Apoteket Upgrades to Machine Learning Forecasting with RELEX

One of Sweden’s leading pharmacy chains moves from time-series forecasting to machine learning for more adaptive and automated demand planning

Apoteket, one of Sweden’s leading pharmacy chains, has completed a migration to machine learning (ML)-based forecasting with RELEX Solutions, improving forecast accuracy by 7.8 pp across its store network and 8.9 pp through its e-commerce channel, achieving results that were significantly above the 2–5 pp improvement typically delivered by RELEX’s ML forecasting engine.

Apoteket serves customers across Sweden through a nationwide network of pharmacies and e-commerce operations, offering both prescription medicines and self-care products. Balancing seasonal demand, promotional activity, and the unique demand patterns of prescription medicines make accurate forecasting a complex planning challenge.

By transitioning to ML-based forecasting, Apoteket adopted a more adaptive and automated approach to demand planning, enabling more efficient, scalable, and data-driven planning processes. The transformation was led from the business side by Elisabet Plate and Linn Andrén, who worked closely with planners and RELEX throughout the implementation.

“Forecast accuracy helps demonstrate the impact of moving from time-series forecasting to machine learning, but the biggest value comes from how we work today,” said Elisabet Plate, Supply Chain Developer, Apoteket. “Seasonal changes, campaigns, and demand fluctuations are now handled much more effectively by the forecasting engine, reducing manual effort and allowing planners to focus on exceptions and business decisions. Forecast accuracy is only one part of the equation, but together with RELEX’s replenishment capabilities, it provides a strong foundation for better planning and continuous improvement.”

“Apoteket’s approach to this project reflects their commitment to getting the configuration right,” said Tiina Kanninen, SVP, Global Customer Success, RELEX Solutions. “That depth of engagement is what made results like these possible, and we are proud to keep building on this partnership.”