Demand-Driven MRP (DDMRP): How Buffers Replace Forecast-Pushed Planning

Key takeaways

• Demand-driven MRP (DDMRP) replaces forecast-driven planning with strategically placed stock buffers that absorb variability.
• It positions decoupling points in the flow, then sizes and manages buffers using red, yellow, and green zones.
• Planning is driven by actual demand pulling through buffers, not by a forecast pushed through the whole BOM.
• DDMRP aims to cut both stockouts and excess inventory by reacting to real consumption.

Demand-driven MRP is a response to a familiar problem: classic MRP amplifies small forecast errors into big swings of shortage and excess. DDMRP breaks that chain by decoupling the flow with buffers and planning to actual demand.

The core idea: decoupling points

DDMRP places decoupling points at strategic spots in the bill of materials and the supply chain. At each one it holds a stock buffer that absorbs variability, so demand signals do not have to travel all the way up the chain and disturbance does not travel all the way down.

Between decoupling points, lead times shrink because each segment only has to replenish to its own buffer, not wait for the entire chain.

Buffer zones: red, yellow, green

Each buffer is sized into three zones. Green sets the reorder frequency and order size, yellow covers demand over the lead time, and red is the safety zone that protects against variability. The position of on-hand plus on-order stock within these zones drives the planning priority and the size of each replenishment.

A worked example

A component has a buffer of 300 units: 100 green, 150 yellow, 50 red. On-hand plus on-order falls to 180, landing in the yellow zone. DDMRP signals a replenishment back up to the top of green, ordering 120 units. There is no forecast in that decision, just actual position against the buffer.

If demand spikes and stock drops into red, the same item jumps to the top of the planner's priority list automatically.

DDMRP vs classic MRP

• Trigger: classic MRP pushes a forecast through the BOM; DDMRP pulls to replenish buffers from real demand.
• Variability: MRP passes it along and amplifies it; DDMRP absorbs it at decoupling points.
• Inventory: DDMRP targets the right buffer in the right place, rather than blanket safety stock.

Where OEE fits

Buffer sizes depend on the replenishment lead time, and lead time depends on real capacity. If OEE on a feeding resource falls, actual lead times stretch and buffers sized for the old lead time run dry. Tracking real performance keeps buffer math honest. Book a Fabrico demo to see how live OEE data supports demand-driven planning. DDMRP also pairs with finite execution; see APS and finite-capacity scheduling.

Common mistakes

• Buffering everything. Decoupling points are strategic; a buffer at every level just hides inventory.
• Static buffer sizes. Zones must adjust as demand and lead times change, or they drift out of date.
• Sizing buffers on optimistic lead times. If real capacity is lower, buffers are too small to protect flow.

Frequently asked questions

Does DDMRP get rid of MRP entirely?

Not exactly. DDMRP keeps the idea of bill-of-materials explosion but changes what drives planning, replacing forecast push with buffer-based pull at decoupling points.

Is DDMRP the same as kanban?

They share a pull philosophy, but DDMRP adds dynamic buffer sizing, priority by buffer zone, and explicit decoupling-point strategy across multiple levels, where kanban is usually a simpler shop-floor signal.