Roadmap for building effective business intelligence for a restaurant chain – 5 key indicators

I have been working in the field of corporate business intelligence for more than 10 years, and based on this experience, I can say that the first obstacle to implementation is often a lack of understanding of what value can be obtained and what data should be analyzed for this.

This article is the first in a series where I will cover the optimal sequence of information processing that allows you to get the maximum value from the data. I will consider the example of a restaurant chain, since my experience was largely related to analytics for the “big three” fast food.

Here is the roadmap we will follow, and today the focus is on financial indicators.

The main source of data that every restaurant has is the cash register system. Even at the level of checks, you can get detailed analytical sections, but not everyone “embroiders” checks to detailed indicators, but only works with totals. And here is what I propose to analyze.

Metric 1: Data from receipts

Like-for-like – comparable sales

This indicator reflects how much the results of the current period have changed relative to the past, and also allows you to identify the reasons for these changes. It gives an objective picture of the dynamics of sales, as it takes into account only the restaurants that worked in the two compared periods when comparing, and excludes new or closed outlets. The use of like-for-like analysis also excludes the influence of seasonality and helps to predict the dynamics of future periods.

Example: 10 new restaurants opened this year compared to the 10 that were open last year. Revenue increased by 2 times, but those establishments that worked in the past period grew only by 10%.

A common mistake here is a one-to-one comparison of days: it is incorrect to compare days if the date fell on days that were different in terms of loading last year. For example, February 27, 2022 is Sunday, and in 2023 it is Monday.

Product mix, dish rating and demand structure

Product mix – a complete product range offered to the buyer.

Ratings are formed as the number of dishes sold per N receipts. Along with the number of dry checks (checks with only 1 dish) and the structure of additional purchases of goods for certain dishes or combo sets, this data allows you to offer optimal marketing campaigns and track their effectiveness.

Lack of such information can cost a company dearly. I’ll analyze it using the example of a promotion, when, upon reaching the check amount of 150 rubles, the client could purchase a dessert for 1 ruble. However, due to an error in setting up the cash system, if the check was 100 rubles and dessert was added for another 100 rubles, the value of the check became 101 rubles. Due to the risk of fraud (fraud), the promotion was prematurely closed within three days.

But as a retrospective analysis showed, the cases of using an error were less than 3% of the total number of promotional checks, while the average check with a share was 132 rubles higher than usual. This information could significantly influence the decision to continue the action, but without analytics it simply did not exist. As a result, the company missed the benefit of stopping the action and lost the funds invested in advertising: billboards, flyers, radio advertising, etc.

There is also an opposite example, when, when analyzing the “Buy a drink for 1 ruble with a coupon” campaign, it was revealed that about 30% of checks contained only a promotional drink without additional purchases, and a slight increase in the average check completely ate up the promotion budget.

Plan-fact analysis/ABC analysis/XYZ analysis 

Plan-fact analysis is a comparison of planned and actual indicators for a restaurant in the same analytical sections for the same period.

ABC analysis is a marketing tool that allows you to highlight the leading and outsider dishes and allows restaurants to abandon unprofitable positions.

XYZ analysis – indicates the regularity of sales: how often customers buy a product.

It should be borne in mind that in reality, the analyzed population is influenced by not one, but several factors at once. It helps to cope with this limitation multivariate ABC analysis.

However, the information that we extract from checks often cannot answer the question of efficiency. How to understand how well a particular restaurant is performing relative to the rest?

For example, one of the restaurants shows one of the highest revenues in the network. At the same time, its financial result is actually negative, since everything is eaten up by the cost of premises in the center with high traffic.

Metric 2: SPMH (ITPH, GCPH) – relative productivity per employee hour

To analyze SPMH, in addition to information from checks, information about the time of work of the staff is required. Accordingly, a system for recording the working hours of employees (passes, biometrics) should be introduced.

Based on SPMH, a schedule (plan) is drawn up, which is then compared with the fact. The difficulty lies in the fact that it is almost impossible to find a boxed solution that would offer the possibility of simultaneous time tracking with biometrics and flexible schedule management.

In addition, when planning the SPMH, it is important to take into account changes within the day and week (peak load by day and hour) in order to get a correct picture when comparing.

The SPMH indicator can be used, among other things, to motivate staff and gamify the process between network points. Abroad, by the way, this is a very common practice.

Metric 3: Food Cost

The calculation of the cost of production in the company is influenced by a lot of features: batch accounting, accounting or not accounting for the logistics and warehouse components in the cost, FIFO (first in – first out – goods that were delivered earlier), LIFO (last in – first out) models – goods that arrived last, should be the first to write off the register) or the average cost (calculation by arithmetic average).

In my practice, some restaurateurs calculated the cost at the level of each ingredient for each day, while others, on the contrary, simplified the calculation to the maximum and included the cost for the dish as a whole for a month throughout the entire network.

Of course, it is ideal when the margin of each check is known, which means that there is an idea of ​​the cost of each product sold in the restaurant. However, when analyzing Food Cost, I recommend finding a balance between the complexity of the cost calculations and the applicability of the results of the analysis.

Metric 4: Labor Cost

This indicator reflects the proportion of revenue that is spent on staffing and, along with Food Cost, is one of the two main indicators of restaurant costs.

The main problem with calculating Labor Cost is that data is often prepared manually, as many components of LC are difficult to automate. In addition, there are many methodological difficulties in the calculations: forecasting benefits, bonuses and overtime, accounting for interns, part-time workers, couriers and cleaners, outstaffing (staffing). Taking into account Labor Cost at the level of each restaurant in the context of different parameters is not an easy task, but analytics can do it.

I consider a system where for each component of LC (bonuses, night cleaning outstaffing, etc.) a percentage is set as an example of a successful solution for monitoring this indicator, and if it is exceeded, a report is automatically sent to the analytical service about what to pay attention to.

A small observation from the analyst: in general, in food retail in 2022, the Labor Cost indicator increased by several points.

Metric 5: EBITDA – earnings before interest, taxes and depreciation

This is one of the most important performance metrics to automate. The correctness of its calculation depends, first of all, on how accurate your data on Food Cost and Labor Cost are – for a restaurant, these are the main cost components in the EBITDA structure.

In addition, the calculation requires data not only from restaurants, but also from the back office (1C Enterprise, warehouse management, ERP, etc.) and head office (for example, fines and costs of a legal entity). Since there are a lot of variable costs, it is important to organize a system for automatically loading and checking data in order to quickly obtain a final indicator.

Resulting Performance Evaluation: P&L – Profit and Loss Statement

Companies approach the formation of a P&L report with a sufficient degree of analytical maturity. This report is a kind of cross-section of the entire activity of the restaurant.

It is important not only to form a general idea of ​​the financial result at the level of the top management of the company, but also to provide each restaurant with information about their financial result. If there are over 500 restaurants in the chain, this can become a problem.

History is ubiquitous when reports are cut and sent by hand. Meanwhile, there are modules for automated distribution of reports that allow you to deliver information in a standardized form to the maximum number of recipients, even without the need for licensing user access.

Instead of total

• The depth of data analysis depends on the maturity of the company, but the opposite is also true: the more fully the company begins to work with data, the faster it develops.

• The revenue is not indicative – the activity of the restaurant chain is too multifactorial to rely only on the dynamics of revenue when evaluating efficiency.

• Performance indicators should always be placed in context: take into account peaks, weekends, seasonality and other nuances that we described above.

• Analysis requires multiple data sources. However, before working with the number of sources, it is important to take care of the quality of the information being analyzed.