How to Know If Your Restaurant Is Doing Well: 12 Metrics That Actually Matter

So, is your restaurant doing well?
You cannot answer that reliably with one number.
A restaurant doing well usually shows strength across four areas:
- Sales and demand: Enough customers are coming in, and revenue is holding or growing.
- Margin and cost control: Enough of that revenue remains after food, labor, and other expenses.
- Operational efficiency: Staffing, menu mix, discounts, and sales channels are working economically.
- Repeatability and trend direction: Performance is healthy across comparable periods, rather than being carried by a few unusually strong days.
That distinction matters because revenue alone can hide a lot.
Sales can rise while profit falls. Average checks can increase while guest traffic declines. Delivery can add substantial revenue while producing weaker margins. A slow Tuesday can look alarming until you discover that Tuesdays are normally slow, a storm passed through town, or last year’s comparison included a local event.
The better question is:
How is the restaurant performing relative to its normal baseline, its costs, its capacity, and the conditions around it?
These 12 restaurant performance metrics give you a practical way to answer it.
A simple restaurant health scorecard
| Metric | What it tells you | Watch for |
|---|---|---|
| Net sales | How much revenue the restaurant actually generated | Growth without corresponding traffic or margin growth |
| Same-store sales growth | Whether comparable sales are improving over time | Comparisons distorted by seasonality or calendar differences |
| Transactions / covers | Whether customer traffic is growing or shrinking | Sales rising while guest count falls |
| Average check | How much each transaction or guest is worth | Price increases masking weaker traffic |
| Sales by daypart | Where demand is strengthening or weakening | One weak period hidden by a strong daily total |
| Sales / menu mix | What products and channels are driving revenue | Growth shifting toward lower-margin items or channels |
| Food cost % | How much food cost consumes each dollar of food sales | Supplier inflation, waste, portions, theft, or poor pricing |
| Labor cost % | How much revenue is being spent on labor | Staffing costs rising faster than sales |
| Prime cost | Combined pressure from COGS and labor | Too little revenue remaining for overhead and profit |
| Contribution margin | How much individual items contribute after direct food cost | Popular items generating little actual contribution |
| Discounts, comps, refunds, voids | Revenue leakage and possible operational issues | Sudden increases or growing dependence on discounts |
| Repeat business | Whether guests are choosing to return | Acquisition masking deteriorating customer retention |
These metrics are most useful when interpreted together. None has a universal number that separates a “good” restaurant from a “bad” one.
The National Restaurant Association’s restaurant cost analysis provides a useful illustration. Food and labor each account for roughly one-third of sales at a typical restaurant, with other operating expenses consuming much of what remains.
More detailed figures in the Association’s 2025 Restaurant Operations Data Abstract reinforce the same point: restaurant cost structures and margins vary meaningfully by service model.
Benchmarks can tell you where to look. They should not tell every restaurant what its numbers must be.
1. Gross sales and net sales
Sales are the obvious starting point, but the number you use matters.
Gross sales generally represent the total value of products sold before discounts and certain deductions.
Net sales represent what remains after relevant discounts, comps, returns, and similar adjustments, depending on how your POS defines them.
For evaluating operating performance, net sales are usually the more useful starting point because they are closer to the revenue the restaurant actually earned.
But even net sales need context.
Suppose this Friday generated $12,400 compared with $10,900 on Thursday. That tells you almost nothing about whether the restaurant improved. Fridays may normally outperform Thursdays substantially.
Instead, compare like with like:
- this Friday versus typical Fridays
- this week’s lunches versus recent weekly lunches
- August versus last August
- the most recent four weeks versus the preceding four weeks
- dine-in sales versus previous dine-in periods
The principle is simple:
Compare a period against the period it reasonably could have been expected to resemble.
A Monday lunch and Saturday dinner do not share the same demand pattern. Neither do December and February for many restaurants.
That is why useful restaurant reporting should move beyond “yesterday versus the day before” toward meaningful weekday and seasonal baselines.
