Optimizing stock levels in DIY retail. How Hornbach was able to reduce capital expenditure by introducing advanced demand forecasting
How can you refine your inventory management and reduce capital expenditure with AI and data-driven insights?Â
Discover how Hornbach, a leading European DIY retailer, revolutionized its supply chain by implementing advanced demand forecasting techniques. This strategic move enabled Hornbach to optimize stock levels across its stores, significantly improving cost efficiency and customer satisfaction
Client Introduction
Head of Store Process Development, Strategy & Innovation at Hornbach
- Industry: DIY Retail
- Role: Head of Store Process Development, Strategy & Innovation
- Company Size: €6.3 billion in revenue; 169 stores in Europe, of which 17 are in the Netherlands (approximately 226,617 square meters).
Hornbach is a prominent DIY retailer operating 169 stores across Europe. Known for a wide range of quality DIY products at competitive prices, Hornbach has remained family-managed since 1877, emphasizing a strong customer-first approach.
Â
Typically, DIY retailers want to provide customers with a one-stop-shopping experience, implying a high service and availability level regardless of the project they wish to complete. Offering such a value proposition is by nature capital-intensive and is, given the growth of online and offline channels, requiring ever more capital expenditure.
The Challenge of Reducing Capital Expenditure
Hornbach’s commitment to customer satisfaction hinges on its ability to offer a complete range of items needed for DIY projects in a single store visit. However, their existing forecasting method, an 8-week sales data analysis, was not sufficiently accurate, leading to overstocking and underutilization of floor space. This not only tied up valuable working capital but also compromised the store layout and customer experience.
Introducing Advance Demand Forecasting
Recognizing the need for a more sophisticated approach to managing inventory more effectively, our team convinced Hornbach to develop a cutting-edge forecasting solution:
Enhanced Data Architecture
Parallel to developing the forecasting model, Crunch worked closely with Hornbach’s data teams to upgrade their data infrastructure. This strategic enhancement supported the advanced analytics capabilities required for modern inventory management.Â
Hornbach now has a more data-science-ready environment that connects to its legacy data warehousing environment. Such a data-science-ready environment runs a number of data-science-enabling technologies like Jupyter Notebooks to visualize data, Git to track code, Docker to handle dependencies, Kubernetes for managing virtual machines, and Airflow for scheduling.Â
Key Performance Indicators
The primary metric for assessing the impact of the new system was the weighted Mean Absolute Percentage Error (wMAPE), which showed considerable improvement in forecasting accuracy.Â
- AI-Powered Forecasting Model
Crunch’s solution involved the deployment of a LightGBM-based machine learning model that improved prediction accuracy significantly.Â
By categorizing inventory articles into demand types (smooth, erratic, intermittent, lumpy) and focusing initially on smooth and erratic articles, the model provided actionable insights that directly supported better stock control.
Results
Optimized Inventory Levels
The implementation of the AI forecasting model enabled Hornbach to more accurately predict future sales and adjust their stock levels accordingly. This not only reduced the necessity for excess working capital but also minimized the risk of stockouts, thereby ensuring that customers always find what they need.
Improved Store ProcessesÂ
In the store, several stock pickers now use their trusted stock-picking app, connected to said data infrastructure, to input current inventories. These inventory numbers are then compared with the forecast and a set of business rules. Based on this comparison, the stock-picking app provides the correct number to ensure proper in-store replenishment.
Conclusion
Hornbach’s journey illustrates the power of AI in transforming retail operations. By adopting advanced forecasting models and improving data infrastructure, Hornbach now enjoys reduced operational costs, improved customer experience, and a competitive edge in the DIY retail market.
Interested in learning more about how data-driven solutions can transform your business? Have a look at our cases for more inspiring examples, or follow us on LinkedIn for the latest updates and insights. For a personalized consultation or to discuss how we can assist your company, please contact us.
