When to Net Demand Against Stock vs Build Fresh in EDGEBIC

An order nets against stock when three gates all pass: the product is stocked, the order is consuming demand rather than a build-to-stock producer, and on-hand covers at least part of the need; otherwise it builds fresh. EDGEBIC by User Solutions checks these gates at scheduling time, so the decision reflects the real stock position when the schedule runs, not a guess made when the order was entered. Getting the product settings right is what makes the decision automatic and correct.

This post lays out the gates and the settings behind them so you can decide per product with confidence. For the step-by-step mechanics of the netting itself, see how EDGEBIC nets demand against stock; both sit under the EDGEBIC planning guide.

The three gates, in order

The engine asks three questions before it will net a demand against stock. All three must pass.

Is the product stocked? A product flagged not stocked skips all netting, receipt, and issue posting. Service items, overhead SKUs, and anything you never hold in finished-goods inventory belong here. If a product is not stocked, its orders always build; there is no stock to net against.

Is this order a consumer, not a producer? A build-to-stock order is a producer: it exists to add stock, so it never consumes. Only make-to-order demand and consuming orders are eligible to net. This keeps a run's supply from canceling its own demand.

Does on-hand cover the need? If available stock is at least the demand, the order can be satisfied from stock in full. If it covers only part, the engine builds the shortfall and consumes the rest. If it covers nothing, the order builds in full.

A fourth guard runs quietly: an order with actuals already logged is never retroactively satisfied from stock, because netting runs only after completed steps are preserved. In-flight work is protected.

The product settings that drive the gates

Two settings do most of the work.

Build method sets the product default: make-to-order (build only for a customer order, ship direct, never touch inventory) or make-to-stock (build ahead, output goes to stock, demand consumes it). A third option builds to stock by default but lets individual orders override to make-to-order.

Is-stocked turns netting on or off entirely. A make-to-stock product must be flagged stocked, or it silently skips posting; the anomaly report flags that exact mismatch.

Any single order can override the product default in either direction. A normally stocked part can run one direct-ship order; a normally custom part can run one build-ahead batch. The override lets you handle exceptions without reclassifying the product. The make-to-stock versus make-to-order decision covers the classification itself in depth.

A worked decision: three orders, one casting

Take a casting, make-to-stock, flagged stocked, with 70 on hand. Three demands arrive.

Order A: customer order for 50. Stocked? Yes. Consumer? Yes (make-to-order demand). Covered? 70 covers 50. Decision: net and consume. The engine posts an issue of −50, marks the order satisfied from stock, spends zero machine time, and leaves 20 on hand.

Order B: customer order for 100, scheduled after A. Stocked? Yes. Consumer? Yes. Covered? Only 20 remain after A consumed 50. This is partial cover. Decision: build the shortfall of 80, consume 20. The order occupies capacity for the 80 it produces, and posts an issue of −20.

Order C: a build-to-stock replenishment for 200. Stocked? Yes. Consumer? No, it is a producer. Decision: build in full. It never nets; on completion it posts a receipt of +200.

Notice the on-hand each order sees changes as earlier orders consume it. That is the reservation map at work, decrementing stock the moment each consumer nets so two orders never claim the same units. The inventory reservation map explained covers that mechanism.

Yield: netting still respects scrap

When the engine builds a shortfall, yield rides on top. A product with a first-pass yield below 100% inflates its build quantity so the net good output covers the need. Order B's 80-unit shortfall on a casting with 90% yield means the engine plans to start about 89 units to reliably yield 80 good ones. Netting decides how much to build; yield decides how much to start.

When to prefer consume, and when to prefer build

The gates make the routine decision automatic, but the policy behind them is yours to set.

Prefer consuming from stock for your runners: high-volume, standard parts you always hold. Classify them make-to-stock with a reorder trigger, and let demand draw down stock while replenishment refills it. Capacity goes to the orders that genuinely need building, and shelf stock does its job.

Prefer building fresh for custom or low-volume work: parts you would never hold, engineered-to-order items, anything where finished stock would just age. Classify them make-to-order so they never net and never post a receipt.

The overrides handle the edges: a rush direct-ship order on a stocked part, or a one-off build-ahead batch on a custom part ahead of a known spike. A handful of correctly classified A-items beats a full catalog of guesses, the same logic behind ABC analysis for inventory and disciplined inventory management for manufacturers.

Why the decision waits until scheduling

Deciding at scheduling time, not order entry, is deliberate. On-hand moves constantly as builds complete and orders consume. An order entered when stock was zero might, by the time it schedules, sit behind a completed replenishment that covers it. By evaluating the gates when the schedule runs, EDGEBIC uses the real position rather than a stale one, and the same order can consume this week and build next week if stock changed underneath it. That is also why rescheduling reverses the prior consume first and re-nets on restored stock, a point covered in how a reschedule keeps inventory consistent.

See the netting decision run against your parts in the EDGEBIC platform overview, or contact US for a demo.