2. Same-store sales growth
Same-store sales growth asks whether an existing restaurant is generating more or less revenue than it did during a comparable previous period.
A simple calculation is:
Same-store sales growth % = (Current comparable sales - Previous comparable sales) ÷ Previous comparable sales × 100
If comparable weekly sales increased from $40,000 to $42,000:
($42,000 - $40,000) ÷ $40,000 × 100 = 5% growth
The important word is comparable.
Depending on the question, useful comparisons might include:
- this Wednesday versus the average of the previous eight Wednesdays
- this month versus the same month last year
- the latest four weeks versus the previous four weeks
- year-to-date sales versus the same year-to-date period last year
Same-store sales become especially useful when combined with traffic.
The National Restaurant Association’s same-store sales and customer traffic data shows why these should be separated. Restaurant sales can improve even while customer traffic remains weaker overall.
That can happen when menu prices or average checks increase while fewer people are visiting.
So when same-store sales change, your next question should usually be:
What produced the change?
3. Number of transactions, covers, or guests
Revenue is essentially the result of two forces:
How many customers bought from you × how much they spent
Depending on your restaurant and reporting system, traffic might be measured as:
- transactions
- orders
- covers
- guests
- checks
Suppose monthly sales rise 6%.
That sounds good.
But imagine transactions fell 5% while average checks rose 12%. The restaurant is technically producing more revenue, but fewer customers are buying.
That is a very different signal from 6% sales growth caused by more traffic.
The first scenario might indicate:
- menu price increases
- larger orders
- a shift toward higher-priced products
- declining customer frequency
- weaker value perception
The second could indicate genuine growth in demand.
Neither interpretation is certain without more evidence. The point is that separating traffic from spend tells you where to investigate.
4. Average check or average order value
Average check measures the revenue generated by the average transaction.
A common formula is:
Average check = Net sales ÷ Number of checks or transactions
If dinner sales were $8,000 across 320 checks:
$8,000 ÷ 320 = $25 average check
Average check can change for many reasons:
- menu prices changed
- employees are upselling more effectively
- customers are buying more add-ons
- alcohol sales increased
- larger parties visited
- product mix changed
- promotions changed
- delivery became a larger share of orders
This is why an increasing average check is not automatically good news.
Consider:
| Period | Transactions | Average check | Sales |
|---|---|---|---|
| Period A | 1,000 | $24 | $24,000 |
| Period B | 850 | $29 | $24,650 |
Sales increased about 2.7%.
But transactions dropped 15%.
You would want to understand whether the restaurant intentionally traded volume for higher-value customers or whether higher prices are masking a meaningful traffic decline.
5. Sales by daypart
A restaurant is rarely one business operating uniformly throughout the day.
Breakfast, lunch, afternoon, dinner, and late-night periods can behave almost like separate businesses.
Suppose daily revenue has stayed around $10,000.
That might appear stable.
But underneath the total:
- lunch fell from $3,500 to $2,700
- dinner grew from $5,000 to $5,800
- everything else remained unchanged
The restaurant has not simply “stayed flat.” Demand has moved.
That can affect:
- staffing
- prep
- inventory
- promotions
- opening hours
- menu design
- marketing priorities
It also changes the questions you should ask.
If lunch is weakening, is it happening every weekday or only Mondays? Are transactions down, or is average spend lower? Has a nearby employer changed its office schedule? Did a competitor introduce a lunch special? Is a local event changing normal traffic patterns?
A single daily sales number hides those questions.
Even a simple comparison such as this Tuesday lunch versus the typical Tuesday lunch over the previous eight weeks can be more useful than another chart of total monthly revenue.
6. Sales mix and menu mix
Your next question is not merely how much you sold, but what produced those sales.
Sales mix can be examined by:
- menu item
- menu category
- dine-in
- pickup
- first-party online ordering
- third-party delivery
- catering
- alcohol
- location
Imagine total revenue grows 10%.
If most of that growth comes from a lower-margin delivery channel while profitable dine-in traffic declines, the economics of the restaurant may have become less attractive even though topline sales improved.
