Are You Into Doing the Math? Restaurant Numbers Every Manager Should Know

Restaurant people tend to fall into two camps: those who love the numbers and those who'd rather be on the floor. You don't have to love them. You do need to be able to do five simple calculations, because they explain most of what happens to your sales, and they turn vague worries ("we're slow") into specific questions ("our Tuesday check average dropped two dollars").

All the examples below use made-up but realistic numbers. Swap in your own.

How do covers and check average add up to sales?

Sales equal covers multiplied by check average, so every sales goal comes down to more guests, more spending per guest, or both.

Covers are the number of guests served. Check average is net sales divided by covers. A restaurant that serves 900 covers in a week with a $32 check average does $28,800 in sales.

That simple formula is powerful because it splits a sales problem in two. If sales are down 10 percent, is it because fewer people came, or because the same people spent less? The fixes are different. Fewer guests is a marketing, reputation or capacity question. Lower spending is a menu, pricing or service question: are servers suggesting appetizers and desserts, is the wine list working, has a popular high-priced item been removed?

Small changes add up. Raising check average by $2 on 900 weekly covers is $1,800 a week, over $90,000 a year, without a single extra guest.

What are table turns, and why do they matter?

Table turns measure how many times each seat is used in a service, and they cap how many covers you can serve.

Turns for a service = covers ÷ seats. A 60-seat dining room that serves 150 guests at dinner turned 2.5 times.

If you're full and turning away guests, more turns are worth real money. But adding turns rarely means rushing people out, which guests notice and resent. It usually means removing slow steps that guests don't value: waiting for a greeting, waiting for the check, waiting for a table to be reset. Time those steps on a busy night. Shaving five minutes off the end of a meal, by dropping the check promptly when guests are ready and resetting tables quickly, can add a partial turn on a full night.

If you're not full, turns aren't the problem. Covers are.

How do you compare busy and slow periods fairly?

Use revenue per available seat hour, which puts every hour of the week on the same scale.

RevPASH = sales ÷ (seats × hours open). If your 60-seat room does $6,000 between 5 and 10 p.m. on a Saturday, that's $6,000 ÷ (60 × 5) = $20 per seat hour. If Tuesday lunch does $900 from 11 to 2, that's $900 ÷ (60 × 3) = $5 per seat hour.

Seeing every daypart in the same unit is useful when you decide what to change. A daypart with low RevPASH might need marketing, a different menu, or shorter hours. A daypart with very high RevPASH and a waitlist might need faster turns or a reservation policy.

What is one guest really worth?

Far more than one check, because a happy guest comes back and brings others.

Here's the basic math of a regular. Suppose a regular spends $40 a visit, comes twice a month, and keeps coming for three years. That's $40 × 24 visits × 3 years = $2,880 in sales from one person.

Now add word of mouth. Say that over those three years the regular brings or sends two friends who become occasional guests, visiting once a month each for two years at $40. That's another $960 each, $1,920 for both. One happy guest is now worth about $4,800 in sales. At a 70 percent gross margin after food cost, that's roughly $3,360 in gross profit.

Your numbers will differ, and your POS or loyalty data can give you real visit frequencies. The point is the size of the multiplier. When a server is deciding whether to comp a $9 dessert for a guest whose entree came out cold, the real comparison is $9 against the guest's future value.

How does a bad night change the math?

A lost regular costs you their future visits plus whatever people hear from them.

The same multiplier works in reverse. A guest who has a bad experience and doesn't return takes their future spending with them, and unhappy guests often tell others, in person or in a review that hundreds of people will read. That's why recovery matters so much: fixing the problem at the table, sincerely and quickly, can keep the guest and sometimes turns them into a bigger fan than before.

How should a manager use these numbers each week?

Look at the same few numbers every week against the same week last year, and ask one question about each.

A simple weekly review covers it:

• Covers by daypart: up or down, and why?
• Check average: what changed in the mix?
• Turns on full nights: where did tables wait?
• RevPASH by daypart: which hours earn the least?
• Regulars: are the familiar names still coming in?

Write the answers next to your prime cost for the week. After a month you'll have a clearer picture of the business than most monthly P&Ls give you, and you'll know which lever to pull first.