Covers per labour hour: the kitchen's real metric

Two Fridays in a row, the kitchen served about 140 covers at dinner. On the first, five cooks worked the service and tickets flowed. On the second, seven cooks worked it, two of them spent half the night wiping down already clean surfaces, and the labour bill was higher for exactly the same output. Nobody noticed, because the covers were the same and nobody looked at the hours. Covers per labour hour would have shown the difference at once.

Covers per labour hour is a simple number: the covers a kitchen served divided by the hours its team worked. It tells you how much output each hour of labour produced. Watched over time, service by service, it shows when a kitchen is overstaffed, when it is stretched, and how much capacity it really has before tickets start to slow.

This guide covers how to calculate covers per labour hour, how to read it by service and by section, what it says about staffing and kitchen capacity, how it links to sales per labour hour and labour cost, and how to use it without turning it into a stick to beat the team with. Every venue and figure in it is an example.

Covers per labour hour measures output per hour worked. If a kitchen team works 40 hours across a dinner service, including prep, and serves 120 covers, it produced 3 covers for every hour worked. If the same team serves 160 covers in the same hours, it produced 4.

It is a form of labour productivity, the general measure of output per hour worked used across many industries. For a kitchen, covers are a natural measure of output, because they reflect the number of guests the kitchen fed.

Most kitchens are judged on food cost and ticket times, and both matter. But neither says whether the kitchen had the right number of people. This measure does. It connects the kitchen's biggest controllable cost, labour, to the thing the kitchen exists to do, which is feed guests.

It also travels well across services. A busy Saturday and a quiet Tuesday have very different cover counts, but their output per hour can be compared directly, which shows whether the rota matched each one.

Count covers for each service: the number of guests served, from the booking system or the till. Use the same source every time, and decide how to count guests who only had drinks or a dessert, so the number means the same thing each week.

Keep services separate. Lunch and dinner have different patterns, different menus and often different team sizes, and combining them hides useful differences.

Count the hours the kitchen team worked for each service, from the rota or the timesheets: the actual hours, not the planned ones, including any overtime. Decide what to include, such as prep before service and cleaning after, and keep it the same each week.

Decide too whose hours count. For a kitchen measure, include everyone working in the kitchen: chefs, cooks, porters if you choose. Some venues measure cooks only, to focus on production. Either works, as long as it stays consistent.

The sum is covers divided by labour hours. A dinner service with 140 covers and 35 kitchen hours produced 4 covers an hour. One with 140 covers and 49 hours produced about 2.9. The second used 14 more hours for the same output.

Work it out for each service over a few weeks, and you will see your kitchen's normal range. That range, not any outside figure, is the benchmark that matters.

Look at the figure service by service. Quiet services often have lower figures, because a minimum team is needed whatever the covers; busy services often have higher ones. That is normal. What matters is whether each service sits within its usual range.

A quiet lunch at 1.5 covers an hour worked may be fine if it is always that way and the minimum team cannot be smaller. A Friday dinner that drops from its usual 4 to 3 deserves a look, because something has changed.

For larger kitchens, the figure can be broken down by section, using the dishes or tickets each station handles. A grill that cooks 90 mains in a service with two cooks for four hours handled about 11 mains per labour hour; a pastry section that plated 60 desserts with one cook for four hours handled 15.

Section figures are rougher, because stations share work and help each other. But they show where hours are concentrated and whether one station has more people than its workload needs, or fewer.

Staff productivity is what this number measures, but it is a measure of the system, not of individuals. A low figure usually means the rota had too many people for the trade, not that the team worked slowly. A high figure can mean an efficient kitchen, or one about to be overwhelmed.

Treat the number as information about how the kitchen was set up for that service: the rota, the menu, the layout and the trade. It is a tool for planning, not for judging people.

On its own, the figure says how much output each hour produced. Read alongside ticket times, it says much more. A figure that falls while ticket times stay good suggests the kitchen had more people than it needed. A figure that rises while ticket times slow suggests it had too few.

The useful range for your kitchen is the band where ticket times stay comfortable and nobody is standing idle. Finding that band, service by service, is the point of measuring.

