The press does not lie. The costing system did.

Shared machines, unequal products.

A thin architectural profile and a thick industrial section pass through the same press, but they do not consume it equally: different die changeovers, different speeds, different scrap, different recovery.

Downstream, the inequality compounds. Some products go straight to packing. Others pass through anodising, with its own tanks, cycle times and energy profile, or through fabrication: cutting, punching, machining, assembly.

The client's costing spread these very different consumptions with broad averages. The commercial consequence was familiar: quotes confident to the third decimal, built on cost figures wrong in the first. Which finishes genuinely earned their price, and which product families were quietly subsidised, nobody could say.

Why model press, anodising and fabrication separately?

Because they are three different economic machines under one roof, each with its own capacity and its own cost of being available.

Averages had been taxing the simple and subsidising the complex.

The margin ranking changed materially. High-volume, low-touch profiles carried cost that belonged to short runs with heavy changeovers. Once true consumption was attributed, the ranking of the product families was re-drawn. CONFIRM: number of product families whose margin ranking changed

Anodising was not one price. Costed on surface area and cycle time rather than a flat per-kilo adder, some finish and thickness combinations were priced well below their consumption. CONFIRM: share of anodised volume found priced below full cost

Batch size was the hidden driver. The same profile could be comfortably profitable at production batch sizes and loss-making in small repeat orders, because die changes and set-up consumed press capacity the price never recovered. The model made the break-even batch size visible per profile. CONFIRM: example break-even batch figure approved by the client

The plant did not have expensive products. It had expensive combinations.

Quoted from consumption, not from averages.

A shared language between finance, production and sales.

The engagement gave the plant minutes and rates instead of allocations and arguments. The first pricing round run on the model delivered CONFIRM: client-approved outcome, e.g. margin uplift on repriced items.

When the scarce resource is press hours, profit per press hour is the number that runs the business.

Fair questions.

Does TDABC cope with scrap and recovery in extrusion?

Yes, and it should. Recovery enters the time equations and material equations directly, so a profile with poor recovery carries its own waste instead of hiding it in plant overhead.

We already have standard costing in the ERP. Why is this different?

Standard costs are usually built on plant-wide or department rates and updated annually. TDABC rates are built per process stage on practical capacity, and time equations cost each order's actual routing and batch size. The difference is largest exactly where quotes are won and lost: small batches and complex finishes.

How long did the model take to build?

Weeks, on data the plant already had. CONFIRM: actual duration The longest step is usually agreeing practical capacity per stage, which is a management conversation, not a data problem.

Can this extend to energy costing?

Naturally. Anodising and press energy are significant pools, and attributing them by cycle time and surface area is exactly what the time equations do. This also gives a defensible base for carbon-per-product reporting later.