All these elements can vary over time, the surface area of a point of sale can change following an expansion of facilities, staff training, a branch can be accredited for new services, a store can change its banner or concept… it is then important to be able to track these events over time to analyse the comparability.
Time variability in the comparable: when performing detailed analysis, time variability must be included in the analysis. Many sales activities are affected by this variability. A month from a year to another doesn’t necessarily have the same number of working days. The variation can go from 3 to 4 days if we take into account the impact of bank holidays.
Thus, this gap can constitute more than 10% of activity. To erase the “month-effect”, we work more and more often on a daily basis, (comparison of December 24th from a financial year to another) or on weekly basis, by spreading months in 4/4/5 (month 1, first 4 weeks to compare with the first 4 weeks of the previous financial year).
However, if we want to keep a monthly analysis, it is necessary to call upon prorate based, which allow to erase the number differential of working days and to make them comparable. For instance, the months of January 2011 and 2012, we apply the following formula: January 2012 = 1.12 x January 2011.
Defining management indicators
It is necessary to identify management indicators, the relevant ratios and business indicators because for every business and every vertical correspond specificities which define the particularities of its activity. We follow the number of missions in the real estate, the revenue by tax rate in the pharmacy (evaluation of mix products) or even the average basket or the stock rotations.
Define the ratios to correlate to the elements of the frame of reference is essential. To complicate the problem, the management indicators usually used are supported on different economic indicators (for example the sale area) and on management data (for example the revenue). Both can vary over time, this creates a new complexity when we want to analyse over two financial year stores that generate revenue between 15 and 20K$ per sq ft. When surface area changed during the current financial year, we will have 6 months of revenue with the initial surface area and 6 months of revenue after expansion.
In order to have a complete analysis, we must store economic data, such as captured in operational tools (sales and financial systems), the elements from the frame of reference, and all other events that occurred during the year. For example, the fact that a store has changed its concept in November 25th of 2011 or that its surface area will change in May 4th of 2012. All these data have to be listed in an Infocenter which has to allow to simply manipulate data and keeping the possibility of comparability at the same time.
Initially, and for a matter of simplicity, we used to exclude from analysis everything that was not comparable.
Every unit that had occurred modifications over the financial year, a significant event like an acquisition, disposal or any other modification of a key element from the frame of reference, would be excluded from the analysis. Therefore, we kept the smallest common denominator to perform the comparison. Technically, a marker identified every point of sale that has occurred a modification and was excluded of the consolidation. We limited ourselves to consolidate the current year with the perimeter of the previous year or the previous year with the current perimeter.
Today, with the evolution of technologies, possibilities of storage both multidimensional and relational, and with the maturity of the solutions, it is finally possible to realize analysis more detailed which rely on data from the Infocenter and which reprocess the comparability, thanks to the creation of dynamic panels.
This panel consists of a list of units that have the criteria on the two periods we want to compare. These panels rely on the whole frame of reference (surface area, type of contract, concept, opening date…) and allow crossing several criteria. Thus, if someone wants to measure the evolution of revenue between March 2010 and March 2011 for a store concept, he has at his disposal several analyses: the constant perimeter (units under concept retained in 2010 and 2011) or the current perimeter (all the units that match criteria retained in 2010 or 2011).
Moreover, within the selection, we can isolate new events that have an impact on the comparability. Therefore, the launch of a new activity, a new line of products, or a marketing campaign can be dismissed in order to reach a more detailed analysis of the performance.
It results from this that detailed analyses are necessary for the decision-making process from executives and give them the tools for a high responsiveness in a constant changing environment.