Forging Quotation Costing: Why It Always Ends Up in Excel (and What That Costs)
Forge quotations end up in Excel because forging costs do not behave like catalogue prices. Every quote reprices raw material against a moving metal index, amortises a die that has to be right the first time, and applies formulas that change with the quantity on each order line. Most ERP implementations model none of that, so the estimator builds the logic in a spreadsheet, and the ERP only hears about the deal after it is won. That gap has a measurable cost, and it is fixable.
We work with metalworking companies on Business Central, and we have walked the shop floor of a French forge to map exactly this process. This article describes what we found, why the Excel habit is rational, and what it takes to move costing into the ERP without breaking what works.
Why does every forge end up quoting in Excel?
The pattern is always the same. The customer sends part drawings and target quantities. The design office studies the drawing and estimates the cost: material, tooling, machining, subcontracting. Sales adds a margin and returns a price. The ERP, meanwhile, holds items, price lists and orders, and none of those exist yet at the moment of quotation.
That is the structural reason, and it is worth stating plainly: in a forge, the item being priced does not exist in the ERP when the quote is made. You are pricing a part that has never been produced, from a drawing, with a die that has not been cut. Standard ERP pricing starts from an item card; forge quotation starts from geometry and a cost model. When the system cannot express the cost model, the spreadsheet fills the void.
There is nothing irrational about it. The estimator gets a tool that follows their reasoning, formulas they can adjust, and total autonomy. The problems start after the quote is accepted.
What actually goes into a forging quotation?
In the forge we visited, every quotation decomposed into five cost blocks, and this structure generalises to most forging and stamping operations.
First, material. In non-ferrous forging this is the volatile block: the price of copper or brass is indexed on a market rate, typically the monthly average or the end-of-month price, sourced from the supplier terms. Two suppliers can index differently on the same metal. If the metal price moves, the selling price of the part moves with it, because the contract says so.
Second, tooling. The die is a fixed cost with a hard property: it must be right from the first cut, because there is no reworking a forging die. Its price is quoted once and stays fixed, and it amortises over the quantities actually ordered, which makes quantity assumptions part of the price itself.
Third, machining and internal operations, valued at machine-hour rates. Fourth, subcontracting, such as surface treatment, priced per operation. Fifth, packaging and transport, which forge customers routinely expect inside the part price rather than as a separate line.
On top of these, the formulas are quantity-dependent: cost per unit changes with the quantity per order line, because setup time, die amortisation and material purchasing conditions all scale differently. A price is not a number; it is a curve.
What does the Excel habit really cost?
The spreadsheet is good at producing the first number. Everything after that first number is where it fails, in four ways.
Re-keying. The quote lives in Excel; the order lives in the ERP. Someone re-types the accepted quote as a sales order, and re-types the cost assumptions nowhere at all. The link between the promise and the execution is broken on day one.
Repricing lag. When the metal index moves, every open price in the book is stale. Contracts allow the adjustment, but if the index logic lives in one workbook, repricing is a manual campaign: someone reopens each quote, recalculates, reissues. In the gap between the index moving and the prices moving, the forge eats the difference.
Key-person risk. The cost model usually lives in one person’s head and one workbook. The formulas encode years of shop-floor knowledge, and nothing else in the company can produce a price. Holidays, sickness or departure of that one person stops quotation.
No feedback loop. This is the expensive one. The Excel estimate is never compared to what the parts actually cost to produce, because actuals live in the ERP and the estimate does not. The company never learns which quotes lost money, so the same estimating errors repeat for years. An estimating error you cannot see is an estimating error you keep.
Can Business Central model forge costing?
Yes, with two honest caveats. Here is how the pieces map.
First, the licence. Manufacturing in Business Central, meaning production BOMs, routings, work centers, machine centers and production orders, requires the Premium licence. Essentials does not include it. If a partner quotes a forge on Essentials, the costing model discussed here is off the table.
Machine-hour rates map to Business Central’s capacity structure: machine centers represent the actual machines and operators, grouped into work centers. A work center can carry the averaged rate of its machine centers, which matches how forges already think, since many cost centers are effectively single-machine anyway. Overheads belong in the machine-hour rate too, and Business Central has dedicated fields for exactly that: indirect cost percentage and overhead rate on work and machine center cards.
Annual rate updates, the yearly ritual of revising standard rates, map to the Standard Cost Worksheet: it lets you prepare changes in advance, simulate their effect on manufactured item costs, then implement them on a chosen date with the inventory revaluation handled.
Customer forecast schedules map to blanket sales orders. Forge customers typically send a rolling forecast and then confirm quantities and dates over time. A blanket order holds the framework quantities without affecting item availability, and each confirmed delivery converts into a sales order, which is precisely the call-off pattern.
Dies map to lot-tracked items. Because a die is an individual physical asset, lot tracking fits, and it brings a discipline for free: Business Central blocks any outbound posting of a lot that is not in stock, so die inventory cannot go negative. Corrections happen as explicit positive adjustments with reason codes, visible and auditable.
The two caveats. Metal price indexation is not a native feature: Business Central has no built-in copper feed. The indexation rule, which average, which supplier basis, which adjustment threshold, is business logic you define, either as a disciplined manual parameter or a small extension. And the quotation calculator itself, from drawing to cost, is design-office reasoning; the ERP’s job is to hold the cost structure that the calculation feeds on, not to replace the engineer.
How do you move quoting out of Excel without breaking it?
Not by decree. The spreadsheet is the current specification of your cost model; killing it first destroys the specification. The order of operations matters.
Start by making the ERP agree with Excel: model the work centers, machine rates, overheads and BOM and routing templates until the ERP prices a known part the way the spreadsheet does. Until the two agree on a part you already produce, nothing should move.
Then move the flow, not the formulas: quote to order inside the system, so that a won quote becomes a sales order without re-keying, carrying its cost assumptions with it. Keep the metal index as a maintained parameter, updated in one place, never hard-coded in a cell.
Last, close the loop. Once orders are produced through the ERP, compare actual production costs, broken down by material, capacity and overhead, against the estimate for each order. That comparison is the payback of the whole exercise: it turns every produced order into a correction signal for the next quote. Excel could never give you that, because it never knew what happened after the quote.
FAQ
Does forge costing in Business Central require the Premium licence?
Yes. Production BOMs, routings, work centers, machine centers and production orders are Premium-only features. On Essentials you can sell and buy, but you cannot model production costing. Budget for Premium users in any forging scenario.
Can Business Central index selling prices on the copper price?
Not natively; there is no built-in market feed. The indexation rule is business logic you define, as a maintained pricing parameter or a small extension. The real win is discipline: one index value, maintained in one place, driving every price that depends on it.
How should forging dies be handled in the ERP?
As lot-tracked items with fixed prices. Lot tracking means the system refuses to post an outbound movement for a die that is not in stock, so die inventory stays honest, and corrections are explicit adjustments with reason codes rather than silent negative stock.
Do we lose our Excel formulas in the transition?
No, you translate them. The spreadsheet is the specification of your cost model: its formulas become machine rates, routing times, overhead percentages and quantity rules in the ERP. The estimator’s knowledge is preserved; what disappears is the disconnection between the estimate and reality.