Average spend per head, split properly

Average spend per head is one of the most quoted numbers in hospitality and one of the least useful in the form most venues read it. It is simple: sales divided by the number of guests. It tells you what the typical guest spent. But a single figure across the whole week blends a quick weekday lunch with a long Saturday dinner, a table of two with a party of twelve, and a first-time visitor with a regular who has been coming for years. The blend hides almost everything worth knowing.

Split properly, the same number becomes one of the most practical tools a manager has. Spend per head at weekday lunch shows whether the set lunch is working. Drinks spend at dinner shows whether the bar is selling. Spend by party size shows whether large tables are as valuable as they look. Spend by new and returning guests shows whether regulars are worth more over time than any single visit suggests.

This guide explains what the figure measures, why one average misleads, and the four splits that show what to change: by service, by food and drink, by party size, and by new and returning guests. It covers revenue per seat alongside spend, what moves spend honestly, and how to set targets and owners for each split.

Spend per head is net sales for a period divided by the number of covers served in that period. If a service takes 4,200 in sales from 120 guests, spend per head is 35. Use net sales, after discounts and before VAT or sales tax, so the figure reflects what the venue actually earned.

Count covers consistently. A cover is a guest who ate or drank, including children if they order, and excluding staff meals and tastings. Decide how to count a guest who only had a drink at the bar, and keep to it. A change in how covers are counted changes the figure without anything real having changed.

The two terms usually mean the same thing: sales divided by guests. Some venues use spend per cover for the dining room and spend per head for the whole venue, including bar guests who did not take a table. Choose one definition, write it down and use it everywhere, so a manager reading the figure in one report can compare it with the same figure in another.

The important point is not the name but the consistency. A figure that sometimes includes bar trade and sometimes does not will move for no reason that anyone can act on.

A single average across the week is a blend of very different kinds of visit. A weekday lunch might average well under half the spend of a Saturday dinner. If the share of lunch covers rises one week, the blended average falls, even if every individual service performed exactly as usual. The manager sees a fall and starts looking for a problem that does not exist.

The reverse happens too. A week with more dinner covers lifts the blended figure, and a real fall in drinks spend at dinner can be hidden inside it. In both cases, the blended number is moved by the mix of services more than by anything the team did.

Here is an illustrative pair of weeks. The figures are examples.

The blended figure fell, so the manager worries. But lunch spend held steady, dinner spend rose, and covers rose overall. Week two was a better week. Only the split shows it.

The first and most important split is by service. At a minimum, separate lunch from dinner. Where the venue opens for breakfast, brunch or late-night trade, give each its own figure. Each service has its own menu, its own guests and its own reasons for visiting, so each has its own normal level of spend.

Read each service against its own history: lunch this week against lunch in recent weeks and the same week last year. A change inside a service is real information. A change in the blended figure caused by a change in the mix of services is not.

Lunch spend is usually shaped by time. Guests have an hour, often less, and choose one or two courses, perhaps a soft drink. A set lunch or a short menu can raise covers while keeping spend steady; a slow kitchen can lower spend by leaving no time for a coffee or dessert.

Dinner spend is shaped by occasion. Guests have more time, order more courses, and drink more. Dinner is where drinks, desserts and sides make the biggest difference, and where the team's selling has the most effect. Treat the two as different businesses that share a kitchen.

Within each service, weekdays and weekends often differ too. A Tuesday dinner may be couples on a quiet evening out; a Saturday dinner may be groups celebrating. A weekend lunch may be long and leisurely, with families and wine; a weekday lunch is quick business.

Where the difference is large, split again: weekday lunch, weekend lunch, weekday dinner, weekend dinner. Four figures are still easy to read, and each one belongs to a service a manager can picture.

The second split separates food spend per head from drink spend per head. The two are sold by different people, priced with different margins and affected by different decisions. A fall in total spend could be a fall in drinks with food holding steady, or the other way round, and the response is completely different.

Drinks usually carry higher margins than food, so a small change in drinks spend can have a larger effect on profit than the same change in food. That makes drinks spend per head at dinner one of the most valuable figures a bar manager can own.

Within food and drink, sales by category shows where the spend goes: starters, mains, sides, desserts; wine, beer, spirits, soft drinks, coffee. Dividing each category's sales by covers shows spend per head on that category, which is often the most actionable figure of all.

If dessert spend per head fell, the pastry section and the floor team know where to look. If wine spend per head rose after a list change, the change worked. Our guide to menu mix covers how to read what guests actually order.

