· Codemax · 6 min read
Short on kitchen staff? Getting more output per man-hour
When you can't hire fast enough, the lever you control is what each paid hour produces. How to measure output per man-hour, where kitchen hours leak, and five fixes that hold.
TL;DR
Output per man-hour is the number of portions, covers or batches a kitchen produces for each paid hour of labour. When a restaurant can't hire fast enough, raising it is the lever it controls: measure it by daypart, roster to the demand curve, move repeated prep to a central kitchen, take ordering, counting and reporting off the kitchen, and cross-train with visual SOPs.
- Measure portions or covers per paid labour hour, by daypart, before changing anything.
- A flat roster hides a crunch at peak and idle hours in the lull.
- Prep repeated in every outlet is often work a central kitchen could batch once.
- Ordering, counting and reporting by hand take hours that should go on cooking.
- Visual SOPs and cross-training keep output steady when someone is off.
When you can’t hire fast enough, the lever you still control is output per man-hour: how much your kitchen produces for every paid hour of labour. Measure it first, by outlet and by daypart. Then go after the hours that don’t produce food: rosters that ignore the demand curve, prep repeated in every outlet, and ordering, counting and reporting done by hand. That is the practical response to the restaurant labour shortage operators across Malaysia keep describing.
The pressure behind Malaysia’s F&B labour shortage (kekurangan pekerja F&B) is well documented. In May, the restaurant associations Primas and Presma said more than 600 Indian and Indian-Muslim restaurants were “on the brink of permanent closure” for lack of workers. In July, Prime Minister Anwar Ibrahim gave ministries a month to resolve restaurant foreign-worker applications, citing “an urgent need to fill vacancies in certain sectors, including the restaurant industry”. And at the TT360 media launch on 3 October, Ethan Choong Wai Hon of Ping Edge Technology Bhd put the operator’s questions in three lines: “I can’t find people. My costs keep going up. How do I keep my food the same across five outlets?”
Hiring is one front. The other is the hours you already pay for.
What is output per man-hour in a kitchen?
Output per man-hour is the amount a kitchen produces for each hour of paid labour: portions per labour hour for an outlet kitchen, covers per labour hour for front of house, kilograms or batches per labour hour for a central kitchen. It is the simplest measure of kitchen productivity, and it tells you whether a labour problem is really a headcount problem or an hours problem.
Take a hypothetical outlet that sells 700 portions on a weekday, with seven kitchen staff on ten-hour days. That is 70 paid labour hours, or 10 portions per labour hour.
Now split the day. Between 12pm and 2pm the kitchen sells 300 portions with all seven people on, which is 14 labour hours, or about 21 portions per labour hour. Between 3pm and 5pm it sells 60 portions with the same 14 hours, or about 4. The daily average of 10 hides both a crunch at lunch and a lull in the afternoon.
Suppose a week of observation then shows roughly two hours a day going on ordering and stock checks, two on end-of-day reporting and four on prep the central kitchen could batch. Take those eight hours out and the same 700 portions need 62 hours: about 11.3 portions per labour hour, or most of a shift freed for the evening peak, without hiring anyone.
A few rules keep the number honest:
- Count paid hours, not rostered hours. Overtime and early starts count.
- Include managers’ admin time. An hour a manager spends texting orders is a kitchen hour.
- Measure by daypart and by station. The daily average is where problems hide.
- Use it to plan, not to blame. Teams that see the number used against them stop recording honestly.
Where do kitchen hours actually go?
Most kitchens lose hours in the same places. Some are the hours lost in hand-offs between systems; the rest sit inside the outlet.
| Where kitchen hours go | Typical waste | Fix |
|---|---|---|
| Flat rosters across the day | Too few hands at peak, idle hands in the lull | Roster to hourly sales and stagger start times |
| Prep repeated in every outlet | Five outlets each making the same sauce, with five different yields | Batch it in the central kitchen or one hub outlet |
| Ordering by walk-around and text | A manager counts shelves and messages the central kitchen every evening | Recommended order quantities from par levels and sales; the outlet reviews and approves |
| Full stock counts | Hours spent rebuilding numbers nobody trusts | Live stock-on-hand, with spot counts on high-value items |
| End-of-day reporting | Sales and wastage re-keyed into a spreadsheet | Captured once, at the POS and at the station |
| Training by shadowing | A senior cook explains the same dish for the tenth time | Image and video SOPs on a phone |
| Rework and remakes | Wrong prep, wrong portion, a dish sent back | Clear recipes and visual standards at the station |
Five fixes for a restaurant labour shortage
Roster to the demand curve. Pull hourly sales for the last four to six weeks for each weekday. Stagger start times so the most hands are on during the two busiest hours, and move prep, cleaning and receiving into the lull. You are not cutting hours. You are putting them where the orders are.
Move repeat prep to the central kitchen. Sauces, marinades, portioned proteins and doughs made once, in volume, take fewer hours than the same work done in every outlet. Without a central kitchen, a batch-prep shift in one hub outlet can supply two or three nearby. The catch is planning: batches have to match what outlets will actually use, or you trade a labour saving for waste.
