Key takeaways
Short answer: Dependent and independent demand are the two fundamental types of demand, and the difference determines whether you forecast it or calculate it. Independent demand comes from outside — external customers ordering finished products — and cannot be derived from anything else, so it must be forecast. Dependent demand is derived from the demand for another item: a component's demand depends on how many of its parent product you plan to make, so it can be calculated rather than forecast. Forecast the independent; calculate the dependent. This distinction is the bedrock of MRP. For the planning system it underpins, see ERP vs MRP.
Independent demand is demand that originates outside the operation and cannot be derived from the demand for any other item — it stands alone. The classic case is a finished product ordered by external customers: how many units customers will buy is driven by the market, not by any internal calculation, so it is independent. Because there is nothing to derive it from, independent demand must be forecast — you predict it from history, market signals, and judgment, accepting the uncertainty that comes with any forecast. Independent demand is inherently uncertain and continuous, and managing it is about forecasting accuracy and the buffers that protect against forecast error. It is the demand that enters the system from the customer side, the starting point from which everything downstream is calculated.
Dependent demand is demand that is derived from the demand for another item — its parent. A component, sub-assembly, or raw material is needed only because something it goes into is being made, so its demand depends directly on the production plan for that parent. If you plan to build 100 bicycles, you need 200 wheels — the wheel demand is not forecast, it is calculated from the bicycle plan and the bill of materials. This is the key property: dependent demand can be calculated precisely rather than forecast, because it follows deterministically from the parent's demand and the BOM. Dependent demand tends to be lumpy (it spikes when a parent is built and is zero otherwise) rather than continuous, and managing it is about calculation and timing, not forecasting — you compute exactly what is needed and when.
The clean distinction, and its single most important consequence, is this: independent demand must be forecast, dependent demand should be calculated. Because independent demand has no parent to derive it from, prediction is the only option, with all the uncertainty that implies. Because dependent demand follows deterministically from a parent's plan and the BOM, it can be computed exactly — and should be, because forecasting it independently would be both unnecessary and error-prone. This is the foundational insight behind material requirements planning: you do not forecast component demand item by item; you forecast (or take orders for) the finished products with independent demand, then calculate all the dependent component demand by exploding the BOM. Treating dependent demand as if it were independent — forecasting components separately — is a classic, costly error that MRP exists to eliminate.
A company makes bicycles. The bicycles themselves have independent demand — customers order them based on the market, so the company must forecast how many bicycles it will sell. Suppose the forecast and orders call for 100 bicycles next month. The components — wheels, frames, chains, pedals — have dependent demand: their requirement is not forecast but calculated from the bicycle plan and the bill of materials. 100 bicycles need 200 wheels, 100 frames, 100 chains, 200 pedals — computed exactly, not predicted. The company forecasts only the independent demand (bicycles) and calculates all the dependent demand (components) from it. If it had instead tried to forecast wheel demand separately, it would have introduced needless error and likely mismatched wheels to bicycles. Forecast the parent; calculate the children.
This distinction is the entire conceptual foundation of material requirements planning. MRP works precisely because dependent demand can be calculated: you start with the master production schedule for finished products (the independent demand, forecast or ordered), explode it through the bills of materials to compute the gross requirements for every component (the dependent demand), net against inventory, and time-phase the orders. The whole logic depends on recognizing that you should not forecast components — you should derive them. Before MRP, operations often did forecast component demand independently, holding excess of some parts and stocking out of others because the components were not tied to the actual parent production plan. MRP's breakthrough was to treat dependent demand as dependent — calculated from the parent — which is why understanding this distinction is essential to understanding how modern planning works.
The demand distinction connects to OEE through the production plan that dependent demand is calculated against. Dependent demand is derived from a master production schedule, and that schedule is only achievable if the floor can actually produce at the assumed rate — which depends on OEE. If a line runs well below its assumed capacity, the parent production falls short, and the dependent-demand calculations that drove component ordering no longer match reality, the same capacity gap behind finite scheduling. Components arrive for a build rate the line cannot hit, or the schedule slips and the careful dependent-demand timing is wasted. Grounding the master schedule in real OEE-adjusted capacity is what keeps the elegant calculate-don't-forecast logic of dependent demand actually working.
Fabrico supplies the real capacity that the production schedule — and therefore all the dependent-demand calculations built on it — must respect. By measuring actual availability, performance, and quality, it reveals whether the line can truly produce at the rate the master schedule assumes, so the parent production plan that drives component requirements is realistic rather than optimistic. When the schedule is grounded in real OEE, the dependent-demand calculations match what the floor actually does. Book a demo to base your production plan on real capacity.
Independent demand comes from external customers and cannot be derived from anything else, so it must be forecast. Dependent demand is derived from the demand for another item (its parent) and can be calculated. Finished products usually have independent demand; their components have dependent demand.
Because it follows deterministically from the parent item's production plan and the bill of materials. If you plan to build 100 products, you can calculate exactly how many of each component you need. Forecasting components separately would introduce needless error.
Finished products sold to external customers typically have independent demand — driven by the market and not derived from anything else, so it must be forecast. The components that go into them have dependent demand, calculated from the finished-product plan.
It is the foundation of MRP. MRP forecasts or takes orders for finished products (independent demand), then explodes the bills of materials to calculate component requirements (dependent demand). The whole logic rests on calculating dependent demand rather than forecasting it.
Yes. A component can have dependent demand for use in production and independent demand as a service or spare part sold separately. Such items need both the calculated production requirement and a forecast for the independent service demand.
Programați o întâlnire individuală cu experții noștri sau înscrieți-vă direct în planul nostru gratuit.
Nu este nevoie de card de credit!
Login
Blog