The third split looks at party size. Couples, small groups and large parties spend differently. Large parties often spend more per head on drinks and sharing dishes, but sometimes less on wine per head when one bottle is shared across many glasses. Couples on an occasion may spend more per head than anyone.

Split spend by party size at least into three bands: two guests, three to six, and seven or more. The result often surprises: the large tables that the venue works hardest to fit in may be no more valuable per head than couples.

Party size matters because it shapes the room. A large party takes several tables, often for longer, and may block a second sitting. If its spend per head is no higher than that of couples, the venue earns less from the same space and time. If it is much higher, it may be worth a set menu, a dedicated area or a minimum spend agreed in advance.

Read party size spend together with seating time. A party of ten that spends a little more per head but sits for three hours may earn less per seat than three couples who each sit for ninety minutes.

The fourth split separates first-time guests from returning ones. New guests are often cautious, ordering mains and a drink; regulars know what they like, order with confidence, and are more likely to try the special or the wine the server suggests. Over time, regulars often spend more per visit, though not always.

This split needs guest records connected to checks, so the venue knows which guests are new and which have been before. Where that is possible, it is one of the most revealing splits, because it shows the value of the work done to bring guests back.

Spend on one visit is only part of a guest's value. Guest lifetime spend, the total a guest spends across all their visits, is what really matters to the business. A guest who spends modestly but comes every fortnight is worth far more over a year than one who spends a lot once and never returns.

The general idea of customer lifetime value applies here: the value of keeping a guest is the sum of their future visits, not the size of a single bill. Our guide to why repeat guests are cheaper than new ones covers the arithmetic.

Spend per head measures each guest. Revenue per seat measures the room: sales divided by the number of seats, for a service or per hour of opening. A venue can raise spend per head and lower revenue per seat at the same time, if guests spend more but sit for much longer, or if fewer seats are filled.

Read the two together. Raising spend while keeping seating times steady raises revenue per seat. Raising spend by slowing service may not. Our guide to table turnover rate covers the seating side.

An average can be pulled by a few very large bills. One table that orders a rare bottle of wine lifts the week's average without telling you anything about the typical guest. Look at the median check, the middle value, and at the spread of check values, alongside the average.

If the average rose but the median did not, a few large bills explain the change. If both rose, the typical guest is spending more. The spread shows whether the venue has a long tail of very small checks, which may point to guests who only had a coffee or a single drink.

Spend per head moves for a handful of reasons: the menu and its prices, the mix of dishes guests choose, what the team suggests and how, the drinks list, the time guests have, and the kind of guests who come. Some of these are within the venue's control; some, like the weather or a local event, are not.

When a split moves, ask which of these changed. A new menu, a price change, a new server, a promotion, a busier or quieter week: each leaves a different mark on the splits. Our guide to increasing average spend without raising prices covers the levers in detail.

The menu shapes spend before anyone speaks. Where dishes sit on the page, how they are described, which ones have a picture or a box around them, and how prices are shown all affect what guests choose. A side dish listed beside each main is more likely to be ordered than one hidden at the bottom of the page.

Our guide to menu design for reading, not for framing covers how layout affects what guests order.

Upselling is suggesting something that improves the guest's visit and raises the bill: a glass of wine that suits the main, a side that completes the dish, a dessert to share. Done well, guests welcome it, because it helps them choose. Done badly, it feels like pressure and damages the visit.

The difference is honesty and timing. Suggest one thing, at the right moment, because it genuinely suits what the guest ordered. Train servers on the menu so their suggestions are real. Never push a guest who has said no.

Drinks are usually the biggest lever on spend per head at dinner. An aperitif suggested as guests sit down, wine by the glass that matches the main course, a digestif or a coffee at the end: each adds to the bill with little extra work in the kitchen.

Watch drinks spend per head by service. A fall may point to a slow bar, a list that guests find confusing, or servers who are too stretched to make suggestions. A rise after training or a list change is evidence that the change worked.

Desserts and coffee are often the first casualties of a slow service: guests who waited too long for their mains do not stay for dessert. That makes dessert spend per head a useful signal of kitchen pace as well as of the dessert menu itself.

A dessert menu brought to the table without being asked, a dessert to share suggested to a couple, a coffee offered as the plates are cleared: small habits like these lift the figure steadily.

Specials can lift spend, because they are often priced a little higher and described by the server in person. They can also lower it, if they are priced below the menu or replace a more expensive dish. Track the spend effect of a special as well as how many sold.

