CPQ Software for Metal Fabrication

Metal fabrication quoting is judgement work performed under time pressure. Weight, cutting time, bend count, weld length and finish all matter, but the enquiry usually arrives as a sketch or a phone call and the answer is wanted the same day.

That pressure pushes estimators towards round numbers and safety margins. Some jobs are won at a price nobody can make, others are lost by a margin that was added out of caution, and neither outcome is visible until the job is either finished or gone.

Why every enquiry is a fresh calculation

A fabricator has no price for a bracket, a hopper or a balustrade. It has a price for this one, at these dimensions, in this grade, at this gauge, with this finish, for this customer. The price is computed rather than looked up, and that single fact is what separates made-to-order fabrication from every other business that quoting software is usually built for.

Configure-price-quote for fabrication therefore means holding the arithmetic rather than a catalogue. Dimensions and a grade go in; your own rules about weight, allowances, rates, process time and margin turn them into a number. There is no list of sizes to maintain, because any size within the allowed range prices itself.

What actually drives the price of sheet and section work

Material weight is the largest term and the one most often estimated. Grade sets the rate per kilogram; gauge sets how many kilograms. They multiply, and they are routinely confused. At 16 SWG (1.6 mm) a square inch of stainless sheet weighs about 0.008155 kg; at 18 SWG (1.2 mm) it is about 0.006116, a quarter less metal for the same panel area.

Grade moves the rate a long way. On one identical work table priced through the same rules with only the steel swapped underneath — SS 202 at ₹168/kg, SS 304 at ₹286/kg, SS 316 at ₹449/kg — the finished price ranged from ₹12,622 to ₹23,977 with no dimension changed. That 90 per cent spread across three grades is why a mental adjustment applied to a 304 price is not a good way to answer a 202 enquiry.

Around the metal sit the terms that are easier to leave out than to price:

• Offcut and yield allowance, because a nested part does not consume its own outline
• Bend allowance, so the blank is larger than the finished face and the price knows it
• Cutting time by process — shear, plasma, laser — and the kerf each one takes
• Weld length by process, not weld count, plus dressing and grinding after it
• Finish, which is a real cost: a hairline or No. 4 polish, pickling and passivation after welding
• Bought-in fittings, and freight and installation where they belong to you rather than the client

How this is normally run today, and what it costs

Almost always a spreadsheet, inherited and adapted. It works, and then it stops being true. Rates were updated in one copy and not another, a wastage percentage was nudged for a difficult job and never nudged back, and rows were added by people who no longer work there. Two estimators pricing a comparable job arrive at different numbers for reasons nobody can reconstruct afterwards.

The specific failure worth naming is that the workbook keeps the total and throws away the working. Once only the number survives, nothing downstream can be derived from it, and every question a customer or an owner asks later — why is this line that price, what did we allow for wastage, what margin did we actually take — becomes an archaeology exercise rather than a lookup.

Margin arithmetic quietly compounds the problem. Margin on selling and markup on cost are different calculations and are used interchangeably in conversation, so a target that is stated as a percentage can be applied two ways and produce two prices. The way out is not a convention but a mechanism: run the same rules twice on every line, once on cost rates and once on selling rates, and report the gap. Then margin is a figure per line rather than an assumption about the job, and discounting becomes a decision instead of an accident.

Why generic quoting software does not fit this

Most quoting software assumes a catalogue: products have prices, quoting is picking products and applying a discount. That is correct for a distributor or a reseller and useless for a shop where no two jobs share dimensions.

Design-led tools sit at the other extreme. You model the job in 2D or 3D and the price falls out of the geometry. Excellent where the drawing has to happen anyway, expensive where it does not, because it makes drawing a prerequisite for quoting and most enquiries never become orders.

There is a test that separates the three in a demo. Ask for a price at 72 by 30, then at 84 by 30, then at 84 by 36. If the price moves proportionately to material, the arithmetic is real. If it moves in steps, somebody has to create every new size and you are looking at a catalogue in a costume. Then ask them to change the grade and explain how the number moved. Grade sometimes changes the rate and sometimes changes the calculation itself, because in a heavier grade some items are genuinely fabricated differently, and a pricing engine that only supports the first cannot express what the business already does.

What the quotation should hand the shop floor

An accepted quotation already contains everything the shop needs to buy and cut, because it had to know all of it to produce the price: which material each part is in, at what grade and gauge, how much of it in the unit it is purchased in, and the blank size including the bend allowance.

Two documents fall out of that without anybody doing the arithmetic a second time. The bill of materials is the buy list, aggregated across every line — so much 16 SWG 304 sheet, so much 50 mm square tube, twenty-four feet, six fittings. The cutting list is the flat pieces at blank size before anything is folded, which is what the person at the shear actually works from.

The cutting list is also where nesting stops being guesswork. Stainless arrives in standard sheets — 1220 × 2440 mm and 1220 × 3048 mm are the usual two — and the question of how many sheets a job consumes is a packing problem, not a division. Laying the pieces out with rotation allowed, choosing the smallest standard sheet that holds the largest remaining piece, and reporting utilisation and offcut area per sheet turns the wastage allowance from a percentage somebody guessed into a figure you can check. Any piece larger than a standard sheet needs flagging rather than silently splitting, because that is a fabrication decision and not an arithmetic one.

When those are rebuilt by hand instead, they drift. The estimator allowed two inches all round for the fold; the person planning the cut allows an inch and a half because that is what he has always done. Neither is wrong, and across forty lines the material variance becomes a mystery that gets explained away as steel being expensive that month. Storing the working rather than only the total is what removes that whole category of variance.

Contract work adds a traceability layer that quoting has to be aware of even when it does not produce it. Structural and pressure jobs are bought against a mill certificate tied to a heat number, and inspection is done against a weld map identifying each joint, its process and its welder. If the grade on the quotation and the grade on the certificate are allowed to disagree, the discovery happens at inspection rather than at estimation.

What happens when steel moves

Material rates need to be held centrally and dated, and the behaviour on a rate change is a fair test of whether a system understands the trade. New quotations should pick up the new rate immediately. Quotations already issued should keep the rate they were priced at, with any change appearing as a numbered revision rather than a silent edit, so what you sent in March is still what you sent in March when the argument happens in July.

The related question is who can change a rule. The answer you want is your own estimator, in the interface, after training. If the answer is that the vendor makes the change in the next release, your pricing moves at the vendor’s speed rather than the steel market’s.

When this is not the right fit

If one person quotes, the volume is low, and no two jobs share any arithmetic, a spreadsheet is still the right answer and the overhead of a system would be paid for nothing.

Nor is this the right tool if you mostly resell equipment other people manufacture. That genuinely is a catalogue business and catalogue software will serve it better. Shops that do both usually want both kinds of tool rather than one bent to cover the other.

The threshold is not really volume. It is more than one person quoting, or the same product going out at different sizes repeatedly. The first means the pricing assumptions need to live somewhere other than in a head. The second means arithmetic that has not changed is being retyped, and every retype is a chance to be wrong.