Take ordering, counting and reporting off the kitchen. When order quantities are recommended from par levels and sales, a manager reviews and approves instead of counting shelves at 10pm. The par levels have to keep up with real demand, or you simply automate the wrong order. Live stock-on-hand turns the weekly full count into spot checks. The wider case for digitalising admin is about giving those hours back to the floor.
Make standards visual. A short video of how a dish is plated, or a photo sequence for a sauce, teaches faster than a laminated sheet and doesn’t need a senior cook standing beside the new hire. It also cuts remakes, which are hours spent cooking the same dish twice.
Cross-train for the peak. Every station should have at least two people who can run it. Start with the station that stops service when one person is missing, then the pass. A simple skills matrix on the office wall shows where you are one sick day away from trouble.
How do you keep the gains?
Track output per man-hour weekly, per outlet and per daypart, next to sales. If it slips at one outlet while sales hold steady, look at the roster and the prep list before you look at the team. If it rises while complaints rise too, you have cut too deep.
Robots will take some physical work off certain stations in time, as the debate around kitchen robots at TT360 shows. Most of the hours you can recover this quarter, though, are in planning and admin.
Where RMS and the AI-Kitchen Command Center fit
Codemax RMS (Resource Management System) takes most of the admin rows in that table off the kitchen. Purchase-order quantities are recommended from par levels and POS data, and outlets order stock from the central kitchen on mobile. Central kitchen production is planned against rolling outlet demand, with demand- and supply-based job scheduling, and pick lists and batch tickets generated automatically. Stock-on-hand stays live through POS sync, and recipes, yields and image or video SOPs are pushed from HQ to every outlet, step by step on a phone.
The AI-Kitchen Command Center watches every outlet, product and shift against its own baseline and flags the ones drifting, so a manager hears about a problem the same week rather than at month-end. It also makes recommendations on stock cover, mix and labour, and sends alerts by email, WhatsApp, Slack or Microsoft Teams. Neither replaces a cook. Both give your cooks back hours they were spending on something else. If you want to map this against one of your own outlets, talk to the team.
A ten-minute check this week
Pick one outlet and one weekday, and answer four questions:
- How many portions did the kitchen sell, and how many paid hours did that take?
- What was the output per labour hour in the busiest two hours, and in the quietest two?
- How many hours went on ordering, counting and reporting?
- Which prep tasks are repeated in every outlet?
If the peak and lull numbers are far apart, or the admin runs to hours, you have found capacity you are already paying for.
Frequently asked questions
How do you calculate output per man-hour in a restaurant kitchen?
Output per man-hour in a restaurant kitchen is total output divided by total paid labour hours for the same period. Use portions sold for an outlet kitchen, covers for front of house, and kilograms or batches for a central kitchen. Count every paid hour, including managers' admin time, and calculate it by daypart as well as by day, because the daily average hides peaks and lulls.
What can a restaurant do about a labour shortage besides hiring?
A restaurant facing a labour shortage can make each paid hour produce more. The practical options are rostering to hourly demand instead of flat shifts, batching repeated prep in a central kitchen or hub outlet, cutting manual ordering, counting and reporting, using image or video SOPs so new staff learn faster, and cross-training so one absence doesn't stop a station.
What is a good output per man-hour for a kitchen?
There is no single good output per man-hour for every kitchen, because the figure depends on the menu, the service style and how much prep arrives ready from a central kitchen. A noodle bar and a fine-dining kitchen will never compare. The useful benchmark is your own: track output per man-hour weekly by outlet and daypart, and compare outlets that run the same menu.
Does a central kitchen reduce labour in outlets?
A central kitchen can reduce outlet labour when it takes over prep that every outlet was repeating, such as sauces, marinades and portioned proteins. Batching that work in one place generally takes fewer hours than doing it in many small kitchens, and outlet staff then spend their hours on cooking and service. Production has to be planned against real outlet demand, or the savings turn into waste.
Which kitchen tasks should be cross-trained first?
Cross-train first on the stations that stop service when one person is missing, usually the busiest cooking station and the pass, then on routine tasks every cook should be able to cover, such as opening checks and receiving deliveries. Short image or video SOPs on a phone make cross-training faster, because staff can follow the standard without a senior cook beside them.
Make every paid hour count.
See how Codemax RMS and the AI-Kitchen Command Center take ordering, counting and prep planning off your kitchen team.
Sources
- 600 Indian, mamak restaurants on verge of closure due to lack of workers, claim associations — The Vibes, 30 May 2026
- Ministers given a month to resolve foreign worker shortage in restaurants — Free Malaysia Today, 24 Jul 2026
- Kitchen robots set to reshape F&B operations — New Straits Times, 4 Oct 2026
Related reading
- Three-day weekend, three-day rush: an outlet checklist for Deepavali Deepavali on Sunday 8 November gives most of Malaysia a three-day weekend. A practical checklist for outlet managers on rosters, stock and on-the-day monitoring, state by state.
- World Food Day: where food waste hides in your operation A stage-by-stage map of where waste hides in a central kitchen and outlet network, what to measure at each stage, and how to start a waste log that people actually use.
- Demand forecasting for multi-outlet F&B: start with the data you have You don't need a data science team to forecast better. A starter method built on the POS history you already hold, from cleaning the data to a weekly review your central kitchen can run.