Optimising management control for a major retail group

In the retail sector, shrinkage (breakage, loss, expiry, theft, etc.) represents a significant source of lost revenue. For our customer, it represents nearly 1% of sales in a sector where margins are limited. Shrinkage can have numerous causes, and identifying them requires the analysis of a vast amount of data. Initially, management control tools were limited to top-down studies, which did not allow teams to carry out ad hoc explorations.

The aim was therefore to design analyses that would enable us to target the probable causes of shrinkage at all levels of the distribution chain, in order to initiate the most appropriate corrective actions.

We chose a visual analysis solution based on Tableau Software. In agreement with the customer, we adopted an iterative approach, building on existing systems and data to deliver results quickly. The first version of the solution has been developed in less than 5 days thanks to a proven methodology and technology adapted to ultra-agile projects, not to mention the responsiveness of our management control teams.

The solution was therefore based on :

• The implementation of advanced graphical representations highlighting elements such as “outliers” or exceptional values,
• The implementation of simpler visualisations for less advanced users,
• The use of cartographic representations to identify systemic or localised problems,
• The use of time curves to highlight any seasonal effects.

The solution was quickly deployed to 8,000 users with a return on investment (ROI) in less than 6 months.

By comparing the results of our analyses, we were able to highlight several very distinct phenomena using a singlea single analytical dashboardThese include identifying problems with accounting entries, purchasing policies, production forecasts, and handling issues in departments and warehouses. Thanks to staff training, these issues have now been resolved.

Thanks to data visualisation, we were able to highlight products that stood out at different levels of granularity. The solution provided business users with great flexibility, thanks to highly responsive analyses on large volumes of data.