Why can two restaurants in the same group, operating under a similar concept, deliver very different profit and productivity? Comparing sites by revenue, operating profit or Food Cost alone rarely explains the difference. A useful restaurant performance comparison starts by making locations genuinely comparable, then tracing the gap through sales, food, labour, operating costs and resource utilisation until management can identify what is actually driving the result.
Restaurant benchmarking should not stop at identifying the best- and worst-performing locations. The objective is to move from the result to the factors behind it, distinguish controllable factors from external conditions, and determine which operating practices can realistically be transferred to other restaurants.
The management logic is: result → normalised KPI → factor → cause → controllable factor → action → measurement of the outcome.
What Should Restaurants in a Group Actually Be Compared On?
Before building a restaurant ranking, management needs to define the business result it is trying to understand. Depending on the decision, that may be operating profit, contribution margin, cash generation, labour productivity, food consumption, sales productivity or the efficiency of a specific asset.
The KPI measures the result or condition. It does not necessarily explain it.
For example, if Restaurant A generates more operating profit than Restaurant B, profit is the result. The first step is to break that result into its main economic components:
Operating profit = Revenue − Food and beverage costs − Labour costs − Other operating costs
This is only the first level of analysis. Each component needs to be decomposed further.
- Revenue may differ because of transaction or guest volume, average spend, menu mix, pricing, discounts, sales channels or trading periods.
- Food and beverage costs may differ because of sales mix, ingredient prices, purchasing conditions, actual consumption, waste, spoilage or other inventory variances.
- Labour costs may differ because of paid hours, staffing structure and the average cost of labour.
- Other operating costs may reflect the size and format of the site, opening hours, occupancy, local operating conditions or differences in the way costs are allocated.
The difference in profit between two restaurants is therefore not a cause in itself. It is the combined outcome of several underlying factors.
This distinction is particularly important in multi-unit operations across Europe and the Middle East. A restaurant in a shopping centre, a hotel-based F&B outlet, a high-street site and a delivery-heavy location may all belong to the same group while operating with different demand patterns, trading hours, space requirements and staffing models. Their results can still be compared, but not without adjusting for those differences.
A consistent restaurant management framework should therefore establish what is being measured, why it matters and which factors management can investigate when performance changes.
Normalise Restaurant KPIs Before Building a Ranking
A larger restaurant will often generate more revenue simply because it has more seats, a larger kitchen, longer opening hours or a greater flow of guests. That does not automatically make it more efficient.
The same problem applies to profit. A smaller site may produce less absolute profit but use significantly fewer labour hours, square metres or other resources to generate it.
For this reason, absolute results should be supplemented with normalised KPIs that relate performance to the relevant operating base.
| Area |
Absolute measure |
Possible normalised measure |
| Sales |
Revenue |
Revenue per trading day, trading hour or seat |
| Guest demand |
Guests or transactions |
Guests or transactions per day, hour or seat |
| Labour |
Total labour cost |
Labour cost per guest, transaction, trading hour or unit of revenue |
| Labour productivity |
Revenue or transactions |
Revenue or transactions per paid labour hour |
| Food and beverage |
Cost of products consumed |
Cost per guest, transaction or unit of revenue |
| Space |
Revenue or profit |
Revenue or profit per square metre |
| Seating capacity |
Revenue |
Revenue per available seat |
| Operating result |
Profit |
Profit per trading day, hour, seat or other relevant resource |
There is no single denominator that makes every restaurant comparable. The normalisation method must correspond to the management question.
Revenue per square metre can help assess the commercial use of space, but it says little about labour productivity. Revenue per paid labour hour may be appropriate for analysing labour deployment, but it does not explain whether the location itself has enough demand.
Restaurants must also be comparable in operating model
Before drawing conclusions, locations should be grouped by the characteristics that materially affect their economics. These may include restaurant format, service model, scale, trading period, opening hours, seating capacity, delivery share and available resources.
A large full-service restaurant should not automatically be benchmarked against a compact café or a delivery-led unit as though the businesses used identical operating models.
Accounting treatment must also be consistent. If one restaurant carries certain operating expenses directly while another receives the same services through head-office allocations, their reported profit cannot be compared fairly until the management-accounting rules are aligned.
