What a Late Specification Change Costs, by the Stage It Arrives At

How to read this sheet

The number to argue about is recovery, and it is the one nobody writes down. It is not the accounting value of the work in progress; it is what a third party would pay for the goods in the state they are in, which for cut panels in one style and one size ratio is very close to nothing. Enter it honestly and the staircase appears by itself - and with it the single operation, usually cutting, that costs almost nothing to perform and takes almost all of the flexibility out of the order. That is the fact worth taking to a buyer, because it converts an argument about whether changes are expensive into an agreement about a date after which they are. The capacity charge is the second thing people leave out and it is the one that decides whether a change is expensive or ruinous. Rework does not only cost what it costs; it occupies a line that had other orders for it. Charge it at contribution when the floor is full and at nothing when it is not, and be honest about which is true - the same change on the same order is a very different number in March and in September. Read the whole column rather than the single figure. The table is a tariff, and a tariff agreed at order confirmation is worth more than any recovery negotiated afterwards, because at that point the buyer is choosing when to decide rather than whether to pay. What is not here. A change that reduces work rather than adding it - a simplified trim, a dropped operation - is outside the model, which only ever adds. Nothing is netted off for material that finds its way into another order later, which happens and is worth watching for. Partial changes across a size or colour break are not modelled; run the affected part as its own order. And the relationship cost of refusing a change, or of charging for one, is real and is not on this sheet: the arithmetic tells you what it costs, not what it is wise to invoice.