WARBURTONS SUPPLY CHAIN
IMPROVING PROMOTIONAL FORECAST ACCURACY WITH MACHINE LEARNING
Discover how Warburtons, a leading bakery business in the UK, partnered with FuturMaster to enhance their promotional forecast accuracy through machine learning. The commercial forecasting team at Warburtons sought continuous improvements in efficiency and forecast accuracy, and machine learning emerged as a potential solution. The webinar presentation by Warburtons' commercial forecasting manager explores their journey with FuturMaster, from the initial problem of promotional forecast accuracy to the results and benefits achieved. The data collection process was a significant challenge, but with the collaboration of FuturMaster, they organized and analyzed the required data to train the machine learning model. Iterative testing and refining led to improved accuracy, surpassing their internal forecasting efforts. By incorporating clean pricing data and stock-out information, they achieved a forecast accuracy of 75%.
Watch the recording session to gain insights into their machine learning trial and its impact on Warburtons' demand planning.
- Introduction to Warburtons
- What is Machine Learning?
- Warburtons use cases
- Statistical approach vs machine learning
- discovering and exploring the data
- Process and challenges
- Results and benefits
Case Study Abstract
By leveraging the power of pattern recognition and learning algorithms, Warburtons aimed to enhance efficiency and accuracy in their demand planning process. The initial phase involved accessing and organizing diverse data sources, including FuturMaster and external pricing records. FuturMaster then trained the machine learning model, iteratively refining it to improve accuracy. By incorporating additional clean pricing data and stock-out information, they achieved a forecast accuracy of 75%, surpassing their internal forecasting efforts.
The presentation highlights the potential uses of machine learning in various aspects of Warburtons' operations and explores the benefits derived from the trial. Download the webinar presentation deck for in-depth information and actionable takeaways.