The same principle applies to groups operating through several legal entities, brands or countries. Before management compares locations, it should ensure that revenue, costs and operational measures follow the same definitions and period boundaries.
Only then does a restaurant ranking become a useful starting point rather than a potentially misleading league table.
Build a Factor Tree for Sales, Food, Labour and Resources
The purpose of factor analysis is to move from the final KPI towards the variables that explain it and, eventually, towards factors that restaurant management can influence.
| Result |
First-level factor |
Second-level factors |
| Operating result |
Sales |
Guest or transaction volume, average spend, sales mix |
| Operating result |
Food and beverage cost |
Sales mix, purchase cost, actual consumption, waste and inventory variance |
| Operating result |
Labour cost |
Paid hours, staffing mix, average labour cost |
| Operating result |
Other operating costs |
Variable costs, fixed or semi-fixed costs, site characteristics |
| Resource efficiency |
Labour |
Output per paid labour hour |
| Resource efficiency |
Space and capacity |
Revenue or profit per square metre, seat or trading hour |
Sales: move beyond the revenue number
A useful starting formula is:
Revenue = Number of transactions × Average transaction value
If the operation tracks covers or guests more consistently than transactions, guest count can be used instead. The important point is that the same definition must be applied across all restaurants being compared.
If one location has lower revenue, management should first determine which component is responsible. A transaction-volume gap may need to be analysed by daypart, weekday, channel, restaurant zone or other relevant dimensions. A gap in average spend may require analysis of menu mix, pricing, discounting and purchase combinations.
It is important to distinguish the factor from the reason the factor changed. Lower transaction volume is a factor behind lower revenue. The underlying cause might be weaker local demand, different opening hours, channel availability, service capacity, seasonality or another condition that needs to be verified separately.
Food Cost: the percentage alone does not explain performance
Two restaurants may report the same Food Cost percentage while arriving there through very different operating processes.
One location may have a less favourable sales mix but tight inventory control. Another may have a stronger theoretical margin but lose part of that advantage through waste, over-portioning, spoilage or unexplained stock variance.
A more useful chain of analysis is:
sales → menu mix → theoretical product cost → actual consumption → variance
This requires comparable sales data, product costs, inventory movements and an agreed costing methodology. The purpose is not simply to identify the restaurant with the lowest Food Cost. Management needs to understand why restaurants require different quantities or values of product to produce comparable sales.
Labour: separate cost from productivity
Total labour cost is highly sensitive to restaurant size and operating hours. For comparison purposes, managers need to analyse both the cost of labour and the output generated by that labour.
A simple productivity measure is:
Labour productivity = Revenue / Paid labour hours
Depending on the operating model, transactions, covers or another output measure may be more useful than revenue.
The labour-cost equation can also be decomposed:
Labour cost = Paid labour hours × Average labour cost per hour
If the difference is driven by hours, managers can then investigate demand patterns, rota design, staffing levels by daypart, operating procedures and productivity. If it is driven by hourly cost, the staffing mix and cost structure need to be examined separately.
This matters in markets where restaurant groups often operate with multicultural teams, varied skill levels and different staffing structures across outlets. A higher Labour Cost ratio should not trigger an automatic instruction to cut hours: reducing labour without understanding guest demand and service requirements may damage sales or service standards.
Assets, seating and trading time are also resources
Restaurant performance is created not only by products and people but also by physical capacity. Useful measures may include:
Revenue per seat = Revenue / Number of available seats
Revenue per square metre = Revenue / Relevant operating area
Revenue per trading hour = Revenue / Restaurant trading hours
These ratios do not prove that a restaurant is well or poorly managed. They show how intensively a specific resource is being used and indicate where deeper analysis may be required.
For example, low revenue per square metre does not establish that the premises are too large. The underlying issue could be weak demand, an unsuitable daypart mix, inefficient use of the dining area, limited utilisation outside peak periods or a concept that requires more space by design.
How to Compare Restaurant Locations in Practice
A useful comparison should connect each variance with a management question. It should also compare restaurants against both peer locations and their own plan, because those views answer different questions.
A restaurant may rank below another location while still meeting its budget. Conversely, the group’s highest-profit location may still be performing materially below its own plan.