A special that sells well but lowers spend per head may still be worth running, for variety or to use stock, but it should be a decision rather than an accident.

A price change raises spend per head directly, unless guests respond by ordering fewer courses or cheaper dishes. After a price change, watch the splits closely for a few weeks: did spend rise by roughly the size of the price change, or less? If less, guests are trading down, and the mix has shifted.

Read covers alongside, too. A price change that raises spend but loses covers may leave sales unchanged and the room emptier.

Promotions usually lower spend per head on the covers they bring in, because they discount or bundle. That is fine if the promotion fills seats that would otherwise be empty. It is a problem if it gives a discount to guests who would have come anyway at full price.

Measure promotions by the spend of the covers they bring and whether those covers were extra. Our guide to offer margin covers the sum to do before launching one.

Discounts and comps lower spend per head directly. A night with many comps, after a kitchen problem, will show lower spend for reasons that have nothing to do with selling. Note them when reading the splits, so nobody mistakes a service recovery night for a selling problem.

Watch the overall level of discounts over time. A slow rise in discounts, from staff deals, regulars' favours or habit, can lower spend steadily without anyone deciding it should.

Set targets for each split rather than for the blend: weekday lunch, weekend lunch, weekday dinner, weekend dinner, and drinks at dinner, for example. Base them on the venue's own history and on what the menu and prices make possible, not on figures from other venues.

Express each target as a range. Spend per head moves a little every week for reasons that do not matter; the range shows when a movement is worth a question.

Give each split an owner. The front of house manager might own spend at weekend dinner; the bar manager drinks spend per head; the head chef dessert spend; the lunch supervisor weekday lunch. Each owner explains movements at the weekly review and proposes what to try.

Ownership turns a number people watch into a number people move. A blended average owned by everyone is owned by no one.

Read the splits once a week, against recent weeks and the same week last year. Look for a split that moved outside its range, and ask its owner why. Most weeks, most splits will be steady, and the reading takes two minutes.

Put the splits beneath the headline figure in the venue's reports, so nobody reads the blend alone. Our guide to restaurant KPIs worth tracking every week shows where the figure sits among the other weekly numbers.

In a group, compare spend per head across venues by split, not by blend. A city-centre venue with heavy lunch trade will always have a lower blended figure than a destination venue that only opens for dinner. Compared service by service, the figures are far more useful and far fairer to each manager.

Share what works. A drinks list change that lifted drinks spend at one venue may work at another.

Here is an illustrative month for a 60-seat venue. The figures are examples.

Actions: the bar manager introduced wine by the carafe for large tables, the floor team suggested an aperitif at seating, and the venue sent a thank-you email after first visits to bring more guests back. The blended figure rose a little; the splits showed exactly why.

In servearound, the dashboard shows spend per cover against the previous period, and analytics gives you the parts to split it. Payment analytics shows revenue by hour, what made up the total and how deep the discounts ran; reservation analytics shows covers by meal period and party size; and order analytics shows the average and median check, the spread of check values, lines on a check and what goes together. The menu report, which can be read for one menu at a time, shows categories, where the revenue sits and every dish that sold with its share of revenue.

For the guest split, guest analytics shows spend per guest, first visits against guests who came back, recency and value, and the best guests here, all time, while each guest profile carries visits and spend. Table analytics adds table revenue and revenue per seated hour. See Analytics.

Net sales divided by the number of guests served, for a service, a day or a week.

Because it blends very different services, and moves when the mix of services changes even if nothing else does.

By service. Lunch and dinner are different businesses, and each needs its own figure.

Yes. They are sold by different people, carry different margins and respond to different decisions.

Large parties often spend differently per head from couples, and take more space and time, so read their spend alongside seating time.

Not if it comes with longer seating times and fewer covers. Read it alongside revenue per seat and covers.

Through the menu, the drinks list, good suggestions at the right moment and a kitchen fast enough to leave time for dessert.

It shows the value of a guest across all their visits, which is often more important than spend on any single visit.

Average spend per head is sales divided by guests, but a single figure for the week blends services, party sizes and guests so thoroughly that it hides what changed. Split it four ways: by service, by food and drink with sales by category inside, by party size, and by new and returning guests, with guest lifetime spend in mind.

Read the splits alongside the median, the spread and revenue per seat, and note promotions, discounts and comps. Give each split an owner and a target range, move it through the menu, the drinks list and honest upselling, and compare venues by split rather than by blend.