Image recognition technology, based on AI, is being used more and more frequently in the FMCG industry as well. Its main task is to strive for maximum accuracy of results in the shortest possible time. Currently, it seems to be a powerful tool in the hands of buyers, sales representatives, and logistics specialists, and the “perfect store” model is no longer just a marketing slogan. Comments Rafal Melanowicz, Director of the AI Department at eLeader.
Incorporating image recognition technology into the perfect store concept is an opportunity for a fresh start in shelf management. The “perfect store” concept in visual merchandising and category management is based on striving to achieve a product display appearance that will ensure the best sales. This is a very attractive model for evaluating a retail outlet, but in practice, it had a fundamental flaw – it relied on time-consuming and imprecise display audits. Without providing certainty as to whether it described the actual state, it was sidelined and gave way to sales data analysis and simple “is/is not” product presence checks. Only by objectifying and accelerating the display audit using artificial intelligence image analysis in the eLeader Shelf Recognition AI application has it become possible to use the perfect store concept to its full potential. We therefore incorporate into the perfect store a system that extracts a wealth of information from a photo taken with a smartphone: from product recognition to specific standard calculations, which are immediately used to calculate scores or even bonuses. On one hand, reports derived from processing actual product photos enable a reliable analysis of how the implementation of display standards affects sales. On the other hand, offloading the real-time calculation of display parameters from humans allows for the implementation of more ambitious and complex standards – in line with business requirements. At the same time, access to results in real-time enables the correction of displays according to guidelines while still during a sales or audit visit.
Overcoming limitations through image recognition technology encourages category managers, sales directors, and analysts to make wider use of indicators such as facing, share of shelf, presence in the “golden zone,” blocking, or other specific standards, and to build perfect store scoring models based on them. Examples of companies using eLeader Shelf Recognition AI on the Polish market include DANONE, NUTRICIA, and Maspex.