Menu mix matters for the same reason.
A restaurant selling more of its highest-revenue items is not necessarily selling more of its highest-contribution items.
This is the basis of menu engineering: examining both popularity and profitability rather than simply asking what sells most. Restaurant365’s menu-engineering guide is a useful deeper walkthrough if you want to analyze individual items.
For a weekly restaurant health review, though, the simpler question is:
What categories, items, and channels are responsible for the change in sales?
7. Food cost percentage
Food cost percentage shows how much of your food revenue is being consumed by the ingredients used to produce it.
A common formula is:
Food cost % = Cost of food used ÷ Food sales × 100
If you used $9,000 worth of food to generate $30,000 in food sales:
$9,000 ÷ $30,000 × 100 = 30%
To calculate actual food usage, operators commonly derive cost of goods sold from:
Beginning inventory + Purchases - Ending inventory
Food-cost benchmarks are useful, but this is an area where generic targets can easily become misleading.
Restaurant-industry guidance often places food cost somewhere around the high-20s to mid-30s as a broad reference range for many concepts. Toast’s restaurant operations guide, for example, discusses food-cost ranges in that general territory.
That is not a universal target.
A steakhouse, pizzeria, coffee shop, cocktail-heavy restaurant, bakery, and food truck can have very different cost structures.
More useful than asking whether your food cost is “supposed to be 30%” is asking:
- What has our normal food cost been?
- Is it rising?
- Which categories changed?
- Is the increase expected?
A rising food cost percentage can result from:
- supplier price increases
- waste
- overportioning
- spoilage
- theft
- recipe changes
- inadequate menu pricing
- shifting sales toward higher-cost items
External cost conditions matter too.
The National Restaurant Association’s food-cost tracker is useful for seeing how wholesale restaurant food prices are moving over time.
That is another reason not to treat an old restaurant benchmark as a permanent rule.
Food cost percentage tells you that something changed. A deeper food cost percentage analysis should tell you why.
8. Labor cost percentage
Labor cost percentage shows how much of restaurant revenue is being consumed by labor.
A common formula is:
Labor cost % = Total labor cost ÷ Sales × 100
Labor should generally include more than hourly wages. Depending on your reporting, total labor may include:
- salaries
- hourly wages
- payroll taxes
- benefits
- bonuses
- other direct labor expenses
Restaurant labor benchmarks vary substantially by concept.
National Restaurant Association analysis of 2024 operating data illustrates the difference. Among limited-service respondents, salaries and wages including benefits represented a median 31.7% of sales overall. Restaurants reporting a pre-tax profit had a median labor ratio of 30.0%, compared with 34.1% among those reporting a loss.
That does not mean 30% is automatically healthy and 34% is automatically unhealthy.
Service model, geography, local wages, menu complexity, opening hours, and staffing requirements all matter.
If you want broader context on wage conditions, the U.S. Bureau of Labor Statistics maintains employment and earnings data for food services and drinking places.
The trend inside your restaurant is often more useful than the benchmark.
Imagine sales increase from $40,000 to $44,000.
Labor rises from $12,000 to $15,000.
Labor percentage changes from:
30.0% → 34.1%
Sales increased 10%, but labor increased 25%.
That does not automatically mean the staffing decision was wrong. Perhaps you deliberately added service capacity. But it tells you that revenue growth is not translating proportionally into labor efficiency.
And labor should often be examined by daypart.
A 35% labor percentage during a slow Monday afternoon might require a different response from the same percentage during a packed Friday dinner.
9. Prime cost
If you only add one restaurant profitability metric to your weekly review beyond sales, prime cost is a strong candidate.
Prime cost = Cost of goods sold + Total labor cost
And:
Prime cost % = Prime cost ÷ Sales × 100
Restaurant365 has a useful guide to calculating restaurant prime cost if you want a more detailed walkthrough of what belongs in the calculation.
