From Economic Signals to SKU Inventory Decisions
From Economic Signals to SKU Inventory Decisions
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
Planning & Budgeting
The gap between a market signal and a product order
Economic indicators can reveal broad demand changes, but an individual stock-keeping unit (SKU) is often too noisy for a direct macroeconomic relationship. Inventory teams still need forecasts for specific products, lead times and service targets. This 2025 study addresses the bridge between those levels.
The method tested by the researchers
Using data from a global manufacturer, the authors selected useful leading indicators at an aggregated demand level, then combined that information with forecasts lower in the product hierarchy. They produced probabilistic SKU-level forecasts and tested the resulting inventory decisions under both backorder and lost-sales settings and across different lead times.
Which level should carry the signal?
Imagine a manufacturer sells several variants of the same product. Total demand for the product family may move with the broader market, while demand for individual variants is driven by promotions, customer orders or substitutions. A family-level signal can inform the overall volume assumption. The demand history of each variant then helps determine how that volume is distributed across items.
Why the inventory test matters
A forecast can be more accurate on average while still missing the periods that matter most for replenishment. A shortage and an excess unit have different consequences, and the time needed to restock changes which forecast errors matter. Testing a forecast through an inventory simulation ties these errors to decisions about how much to order and when. The results depend on the assumptions used in that simulation.
What the case found
The published case reports better forecasts and inventory performance when leading information was introduced at upper levels of the hierarchy and combined with product-level patterns. The paper also found that inserting those indicators directly into noisy SKU-level forecasts could hurt performance. The method of integration—not simply adding more variables—mattered.
What planners can apply
For operations, the lesson is to connect a broad external signal to the level where replenishment decisions are actually made, then test both forecast and inventory outcomes. Morrow can make company-relevant external signals accessible, but this study does not test Morrow. Its results are one manufacturer’s research case, not a guaranteed stock reduction or service-level improvement for users.
Want to know whether external indicators can improve your company’s forecast?
Want to know whether external indicators can improve your company’s forecast?
More to Discover
From Economic Signals to SKU Inventory Decisions
From Economic Signals to SKU Inventory Decisions
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
Planning & Budgeting
The gap between a market signal and a product order
Economic indicators can reveal broad demand changes, but an individual stock-keeping unit (SKU) is often too noisy for a direct macroeconomic relationship. Inventory teams still need forecasts for specific products, lead times and service targets. This 2025 study addresses the bridge between those levels.
The method tested by the researchers
Using data from a global manufacturer, the authors selected useful leading indicators at an aggregated demand level, then combined that information with forecasts lower in the product hierarchy. They produced probabilistic SKU-level forecasts and tested the resulting inventory decisions under both backorder and lost-sales settings and across different lead times.
Which level should carry the signal?
Imagine a manufacturer sells several variants of the same product. Total demand for the product family may move with the broader market, while demand for individual variants is driven by promotions, customer orders or substitutions. A family-level signal can inform the overall volume assumption. The demand history of each variant then helps determine how that volume is distributed across items.
Why the inventory test matters
A forecast can be more accurate on average while still missing the periods that matter most for replenishment. A shortage and an excess unit have different consequences, and the time needed to restock changes which forecast errors matter. Testing a forecast through an inventory simulation ties these errors to decisions about how much to order and when. The results depend on the assumptions used in that simulation.
What the case found
The published case reports better forecasts and inventory performance when leading information was introduced at upper levels of the hierarchy and combined with product-level patterns. The paper also found that inserting those indicators directly into noisy SKU-level forecasts could hurt performance. The method of integration—not simply adding more variables—mattered.
What planners can apply
For operations, the lesson is to connect a broad external signal to the level where replenishment decisions are actually made, then test both forecast and inventory outcomes. Morrow can make company-relevant external signals accessible, but this study does not test Morrow. Its results are one manufacturer’s research case, not a guaranteed stock reduction or service-level improvement for users.
Want to know whether external indicators can improve your company’s forecast?
Want to know whether external indicators can improve your company’s forecast?
More to Discover
From Economic Signals to SKU Inventory Decisions
From Economic Signals to SKU Inventory Decisions
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
A 2025 manufacturer case found that leading indicators at aggregate level can improve SKU-level forecasting and inventory decisions when combined with hierarchical forecasts.
Planning & Budgeting
The gap between a market signal and a product order
Economic indicators can reveal broad demand changes, but an individual stock-keeping unit (SKU) is often too noisy for a direct macroeconomic relationship. Inventory teams still need forecasts for specific products, lead times and service targets. This 2025 study addresses the bridge between those levels.
The method tested by the researchers
Using data from a global manufacturer, the authors selected useful leading indicators at an aggregated demand level, then combined that information with forecasts lower in the product hierarchy. They produced probabilistic SKU-level forecasts and tested the resulting inventory decisions under both backorder and lost-sales settings and across different lead times.
Which level should carry the signal?
Imagine a manufacturer sells several variants of the same product. Total demand for the product family may move with the broader market, while demand for individual variants is driven by promotions, customer orders or substitutions. A family-level signal can inform the overall volume assumption. The demand history of each variant then helps determine how that volume is distributed across items.
Why the inventory test matters
A forecast can be more accurate on average while still missing the periods that matter most for replenishment. A shortage and an excess unit have different consequences, and the time needed to restock changes which forecast errors matter. Testing a forecast through an inventory simulation ties these errors to decisions about how much to order and when. The results depend on the assumptions used in that simulation.
What the case found
The published case reports better forecasts and inventory performance when leading information was introduced at upper levels of the hierarchy and combined with product-level patterns. The paper also found that inserting those indicators directly into noisy SKU-level forecasts could hurt performance. The method of integration—not simply adding more variables—mattered.
What planners can apply
For operations, the lesson is to connect a broad external signal to the level where replenishment decisions are actually made, then test both forecast and inventory outcomes. Morrow can make company-relevant external signals accessible, but this study does not test Morrow. Its results are one manufacturer’s research case, not a guaranteed stock reduction or service-level improvement for users.