Manufacturing companies strive to produce the right quantity of products to meet customer demand at all times. Failing to meet customer demand and excessive inventory are scenarios manufacturers are constantly working to avoid and mitigate. Mismatches like these between supply and demand can cost the company revenue.
Our client is one of the world’s largest toy manufacturers by revenue. A data-driven process helped the organization identify risks and opportunities to shift production and better meet forecasted customer demand.
Our client relies on multiple data sources and systems to guide their production planning process. They recently implemented a demand forecasting system that projected the inventory targets for large retail chains in North America.
Due to several factors, including disruptions in the global supply chain during the COVID-19 pandemic, the company’s production struggled with providing the necessary supply.
Their supply planners had to go through an arduous manual process to calculate shortfalls on a product-by-product basis across their portfolio of over 1,000 products carried by over 5,000 retail stores.
The company wanted a tool that could calculate an allocation plan driven by the forecasted demand of customers but constrained by the supply they could actually provide.
The solution would support numerous use cases:
Which products does the company need to make more of?
Which products can the company cut production on?
Should manufacturing of some products be shifted between weeks?
Which customers are likely to have orders that cannot be filled?
Which products are likely to be out of stock at retail stores?
Collaborating with stakeholders, the Keyrus team developed an Alteryx workflow to calculate an allocation plan for an 18-week horizon. The workflow connects to multiple types of data sources, including SAP Hana, SAP BW, SQL Server, and Impala.
After preparing and blending the data sources, the planned product supply is allocated to customers in two rounds. The first round of product allocation is prioritized by demand and the second is based on forecasted consumer sales.
The first round of allocation prioritizes the demand needed to support promotions and new launches. These allocations ensure that prior commitments are met before allocating towards the forecasted demand.
In the second round, the allocation is primarily driven by forecasted consumer sales at retail stores with the objective of holding a minimum inventory level at each store. Since an allocation in the first week impacts the inventory levels for the whole horizon, the allocation must be conducted in an iterative manner. Key measures such as allocation quantities, store inventories, and remaining supply available must also remain updated.
In the event that not all demand can be met, the workflow spreads the shortfall out as equally as possible across customers to ensure fairness.
A Tableau dashboard visualizes the results of the analysis and helps the client quickly identify some projected key performance indicators in the supply chain:
Allocation quantity by product, customer, and week
Supply shortfall by product and week
Excess supply by product and week
Retail inventory levels by customer, product, and week
The dashboard saves supply planners significant time when identifying risks and opportunities. It also provides customer relations managers a more complete picture of how their customers can be supported in the upcoming weeks. Executives can also easily glean insight into the projected health of the supply chain from the dashboard.
Future opportunities of this initiative include adding capabilities for scenario planning and feeding back into the production planning process, further optimizing the company’s supply chain operations.