Suppose weekly sales are $30,000:
- COGS: $9,000
- labor: $9,300
Prime cost is:
$9,000 + $9,300 = $18,300
Prime cost percentage:
$18,300 ÷ $30,000 × 100 = 61%
Broad industry guidance often places prime cost somewhere around 60% to 65%, depending on concept.
The National Restaurant Association’s 2025 Restaurant Operations Data Abstract provides useful real-world comparative data, while Restaurant365’s prime-cost benchmarking guidance explains how targets can differ between full-service, quick-service, and fast-casual concepts.
Again, this is a benchmark, not a law.
A restaurant with a 63% prime cost and unusually low rent may have healthier economics than one at 58% with extremely high occupancy expenses.
Prime cost is powerful because it shows whether the two largest controllable pieces of the operation are leaving enough room for:
- rent
- utilities
- insurance
- payment processing
- repairs
- marketing
- administrative expenses
- debt service
- profit
Track the direction, not just the number.
10. Contribution margin by item or category
A high-selling item is not necessarily a high-value item.
For a simplified menu-item calculation:
Contribution margin = Selling price - Direct item cost
Consider two dishes:
| Item | Price | Food cost | Contribution |
|---|---|---|---|
| Dish A | $20 | $9 | $11 |
| Dish B | $14 | $2.50 | $11.50 |
Dish A produces more revenue.
Dish B contributes more dollars after ingredient cost.
Restaurant365’s explanation of menu contribution margin uses the same basic idea: selling price minus the cost associated with the item.
Now combine contribution margin with sales volume and the picture gets more interesting.
This does not mean you should automatically remove lower-margin products. Some may attract customers, complement profitable items, or be strategically important to the menu.
The point is narrower:
Revenue does not tell you how economically valuable a sale is.
That is why menu mix and contribution should eventually be reviewed together.
11. Discount, comp, refund, and void rate
Small forms of revenue leakage can become meaningful when they move together.
Track at least:
- discounts
- promotions
- comps
- refunds
- voids
Then compare them with net sales or transactions.
For example:
Discount rate = Total discounts ÷ Gross sales × 100
The goal is not necessarily to drive the number to zero.
A discount can be intentional marketing.
A comp can be the right service-recovery decision.
A void can be an innocent correction.
What matters is change.
A sudden increase might point toward:
- a new promotion
- more customer complaints
- ordering errors
- employee training problems
- incorrect POS workflows
- discount dependency
- potential abuse
If discounting rises while traffic and repeat business remain flat, for example, you may be paying more to generate essentially the same demand.
That deserves investigation even if total sales have not moved much.
12. Repeat business and customer retention
A healthy restaurant does not have to reacquire every customer from scratch.
Repeat business might be measured through:
- returning-customer percentage
- loyalty activity
- repeat online orders
- visit frequency
- reservation history
- POS customer profiles
Not every independent restaurant can identify every guest.
Cash transactions, walk-ins, shared payment methods, and anonymous orders make perfect customer-level retention measurement unrealistic.
That does not make retention useless.
Use whatever signals your systems can reliably provide.
The more practical question is:
Are people continuing to choose us again?
If acquisition looks healthy but repeat behavior is deteriorating, topline sales may take longer to reveal the problem.
For restaurants with limited customer identification, even directional measures such as loyalty activity, repeat online ordering, or reservation history can provide useful clues.
Restaurant metrics become more useful when you read them together
This is where restaurant performance analysis becomes more useful than a dashboard full of isolated KPIs.
Individual metrics tell you what happened.
Combinations begin to tell you where to look.
Sales up + transactions down + average check up
Possible interpretation:
Higher prices, larger tickets, or product mix are compensating for weaker traffic.
Questions to ask:
- Did menu prices recently increase?
- Which dayparts lost transactions?
- Are repeat visits changing?
- Is traffic declining across the market or specifically at your restaurant?
Sales up + food cost rising faster
Possible interpretation:
More revenue is not necessarily producing proportionally more margin.
Investigate:
- ingredient costs
- menu mix
- pricing
- portions
- waste
- vendor changes
Sales down + transactions stable
Possible interpretation:
Customers are still coming, but spending less.
