Is it possible to improve stock management and warehouse operations to minimize negative economic impacts in large-scale distribution companies? The answer is yes, thanks to the integration of predictive models into management processes, following a data-driven approach. This is the case with the project developed by Adiacent for Sidal.
Sidal, which stands for Italian Society for the Distribution of Groceries, has been operating in the wholesale distribution of food and other goods since 1974. The company primarily serves professionals in the horeca sector (hotels, restaurants, and cafes) and the retail sector, including grocery stores, delicatessens, butchers, and fish markets. In 1996, Sidal strengthened its market presence by introducing physical cash & carry stores under the Zona brand in Tuscany, Liguria, and Sardinia. These stores offer a wide range of products at competitive prices, serving professionals efficiently and effectively. Today, Zona operates 10 cash & carry stores, employs 272 workers, and reports an annual turnover of 147 million euros.
Solution and Strategies
· Analytics Intelligence
Zona decided to optimize its stock management and warehouse operations through artificial intelligence, aiming to better understand product depreciation and introduce specific strategies, such as the “zero sales” approach, to mitigate the negative economic impact of product devaluation.
In this initiative, Adiacent, Zona’s digital partner, played a key role. The project, launched in January 2023 and operational by early 2024, was developed in five phases. The first phase involved analyzing available data, followed by the creation of an initial draft and a proof of concept to test the project’s feasibility. Subsequently, prescriptive and proactive analysis models were developed, and the final phase focused on data tuning.
Data analysis and the creation of the algorithm
During the data analysis phase, it was essential to inventory the available information and thoroughly understand the company’s needs to design robust and structured technical solutions. While creating the proof of concept, Zona’s main requirements emerged: the creation of product and supplier clusters, the categorization and rating of items based on factors such as their placement in physical stores, profitability, units sold, depreciation rate, and expired units, as well as the categorization of suppliers based on delivery times and unfulfilled orders.
Product depreciation posed one of the most significant challenges. Using an advanced algorithm, the probability of a product depreciating or expiring was predicted, enabling proactive stock management and reducing the negative economic impact of waste. This strategy aims to optimize the company’s turnover, for example, by moving products close to their expiration date between stores to make them available to different customers, while also improving warehouse staff productivity.
The evaluation is based on a wide range of data, including the number of orders per supplier, warehouse handling, and product shelf life. To ensure timely and effective management of product depreciation, call-to-action procedures were implemented, with detailed reports and notifications via Microsoft Teams.
Forecasting to Optimize Processes
Thanks to these implementations, an integrated predictive system was created to identify potential product depreciation and provide prescriptive mechanisms to reduce its negative effects, maximizing overall economic value. The “zero sales” strategy plays a crucial role in stock management and the optimization of Zona’s warehouse operations, enhancing customer experience, improving stock and operational cost management, maximizing sales and profitability, and enabling smarter supply chain management.
Special attention was given to training four key prescriptive models, each designed to make specific predictions: average daily stock, minimum daily stock, warehouse exits/total monthly sales, and warehouse exits/maximum daily sales. The development process followed a data-driven approach, and each model was designed to adapt to new warehouse, logistics, and sales needs, ensuring long-term reliability.
Looking to the future, “The integration of artificial intelligence – stated Simone Grossi, Zona’s buyer – may open new paths toward personalizing the customer experience. Advanced data analysis could enable us to predict customer preferences, offering personalized promotions and targeted services.”
