MORE THAN ANY OTHER FIELD, retailing is in a constant state of flux. It goes without saying that, in order to make a pertinent analysis, we need to introduce comparability of elements, what our Anglo-Saxon friends call "like for like". Using this method of analysis means excluding all effects of expansion, acquisition, or any other event that artificially modifies the evolution of these figures. For example, if we're looking to compare two financial years, we'll exclude any sales resulting from an acquisition during the year.
Keeping track of comparability parameters is a complex business. Beyond a simple acquisition, which is the simplest case, many other parameters can vary: a temporary closure lasting a few days, work that changes the sales area, equipment that evolves... and comparability is put to the test. These parameters are numerous, and they are all linked to temporal events, i.e. a start date, an end date and a nature that makes it possible to understand and understand these evolutions. The challenge of such fine-tuned comparability management is to decipher and understand the business, which is essential for all management. Comparability is the key to the benchmarking needed to assess performance. The absence of like for like is like not having a GPS in your car: you don't know where you are anymore, and it's difficult to find your way.
Definition of the repository and its complexity
Defining a benchmark: The starting point for any comparable analysis is to define a frame of reference that enables a precise description of your network and relevant analysis criteria. In sales networks with many points of sale, this is often a tedious task, but one that can be very structuring. There are generally four types of elements in this frame of reference: economic (location area, surface area, approvals, workforce, concept, equipment, opening hours, creation date, etc.), legal (type of contract, closing month, type of company, start and end of contract, etc.), organizational (coordinator, trainer, geographical area, etc.) and management/ratios (sales range, margin rate range, fixed asset obsolescence rate, etc.).
All these elements can vary over time: the surface area of a sales outlet can be modified following expansion work, a branch can be accredited for new services following staff training, a store can change its name or concept following work... It is therefore important to be able to trace these events over time in order to analyze their comparability.
Time variability in the comparable : Time variability must be taken into account in fine analyses. Many business activities are influenced by this variability. A month does not necessarily have the same number of working days from one year to the next, and the variation can be as much as 3 or 4 days if the impact of public holidays is taken into account. This difference can represent more than 10% of activity. To erase the month effect, we increasingly work on a daily basis (comparing December 24 from one year to the next) or on a weekly basis, dividing the months into 4/4/5 (month 1, first 4 weeks compared with the first 4 weeks of the previous year). However, if you want to keep to a monthly analysis, you'll need to use prorations to erase the difference in the number of working days and make them comparable. For example, to compare the months of January 2011 and 2012, we apply the following formula: January 2012 = 1.12 x January 2011.
Definition of management indicators
It's essential to identify the right management indicators, ratios and business indicators, as each business and vertical has its own specific features that define the particularities of its activity. For example, we track the number of mandates in real estate, sales by VAT rate in pharmaceuticals (product mix evaluation), or the average basket or stock rotation.
Defining the ratios to be correlated with the benchmark elements is essential. To complicate matters further, the management indicators generally used are based on different economic indicators (e.g. sales area) and management data (e.g. sales). Since both can vary over time, this creates a new complexity when we want to analyze stores with sales of between €15 and €20 per m2 over two financial years. When the surface area has changed during the year, we'll have, for example, 6 months' sales with the initial surface area and 6 months' sales after renovation.
For the analysis to be complete, we need to store both the original economic data, as captured in the operational tools (sales management or financial management), the elements of the repository, as well as all events occurring during the year. For example, the fact that a store changed concept on November 25, 2011, or that its surface area will change on May 4, 2012. All this data needs to be stored in an info-center, which must enable simple manipulation of the data while maintaining comparability.
The solution
The need for a dynamic data processing/analysis tool. Initially, and for the sake of simplicity, anything that was not comparable was excluded from the analyses. All units that had undergone changes during the year, or had undergone a significant event such as an acquisition, disposal, or any other modification to a key element of the frame of reference, were excluded from the analysis. In this way, the lowest common denominator was retained for comparison. Technically, a marker identified any sales outlet that had undergone a change, and excluded it from the consolidation. We would then simply consolidate the current year with last year's perimeter, or last year with the current perimeter.
Today, with the evolution of technologies, the possibilities of both multidimensional and relational storage, and the maturity of solutions, it is finally possible to carry out finer-grained analyses based on data from the infocenter and reprocessing comparability, thanks to the creation of dynamic panels.
This panel consists of a list of units with the same criteria over the two periods to be compared. These panels are based on the entire frame of reference (surface area, type of contract, concept, opening date, etc.) and make it possible to cross-reference several criteria. For example, if you want to measure sales growth between March 2010 and March 2011 for a given store concept, you can choose between two types of analysis: constant perimeter (units with the same concept in both 2010 and 2011) and current perimeter (all units meeting the criteria in 2010 or 2011).
What's more, within this selection, we can isolate new events that have an impact on comparability. For example, the launch of a new activity, a new product line, or a marketing campaign can be excluded to enable a precise analysis of performance.
It is clear, therefore, that detailed analyses to support management decision-making are necessary, and make it possible to be highly responsive in a changing world.
We'll be hosting a free webinar on this topic on July 2 from 11:00 to 11:45.
