Labour cost percentage, by service not by month

Labour cost percentage is labour cost divided by sales, and most venues look at it once a month. The month's labour cost goes in, the month's sales go in, and a single percentage comes out. If it is close to the target, everyone relaxes. The trouble is that a monthly figure is an average of dozens of very different services, and averages hide things. A busy Saturday dinner and a quiet Tuesday lunch have very different labour cost percentages, and the monthly figure blends them into one number that describes neither.

Hidden inside a healthy monthly average there may be services that lose money every week: a lunch with more staff than guests need, an early evening shift that starts an hour before anyone arrives, a Sunday close with a full team for the last two tables. Each one costs a little, every week, and none of them shows up in the monthly figure. Equally, the average may hide busy services that are badly understaffed, with service suffering and staff stretched.

This guide explains how to read labour cost percentage by service rather than by month. It covers how to split labour and sales by service, how to read the results alongside sales per labour hour, which services usually stand out, what to do about them through staff scheduling, how service-level labour connects to prime cost, and how to keep the analysis simple enough to do every week.

A monthly labour cost percentage is useful for the accounts and for comparing one month with another. It is much less useful for deciding what to change, because it does not say where the cost sits. A month at 28 per cent might be every service at about 28 per cent, or weekend dinners at 20 per cent and weekday lunches at 45. The actions needed are completely different.

The monthly figure also arrives too late. By the time it is calculated, the month's rotas have been worked and paid. A problem found at the end of the month has already cost four weeks.

Reading the figure by service, every week, shows where the cost is and gives time to act. Larger operators use workforce management systems for this; a single venue can do the same thing with a simple weekly sheet.

A service is a meal period at the venue, such as weekday lunch, weekday dinner, weekend brunch, Saturday dinner or Sunday lunch. For labour, it is the block of time during which a particular team works to serve a particular trade.

Choose services that match how you staff the venue. If lunch and dinner are staffed separately, split them. If brunch on Saturday is different from brunch on Sunday, split those too. Keep the list short enough to manage, usually somewhere between eight and fifteen services a week.

To split labour by service, assign each hour worked to the service it supported. For hourly staff, that is usually straightforward from the rota or timesheets: a server working from 11:00 to 15:00 is on lunch; one working from 17:00 to 23:00 is on dinner.

Some hours are shared, such as a morning prep cook whose work supports both lunch and dinner, or a manager on a salary. Allocate them sensibly, for example in proportion to each service's sales, or simply split them evenly, and keep the method the same every week.

Include the full labour cost, not just wages. Add on-costs, leave accruals and benefits as a percentage of the hours worked. Our guide to labour cost in a restaurant, counted properly covers what belongs in the figure.

Sales by service come from the till or point of sale: the sales rung during each service's hours, excluding tax. Most systems can show sales by hour, which can be grouped into services.

Use the same service boundaries for sales and labour, so each service's labour is compared with the sales it supported. If a service's staff start an hour before guests arrive, those hours still belong to that service, because they are part of what it costs to run.

With labour and sales split, divide each service's labour by its sales. The result is the labour cost percentage for that service. Lay them out side by side for the week, and the pattern appears: which services are efficient, which are expensive and which are stretched.

Expect variation. A quiet service with a minimum team will always have a higher percentage than a busy one, because the fixed part of labour is spread over fewer sales. The question is not whether every service has the same percentage, but whether each service's percentage is reasonable for what it is, and whether any service is costing more than it brings.

Compare each service with itself over time, too. A Tuesday lunch whose percentage is rising week after week is telling you something.

Sales per labour hour is sales divided by hours worked, for each service. It shows how much each hour of staff time brings in, and it complements the percentage. A service with high sales per labour hour is busy relative to its staffing; one with low sales per labour hour has more staff than its trade needs, or more than it can use.

Covers per labour hour does the same with guests instead of sales. It is useful because it does not change when prices change, so services can be compared across seasons and price rises more fairly.

Read the percentage and the per-hour figures together. A service with a high labour percentage and low sales per labour hour is overstaffed; one with a low percentage and very high sales per labour hour may be understaffed.

In many venues, the same kinds of service stand out. Weekday lunches are often expensive, because trade is thinner but the venue still needs a kitchen and floor team. Early evenings can be expensive if the full dinner team starts long before the first guests. The last hour of service, especially on quiet nights, can carry a full team for a handful of tables. Sunday evenings and Monday services are often quiet.

At the other end, busy weekend dinners often have the lowest percentages, sometimes low enough to suggest the team is stretched, with service and staff suffering. The aim is not to push every service to the lowest possible percentage, but to bring the expensive ones down and make sure the busy ones are properly staffed.

When a service's labour cost percentage is consistently high, there are a few options. The first is staff scheduling: adjust the rota so start and finish times match the trade, with fewer staff at the quiet edges of the service and more at the peak. Our guide to staff scheduling against the trading curve covers how.

The second is to grow the service's sales: a set lunch, an early evening menu, an event or a promotion that fills the quiet hours with the team already in place. Our guide to promotions that fill the right seats covers how to choose one.

The third, for services that remain unprofitable however they are run, is to change the hours: open later, close earlier, or close for that service entirely. That is a bigger decision, and it should be made with the whole picture, including guests' expectations and the venue's reputation.

When a service's percentage is very low and sales per labour hour very high, the team may be stretched. Look at service measures for that service: ticket times, guest feedback, complaints, remakes. If they are slipping, adding hours at the peak may pay for itself in better service, more return visits and fewer mistakes.

Stretched services also wear out staff. A team that is always overstretched on Saturdays is more likely to leave, which brings recruitment and training costs. A slightly higher labour percentage on the busiest night may be the right trade-off.

