Demand Planning & Forecasting Resources

Demand planning has a math problem and an organizational problem, and most teams only solve for the first one. The forecasting models get more sophisticated every year — more inputs, tighter algorithms, better handling of seasonality and promotions — yet forecast accuracy at the SKU-location level often doesn't move much, because the model was never the real constraint. The real constraint is that demand signals live in one system, planning assumptions live in another, and the people closest to actual customer behavior — sales, store ops, field teams — rarely have a direct channel into the number that gets baked into a plan.

That's the silo problem showing up in forecasting specifically: marketing knows about a promotion before supply chain does, sales hears about a customer's pullback before it hits the pipeline, and by the time the forecast catches up, the inventory decision has already been made on stale information. Closing that gap isn't just about better statistical methods, though those matter. It's about making sure the right signal reaches the right decision-maker while there's still time to act on it, and giving planners a way to apply judgment where the data alone can't decide.

The articles below cover forecasting methods, demand sensing, planning process design, and the organizational habits that separate teams whose forecasts hold up from teams that replan every week.