Join RELEX at NRF 2025 – Retail’s BIG SHOW! January 11-14, 2025 | Request a meeting

menu-close

Case study: Atria
AI-Powered Demand Planning Accuracy in Consumer Goods Manufacturing

icon-checkmark

98.1%

forecast accuracy on the week level

icon-checkmark

13%

reduction in manual forecast changes

As one of Finland’s largest suppliers of meat products and the leading food company in Northern Europe, Atria has been using RELEX demand planning solution for several years, continuously improving their planning processes.

To further improve their ability to manage the intricacies of meat products—a complex and challenging category—they recently introduced RELEX machine learning capabilities into their supply chain processes to drive substantial improvements to forecast accuracy.

Managing a Complex Supply Chain

Supply chain management in the meat product sector is complex. For example, Atria’s end products have short shelf lives, requiring highly accurate day-level forecasts to ensure high availability and minimal waste. In addition, Atria needed accurate long-term planning to balance supply with demand, as breeding and raising animals at the start of the supply chain can take anywhere from 10 weeks to several years.

Finally, Atria needed to better manage the impact of seasonality on their demand throughout the year, as well as the impact of their retail clients’ business decisions. They wanted to improve their ability to capture the impact of retailer decisions such as periodic assortment changes, promotions, or how they choose to prioritize products and stock up in preparation for big holidays like Christmas.

Building from a Solid Foundation in Demand Planning

At the start of Atria’s journey with the RELEX demand planning solution, their primary goal was to drive more granular forecasting on the retail chain level. Granular forecasting was central to their ability to effectively account for their retail customers’ various assortment decisions, promotions, and holiday- and event-driven sales. RELEX successfully increased forecast automation, enabling Atria to move to the level of forecast granularity they needed.

In addition, Atria was able to rationalize their application portfolio, replacing two custom-built planning tools for short- and long-term planning with the RELEX unified demand planning solution. This unification allowed them to seamlessly combine short-term demand sensing with long-term demand planning within the same tool—even supporting their rolling budget planning.  

Following the success of the implementation in Finland, Atria expanded the use of the RELEX demand planning solution in additional countries as they refined their planning processes.

AI-Powered Demand Forecasting Further Improves Planning Accuracy

Results at a Glance:

  • 98.1% forecast accuracy on the week level
  • 13% reduction in manual forecast changes

Atria chose to implement RELEX machine learning-based forecasting capabilities to help them better manage the complexities of the meat products sector.

Machine learning-based forecasting drove further improvements to Atria’s forecast accuracy, which is now 98.1 % on the week level, and led to a 13% reduction in manual changes to the forecasts. These KPIs were driven in large part by machine learning’s ability to more accurately forecast demand during high-volume holiday periods, when shoppers tend to splurge on more festive foods.

Atria’s new forecasting approach now automatically detects and reacts to step-changes in demand—a capability that proved especially valuable throughout the COVID-19 pandemic, which impacted demand through lockdowns and social distancing recommendations. Atria has also seen more stable forecasts, even following large sales peaks during their busiest seasons. This stability has greatly improved their long-term planning.

“Thanks to our initial RELEX implementation, we were able to begin from a solid foundation of already high forecast accuracy,” said Tapani Potka, SVP, Delivery Chain Management for Atria. “However, following the introduction of machine learning, we were extraordinarily impressed to see just how much further RELEX algorithms could improve our forecast accuracy and stability—improvements critical to our ability to manage the complexity specific to our industry and business model.”

Machine Learning Unlocks New Opportunities

One of machine learning’s greatest competitive advantages is the ability to leverage large amounts of external data to improve planning accuracy on a scale not possible without the support of AI. Machine learning-based demand forecasting has unlocked new potential for Atria’s development, including opportunities to leverage retail price data in planning and incorporate data on future assortment changes from retail partners into automatically updated forecasts.

“For nearly a decade, we’ve had an exciting, innovative partnership with RELEX. We continue to reap the benefits of that relationship while expanding our use of their solution’s capabilities,” said Pekka Korpeinen, Director, Steering & Planning for Atria. “It’s apparent to me that RELEX and Atria share the same level of ambition when it comes to exploring how consumer packaged goods companies can derive maximum value from modern planning technologies. The addition of RELEX machine learning-based forecasting has greatly contributed to the continued success of our business.”

The results

icon-checkmark

Automatic detections and reactions

to step-changes in demand

icon-checkmark

More stable forecasts

even after large sales peaks during Atria's busiest seasons