Every service needs a minimum team to open safely and serve properly: someone in the kitchen, someone on the floor, someone responsible. That minimum is a fixed cost for the service, whatever the trade. On a very quiet service, it alone may make the labour percentage high.

Work out the minimum team for each service, and the sales needed to cover it at a reasonable percentage. If a service rarely reaches that level, it is a candidate for growing sales or changing hours.

Labour is half of prime cost, the combination of food and drink cost and labour cost. Reading prime cost by service, as well as labour, shows the full picture of each service's direct costs. A lunch with a modest labour percentage but a high food cost, because of a generous set menu, may be less profitable than it looks.

Our guide to prime cost, the one number to watch weekly covers the combined measure. Even a rough split of food cost by service, based on the menu mix, adds useful context.

Much of the waste in a service's staffing sits at its edges: the first and last hour. Teams often start well before the first guests arrive and stay on after the last tables have been served, because shifts are written in round blocks rather than around the trade. A dinner team that starts at five for a service that fills from seven, or a lunch team that stays until four when the last guests leave at three, is paid for hours the service does not use.

Look at when guests actually arrive and leave in each service, and when the kitchen and floor genuinely need each person. Stagger start and finish times, so the team builds up as the room fills and winds down as it empties. Some tasks, such as preparation and setting up, need people before guests arrive; plan exactly who and for how long.

Holidays and events change both sides of the calculation. A public holiday may bring a busy lunch that is usually quiet, or empty a business district completely. A private event changes a service's sales and staffing in one go. Holiday pay rates, where they apply, raise the cost of the hours worked.

Mark these weeks in the sheet and read them separately. Compare a holiday service with the same holiday last year rather than with an ordinary week, and use what you learn to staff the next one better.

Once you have several weeks of service-level figures, you can set a sensible target range for each service: what a well-run Tuesday lunch, Friday dinner or Sunday lunch should cost in labour as a share of its sales. Targets set from your own figures are far more useful than a single target for the whole venue.

Review the targets each season, because trade changes with the time of year. A summer lunch target may differ from a winter one.

The point of reading labour by service is not only to save money but to spend it where it does the most good. Hours taken from an overstaffed quiet lunch can be moved to the peak of a stretched Saturday, where they improve service, reduce stress and protect the guest experience. The total may barely change while the venue runs better.

Think of the rota as a budget of hours to place well, rather than a cost to minimise.

The analysis only helps if it is done regularly. A simple weekly sheet works: one row per service, with hours worked, labour cost, sales, labour cost percentage and sales per labour hour. Fill it in each week from timesheets and sales, and look at it before writing the next rota.

Keep it simple. The aim is to see which services stand out and whether last week's changes worked, not to produce a perfect allocation of every minute.

Compare like with like. A Tuesday lunch in a holiday week is not comparable with a normal Tuesday lunch; a Saturday with a private event is not comparable with an ordinary Saturday. Note unusual weeks and leave them out of comparisons, or compare them with the same occasion last year.

Look at trends over several weeks rather than single weeks. One expensive Tuesday lunch may be bad luck; four in a row is a pattern.

In a group, reading labour by service matters even more, because each venue has its own pattern. One venue's weekday lunches may be busy because of nearby offices; another's may be empty. A group-wide monthly figure hides those differences completely.

Ask each venue to keep the same simple weekly sheet, with the same service definitions, so the group can compare like with like. Look for services that are expensive everywhere, which may point to a shared policy such as standard shift lengths, and for services that are expensive at only one venue, which point to something local.

Share the service-level picture with the managers and senior staff who write rotas and run services. They often know why a service is expensive: a delivery that arrives late, a regular group that always comes at the end of the night, a prep task that runs into service. Their knowledge makes the fix better.

Frame it as making the rota fit the trade, not as cutting hours. Staff respond better to "we are starting the dinner team when guests actually arrive" than to "we need to reduce labour".

Here is an illustrative week for a venue, by service. The figures are examples.

servearound does not track hours or labour cost, but it provides the sales side of the split. Payment analytics shows revenue by hour and by day, and order analytics shows when checks are rung across the week, so sales can be grouped into the same services as your rota. Reservation analytics shows covers by meal period and by day and hour, so covers per labour hour can be calculated for each service.

Kitchen analytics shows ticket times by day and hour, which helps you see whether a service with a very low labour percentage is stretched. See Analytics.

Each service's labour cost divided by that service's sales, so you can see where labour is well spent and where it is not.

No. Quiet services will be higher because of the minimum team. Look for services that are unreasonably high or very low.

Allocate them sensibly, such as by each service's share of sales, and keep the method consistent.

Sales divided by hours worked for a service, showing how much each hour of staff time brings in.

Adjust the rota, try to grow its sales, or, as a last step, consider changing the hours.

Weekly, before writing the next rota.

Not always. A quiet service may simply need a minimum team, and a service with a high percentage may still be worth running for the guests and reputation it brings. Read it alongside sales per labour hour and the service's purpose.

Yes. A rota with actual start and finish times written in after each shift is enough to start. Accuracy improves as the habit settles, and the team quickly gets used to noting the real times.

Labour cost percentage by month hides the services that cost money and the ones that are stretched. Split labour and sales by service, using full labour cost and actual hours, and read each service's percentage alongside sales per labour hour and covers per labour hour. Expect quiet services to be higher, and look for those that are consistently out of line.

Use staff scheduling to fit the rota to the trade, grow sales on quiet services, protect service quality on busy ones, and consider changing hours only when nothing else works. Read labour cost together with food cost as prime cost, do the analysis weekly before the next rota, and share it with the managers who run each service.