Order Acceptance Under a Binding Constraint

How to read this sheet

The whole sheet depends on one claim: that a single resource is what runs out. Where two resources bind on different orders - stitching on one, finishing on another - the ranking here is not the answer and no single ranking is, because the problem stops being a ranking and becomes an allocation across two scarcities. The honest use is to identify the resource that binds first, rank against that, and re-run when it moves, which it does between seasons. Whole-order selection is a knapsack, and taking orders in rank order is not guaranteed to be the best possible set: an order skipped because it does not fit can leave hours nothing else fills. Idle hours are therefore reported for both rules rather than assumed away, and where they are large the remedy is usually to ask a buyer whether a delivery can be split rather than to accept the sheet's answer. The price at which a refused order would earn its hours is the rate of the worst order still accepted, applied to that order's minutes - it is a negotiating position, not a valuation, and an order can clear it and still be refused because it does not fit, which the sheet distinguishes because the two have different remedies. Fixed cost is absent deliberately: it does not change with the decision, and absorbing it per piece is what makes a fast cheap order look unprofitable. Nothing here values the customer relationship, the cost of refusing a buyer whose other programmes you keep, or the learning curve on a repeat style - and the first of those reverses real decisions that this arithmetic would settle the other way.