Case study: supply chain forecasting as a service in consumer goods

For complex supply chain networks in a VUCA world, expertise in supply chain planning is key to meet demand most efficiently and reliably possible. It is essential for a competitive advantage. Suppose you do not have this expertise in-house. In that case, supply chain planning as a service can be a good alternative by utilizing the expertise of specialized data scientists to increase the quality of your forecasts. 

 

Our customer and their needs

Our customer is a sales organization of a multinational in consumer goods. The client followed an entirely manual approach for demand planning in predicting future demands. The average forecast accuracy was below their targets and expectations, severely impacting supply reliability (product availability, service levels, customer retention) and costs (inventory liabilities, scrapping, slow-moving stock). 

 

The main objectives of this initiative were the improvement of the forecast accuracy with a positive impact on win rates, reduction of order cancellations, and higher customer retention​ while simultaneously reducing the costs for scrapping and slow-moving stock and lowering inventory liabilities​. 

 

The project: demand analytics and forecasting as a service

We carried out a comprehensive quantitative analysis of product-line specific demand patterns to split demand patterns in different demand elements for forecasting​to reach the defined objectives. We built a machine learning-based forecast algorithm, using different internal data types (like open orders, opportunities, prices) and external data (web-scraping, market analysis, economic factors) to predict future demand with AIOintelligence

 

We designed and implemented demand analytics consisting of demand pattern analysis, demand segmentation, order history and order volume development, and forecast accuracy review dashboards in AIOinsights. We implemented forecasting algorithms and dashboards in a forecasting as a service approach, delivering weekly data updates and demand forecasts for a horizon of 18 months into the future. 

 

The result: up to 20% higher forecast accuracy

The project resulted in transparency on demand patterns and demand volumes over time, supporting data-driven decisions. Demand review meetings and alignments are now supported by transparent forecast accuracy and value-add reporting. Forecast accuracy on the appropriate planning level has increased up to 20% compared to the pure manual approach. Part of our service is feature engineering to ensure service quality and continuous improvement and coaching demand planners in incorporating the AI/ML-based forecast results in their daily planning routines. 

 

Our Supply Chain Planning managed service provides you the expertise to implement a seamlessly aligned end-to-end planning approach. Get in touch and see how we can help your business improve its competitiveness. 

 

New call-to-action

Dr. Adrian Reisch
By Dr. Adrian Reisch Dec 10, 2021 7:00:00 AM
Supply Chain Brief