Look at:
- average check
- category mix
- alcohol
- add-ons
- party size
- discounts
- channel mix
Sales flat + transactions up + average check down
Possible interpretation:
Demand may actually be improving even though revenue appears unchanged.
Perhaps promotions attracted more customers, customers shifted toward lower-priced items, or a channel with smaller orders grew.
Whether that is good depends on margin and whether those customers return.
Sales up + labor percentage down
Often a promising operating signal.
Revenue is growing faster than labor expense.
But check service quality and workload before assuming the restaurant should push labor even lower.
Sales down on one Tuesday
Possibly nothing.
Compare it with:
- typical Tuesdays
- weather
- nearby events
- holidays
- school calendars
- road closures
- promotions
- seasonality
- competitor activity
A number becomes useful when you know what it should reasonably be compared with.
Context creates meaning.
Your own baseline may matter more than an industry benchmark
Restaurant benchmarks answer questions like:
Is this number unusual enough that I should investigate it?
Your own baseline answers:
Is this unusual for my restaurant?
That distinction matters.
A neighborhood café should not blindly compare itself with:
- a fine-dining restaurant
- a bar
- a drive-through QSR
- a food truck
- a hotel restaurant
- a national multi-unit chain
Their labor models, rent, menu economics, alcohol mix, throughput, guest expectations, and operating hours are different.
This is also how the National Restaurant Association’s Restaurant Operations Data Abstract is most useful: as comparative operating data, rather than a universal restaurant scorecard.
For your own restaurant, start building operating baselines such as:
- typical Monday sales
- typical Friday dinner sales
- normal Tuesday lunch transaction count
- average check by daypart
- average order value by sales channel
- trailing eight-week food cost percentage
- typical Friday dinner labor percentage
- average weekly discount rate
Then look for deviations.
Suppose Tuesday normally generates between $5,600 and $6,100 and this Tuesday produces $4,200.
That deserves attention.
Suppose Saturday falls from $12,500 last week to $11,900 this week, but your normal Saturday range is $11,500 to $13,000.
That may just be normal variation.
Generic restaurant benchmarks remain useful. They help expose structural problems you might otherwise normalize.
But a restaurant-performance system becomes much more powerful when industry benchmarks are combined with the restaurant’s own history.
Leading indicators and lagging indicators
Most financial restaurant metrics describe something that has already happened.
These are lagging indicators.
Examples include:
- monthly revenue
- profit
- food cost
- labor percentage
- prime cost
They are essential, but by the time a monthly profit-and-loss statement shows a problem, the underlying behavior may have been changing for weeks.
Earlier signals can sometimes provide additional context.
Depending on the restaurant, those could include:
- reservation volume
- advance catering orders
- online ordering activity
- changes in transaction traffic
- menu-item momentum
- local events
- seasonal occasions
- competitor promotions
- weather
- review activity
- local search behavior
These signals are not guarantees of future demand.
A festival nearby does not guarantee a busy Saturday. A competitor promotion does not guarantee lost customers. Reservation volume does not perfectly predict final revenue.
But they can help explain what is changing around the restaurant before those effects fully appear in monthly financial statements.
For broader market context, the National Restaurant Association maintains a useful collection of restaurant economic indicators covering areas such as sales, traffic, food costs, menu prices, employment, and consumer conditions.
A useful operating view therefore includes both:
What already happened?
and
What is happening that may affect what comes next?
Profit ultimately matters, but one monthly number is not enough
Restaurant owners ultimately need the business to generate profit.
There is no KPI that makes poor profitability irrelevant.
But monthly net profit alone is a weak diagnostic tool.
If profit deteriorates, the number does not automatically tell you whether the cause was:
- traffic
- average ticket
- food cost
- labor
- discounts
- channel mix
- occupancy cost
- utilities
- some combination of them
And by month-end, the underlying change may already be several weeks old.
The National Restaurant Association’s 2025 operating data shows how narrow restaurant margins can be and why relatively small shifts in major costs can materially affect the bottom line.
