you in black and white, week after week, with what they're worth. Half an hour of floor time a day, seven days a week at $18 an hour, pays my subscription twice over; the rest is yours.

Sources

Checked September 21, 2026. The American numbers come from surveys of operators, not audited books; the medians are the ones published by the associations. When a study comes from a sector other than restaurants, it says so. If a number changes at the next edition, this page changes too.

1. National Restaurant Association. Elevated labor costs had a significant impact on restaurant profitability in 2024 (Restaurant Operations Data Abstract, 2025 edition, more than 900 restaurateurs). restaurant.org
2. National Restaurant Association. Elevated costs continue to pressure restaurant profitability (2026). restaurant.org
3. Restaurants Canada. Industry outlook report, February 12, 2026 (November 2025 data). www.restaurantscanada.org
4. Restaurants Canada. Industry outlook report, May 4, 2026. www.restaurantscanada.org
5. Williams, Lambert, Kesavan et al. (UC Hastings, University of Chicago, UNC Kenan-Flagler). The Stable Scheduling Study, 2018: 28 stores, San Francisco and Chicago, November 2015 to August 2016. worklifelaw.org
6. Gallup. Meta-analysis of employee engagement (183,806 business units, 736 studies, 2024). www.gallup.com
7. Hinkin and Tracey, Cornell (Cornell Hotel and Restaurant Administration Quarterly). The cost of turnover: US$5,864 per hourly employee replaced, a figure the whole industry has quoted since. higherme.com