Cost per unit is not a number, it is a calculation
Almost every brand we talk to has a cost per unit in a spreadsheet. Almost none can say when it was last true.
Ask a founder what their product costs to make and you usually get a confident answer to four decimal places. Ask when that figure was last recalculated and the confidence goes.
It is not carelessness. The figure was correct when it was worked out. What changed underneath it was sugar, or a co-packer raising rates, or a run that yielded ninety-six per cent instead of ninety-nine and quietly spread the same cost across fewer sellable units.
That last one is worth sitting with. Yield loss never appears on an invoice. Nobody bills you for the three per cent that did not make it. The cost simply redistributes onto the units that survived, and the only way to see it is to compare what went in against what came out — which requires someone to have recorded both.
So we made cost per unit derived rather than stored. Ingredients from the recipe times current supplier prices; packaging and freight from your rate lists; tolling from your manufacturer’s own rate card at the band your run qualifies for. Change any input and everything downstream moves. There is no field to update because there is no field.
The test we held ourselves to was a real one: Little Rick’s actual numbers — an order totalling $69.35 and a cost of $0.1232 per can — reproduced to the cent from the recipe rather than typed in. If the arithmetic cannot survive contact with a real batch it is not worth building.