For plan-versus-actual analysis, use the sequence:
plan → actual → variance → factor → cause → action
For a practical multi-unit review, work through the analysis in the following order:
- Define the result. Decide which business outcome needs to be explained: operating profit, contribution, cash generation or the efficiency of a particular resource.
- Create a comparable peer group. Check format, scale, trading period, opening hours, service model and accounting treatment before comparing sites.
- Normalise the KPIs. Add measures per guest, transaction, paid labour hour, seat, square metre, trading day or other relevant resource.
- Identify material gaps. Focus on the differences large enough to influence the economic result rather than reviewing every metric equally.
- Break each result into factors. Move from profit to sales and costs; from sales to volume and average spend; from labour cost to hours and average cost; from food cost to mix, product cost and consumption variance.
- Investigate the reason each factor changed. Use operational dimensions such as daypart, channel, menu category, staffing period or inventory movement where the data are available.
- Separate controllable and external factors. Do not hold restaurant management responsible for a variable it cannot directly influence.
- Identify the operating practice behind the difference. Transfer a specific process or decision, not simply the KPI achieved by the better-performing site.
- Define the expected effect. Specify which factor should change and which financial or operating result should improve.
- Measure the result after implementation. Compare plan and actual performance again and verify that the intended factor changed rather than assuming the intervention worked.
This approach changes benchmarking from a reporting exercise into a management cycle.
Turn Restaurant Rankings into Transferable Operating Practices
Restaurant groups naturally want a simple ranking from best to worst. The problem is that a single composite score can hide economically important differences.
One restaurant may perform strongly in sales but poorly in labour productivity. Another may control food consumption well while under-utilising available space. A third may produce exceptional profit because of an unusually strong location rather than because its operating practices are superior.
A restaurant ranking is therefore most useful as a navigation layer. Once a location is identified as an outlier, management should be able to move into its factor profile:
sales → food and beverage → labour → other operating costs → resource utilisation → profit → plan versus actual
Transfer the practice, not the KPI
The real value of a multi-unit comparison appears when management can explain which controllable factor gives one restaurant an advantage.
Suppose a comparable site consistently produces more revenue per paid labour hour. The KPI itself cannot be copied. Management must determine which underlying practice supports it: better shift scheduling, different deployment by daypart, more effective workstation organisation, stronger demand matching or another verified operating difference.
The practice can then be tested in another restaurant, with the same factor and outcome measured before and after the change.
The sequence is:
performance gap → factor → verified cause → operating practice → change → repeat measurement
This is especially important when comparing restaurants across different cities or countries. Seasonality, tourism flows, local demand, delivery penetration, trading patterns and the characteristics of the property may differ materially. A practice should only be transferred when management has reasonable evidence that the performance advantage comes from the practice rather than from the operating environment.
Separate controllable factors from external conditions
Some variables can be managed directly: menu mix, pricing decisions, discount rules, purchasing arrangements, staffing schedules, waste control, operating processes and parts of capacity utilisation.
Other factors may be partly or largely external, including site characteristics, local demand patterns and seasonality.
The distinction matters because correlation is not proof of causation. A restaurant with higher labour productivity may also occupy a location with stronger demand. The productivity difference alone does not prove that its staffing model is better.
Move from one-off analysis to regular performance management
Restaurant comparisons become more useful when they are embedded in a consistent management model rather than prepared as isolated reports.
For each key result, the group can define the KPIs, factor tree, analytical dimensions and management responsibilities required to explain significant variances. This creates a repeatable route from the consolidated result to the operational cause.
The underlying model can form part of a broader multi-unit restaurant performance management system covering management reporting, budgeting, plan-versus-actual analysis and operational review.
Within the RestoFactor approach, methodology comes first: define the result, identify its factors, determine which data are needed, decide which variances require management action and establish how the outcome will be checked. Software can then automate data collection, calculations, reporting and recurring control of the model already defined by management.
A strong restaurant comparison should therefore answer more than “Which location is number one?” It should show why restaurants produce different economic results, which part of the difference is controllable, what action should be taken and how management will know whether the action worked.
The objective is not to rank restaurants. It is to turn differences between restaurants into better management decisions across the group.