An overstaffed service shows a low figure, often well below its usual range, with ticket times as fast as normal or faster, and cooks finding things to do. Labour cost for that service is higher than it needs to be.

When you see it repeatedly on the same service, trim the rota: one fewer cook, a later start, an earlier finish. Watch the next few weeks to make sure ticket times stay where they should.

An understaffed service shows a high figure, above its usual range, with ticket times rising, more remakes and a team under strain. It looks efficient on paper and feels chaotic in the kitchen.

High figures are not always good news. If they come with slow tickets and a tired team, the kitchen is past its comfortable capacity. Our guide to understaffing costs more than the shift you saved covers what that costs.

Covers per labour hour helps measure kitchen capacity: how many covers the kitchen can produce in an hour, with a given team, before tickets slow. Look at the services where ticket times stayed good and the figure was highest. That is close to the kitchen's comfortable capacity with that team.

Knowing capacity helps with bookings as well as staffing. If the kitchen comfortably handles 40 covers an hour with five cooks, booking 70 covers into one hour on a Friday will overwhelm it, however well it is staffed.

Always read covers per labour hour alongside ticket times. Ticket times show whether the kitchen kept up; the output figure shows how many people it took to keep up. Together, they show whether the kitchen was overstaffed, understaffed or about right.

Our guide to ticket times: the only kitchen metric that moves covers how to measure them. A kitchen that tracks both can find its right staffing level with confidence.

Prep hours, the time spent preparing before service, are part of the labour that produces covers, but they behave differently. Prep depends on the menu and the forecast more than on the minute-by-minute flow of service. Including it in the measure is right, but it helps to see it separately too.

If the figure falls, check whether service hours or prep hours rose. A rise in prep may come from a more complex menu or from over-prepping; a rise in service hours points to the rota.

Sales per labour hour is the money version: sales divided by labour hours. It accounts for how much guests spend, not just how many there are. A kitchen that serves fewer covers but more expensive dishes may have lower output per hour in covers and higher sales per labour hour.

Use both. Covers per labour hour suits the kitchen, because cooking effort follows guests more than prices. Sales per labour hour suits the whole venue's finances. Our guide to sales per labour hour, by shift and by section covers it in detail.

Covers per labour hour connects directly to labour cost. Divide the average hourly cost of the kitchen team by covers per hour worked, and you have the kitchen labour cost per cover. At an average of fifteen pounds an hour and 4 covers an hour, each cover carries about £3.75 of kitchen labour.

That figure can be compared with the spend per cover and the food cost, to see how much of each guest's spend goes on the kitchen's labour. Our guide to labour cost percentage, by service not by month covers the percentage view.

Resist comparing your figure with numbers from other venues or articles. Kitchens differ too much in menu, layout, equipment and service style for outside numbers to mean much. A fine-dining kitchen will always produce fewer covers per hour than a pizza kitchen, and both can be well run.

Compare with your own past instead: the same service, week by week and season by season. A falling trend on Friday dinners, or a gap between two similar services, is far more useful than any outside benchmark.

The menu shapes kitchen output per hour more than almost anything else. Complex dishes with many components, dishes cooked to order from raw and long menus with many different preparations all need more hours per cover. Simpler menus, with more prepared in advance, need fewer.

When a new menu arrives, watch the figure. If it falls sharply, the new dishes may be taking more labour than planned. That may be worth it for the right dishes, but it should be a decision, not a surprise.

Equipment and layout limit how productive a kitchen can be. A single grill, a pass too far from the stations, a dish-wash that backs up at nine: each caps how many covers the team can produce, however many people are on.

If the figure stalls at a certain level and adding people does not raise output, look for a bottleneck. Our guide to kitchen bottlenecks: finding the station that stalls covers how to find it.

Skill matters. A trained, settled team produces more covers per hour than one full of new starters, with fewer remakes and less time lost to questions. When the team changes, expect the figure to dip and then recover as people learn.

Investing in training, written recipes and clear station set-ups raises productivity without asking anyone to rush. It is usually the best long-term way to improve the number.