Those industry figures are reference points, not profit targets for your restaurant.
Profit tells you the result.
The operational metrics above help explain the drivers.
From restaurant reporting to restaurant intelligence
Most modern POS systems can tell you how much you sold.
Many can also report transactions, average checks, item mix, discounts, labor, and other restaurant KPIs.
The harder questions are different:
- Is today’s performance actually unusual?
- What changed?
- Is the decline concentrated in one daypart?
- Are sales falling because of traffic or average ticket?
- Is revenue growing without margin improving?
- Is something happening locally that could explain the change?
- Does this deserve attention now?
- What could the owner reasonably do about it?
That is the layer Allwhile is being built around.
Allwhile is an agentic business-intelligence product for independent businesses, initially focused on restaurants.
It is designed around three practical questions:
- How is my business doing?
- What’s happening around me?
- What can I do about it?
The idea is to connect signals from the business, such as POS performance, with relevant outside context such as local events, seasonal occasions, competitor activity, market changes, and potential opportunities.
Instead of asking an owner to manually inspect every chart, the goal is to surface the changes that may actually deserve attention.
Your existing systems record the numbers. Allwhile helps determine which changes are worth noticing.
That is complementary to POS reporting, bookkeeping, accounting software, and professional financial advice, not a replacement for them.
A 10-minute weekly restaurant health check
You do not need a sophisticated analytics operation to start using these ideas.
Once a week, run through this sequence.
1. Compare net sales with a meaningful baseline
Look at the week by comparable weekday rather than only comparing the total with last week.
Ask:
Where was performance meaningfully above or below normal?
2. Separate traffic from average check
If revenue moved, determine whether:
- transactions changed
- average check changed
- both changed
Do not treat those as the same kind of growth or decline.
3. Review food and labor trends
Look beyond one week’s percentage if it is noisy.
Ask whether either cost has been moving consistently away from its recent baseline.
4. Calculate prime cost
Combine COGS and labor.
If prime cost is deteriorating, identify which component is responsible before trying to fix it.
5. Find unusual dayparts or categories
Look at lunch versus dinner, weekdays versus weekends, and major menu categories.
Find where the change actually occurred.
6. Check discounts, comps, refunds, and voids
Look for changes rather than simply checking whether they exist.
7. Look at repeat business where possible
Use loyalty, online-ordering, reservation, or POS customer data.
You do not need perfect customer identification to watch directional trends.
8. Ask what happened around the restaurant
Consider:
- weather
- events
- holidays
- school calendars
- major local activity
- promotions
- competitor activity
Do not automatically attribute performance to these factors. Use them as possible explanations worth testing.
9. Pick one or two things that deserve action
The goal is not to optimize every metric every week.
It is to find the few changes that matter enough to investigate.
Maybe lunch traffic is falling.
Maybe food cost has climbed for four consecutive weeks.
Maybe delivery revenue is booming but contribution is weak.
Maybe nothing important changed at all.
That last outcome is useful too.
A good performance review should help you decide when not to react.
The goal is not more restaurant KPIs
A healthy restaurant is not simply one where revenue keeps increasing.
You need to know:
- Are we attracting enough demand?
- Are guests spending enough?
- Are we keeping enough of the revenue?
- Are food and labor costs moving in the right direction?
- Is growth coming from economically useful products and channels?
- Are customers returning?
- Are any changes significant enough that we need to respond?
That requires more than a collection of restaurant benchmarks.
It requires comparison.
A 32% labor cost means one thing relative to a 27% historical baseline and something else relative to a 35% baseline.
A 10% sales decline means one thing on a random Tuesday and something else when it repeats across six comparable Tuesdays.
A record revenue month means one thing when traffic is growing and margins are stable, and something else when higher prices are hiding declining transactions.
The useful dashboard is not the one with the most charts. It is the one that tells you what changed, why it might matter, and where to look next.
That is the kind of restaurant intelligence Allwhile is being built to provide: watching the business and the market around it, connecting the signals, and surfacing the changes that may deserve your attention.