Once you know each service's comfortable range, use it to build the rota. Divide the forecast covers by the target figure for that service, and you have the hours the kitchen needs for that service. If Friday dinner is forecast at 140 covers and the comfortable range is about 4, the kitchen needs around 35 hours.

That turns the rota from habit into a calculation. Our guide to staff scheduling against the trading curve covers matching those hours to the busy part of the service.

Do not use the number to push individuals to work faster, or to compare cooks with each other. The number measures the whole system, and using it to blame people damages morale and encourages corners to be cut on quality and safety.

Do not chase the highest possible number, either. A kitchen run permanently at the top of its range has no slack for a busy night, a sick day or a new starter. The aim is the right number for each service, not the biggest one.

The same idea works for front of house: covers per server hour, or covers per front-of-house hour. It shows whether the floor had the right number of people for the guests it served, and works best read alongside measures of service, such as how quickly tables are greeted and how long guests wait for the bill.

Floor service depends on more than numbers, such as the layout of the room and the style of service, so treat the floor figure with the same care as the kitchen one: as a planning tool, compared with your own past.

Each week, look at the figure for each service alongside ticket times. Note any service outside its usual range, and ask why: a quiet night with a full rota, a busy night with a short one, a new menu, a new starter, a broken piece of equipment.

Adjust next week's rota where the pattern repeats. Over a few months, labour hours match trade more closely, labour cost falls on quiet services and busy services get the people they need.

A refit, new equipment, a different pass or a move to a new menu format changes how much the kitchen can produce in an hour. After any such change, watch output per hour for a few weeks to find the new normal, rather than judging the kitchen against the old range.

Some changes raise it, such as a second grill that removes a bottleneck; others lower it for a while, as the team learns. Either way, the new range becomes the benchmark for the rota.

Share the figure with the head chef and the team, along with ticket times. Chefs often welcome a number that shows when a service had too many people or too few, because it supports their own sense of how the night went.

Ask for their view when the figure looks odd. A chef may know that one cook spent the evening on prep for tomorrow, or that a new starter was shadowing, which explains a low figure better than any report.

An invented bistro's kitchen produces 140 covers on most Friday dinners. Over six weeks, the head chef works out covers per hour worked for each Friday. On the four Fridays with five cooks, it averages about 4, with comfortable ticket times. On the two with seven cooks, it averages about 2.9, with ticket times no faster.

On Tuesday dinners, with four cooks for about 60 covers, it averages around 2, and tickets are very quick. The chef sets Friday at five cooks, trims Tuesday to three, with a fourth on call for large bookings, and watches ticket times. The kitchen's weekly labour hours fall noticeably, and ticket times stay where they were.

servearound does not record hours worked: there are no timesheets or clocking in, and rotas are coming, so for now the labour hours in this guide come from your rota or payroll. What it gives the other half of the sum is the output: reservation analytics show covers by day, by meal period and by day and hour, and kitchen analytics show tickets completed by station, with each station's share, median queue and prep time, P90 total and remakes.

Kitchen analytics also show ticket time by day and hour, queue depth against ticket time and which items cost the most time, so a change in output per hour can be read against whether tickets slowed. See Analytics.

Covers served divided by the hours the team worked, for a service.

It shows whether the kitchen had the right number of people for the trade.

Everyone working in the kitchen, or cooks only, as long as it stays consistent.

There is no universal one. Find your own range for each service, where ticket times stay comfortable.

Often that the service was overstaffed for its trade.

No. With slow tickets, it means the kitchen is stretched.

Divide the average hourly cost by covers per hour worked to get kitchen labour cost per cover.

No. It measures the whole system, not people.

Covers per labour hour is the kitchen's real metric: covers served divided by the hours the team actually worked, for each service. Work it out for a few weeks to find each service's comfortable range, and read it alongside ticket times. A low figure with good ticket times means overstaffing; a high figure with slow tickets means the kitchen is stretched.

Use it to find your kitchen capacity, to build the rota from the forecast and to link labour hours to labour cost per cover and sales per labour hour. Compare with your own past rather than other venues, watch it when the menu or team changes, look for bottlenecks when it stalls and never use it to blame individuals.