The pound is a unit, not a finished product
In extrusion sourcing, the recurring mistake is not simply paying too much for aluminum. It is comparing two numbers that describe different products.
One supplier’s $1.58 per pound may represent mill-finish material leaving the factory. Another supplier’s $2.85 per pound may already include die amortization, anodizing, cutting, packaging, inspection, and delivery. Treating those figures as competing prices can turn a seemingly cheap bid into an expensive purchase order.
A pound measures weight. It does not tell you whether the material is finished, accepted, delivered, or ready for assembly.
The weakness becomes clearer when profile weights differ. Imagine two profiles that perform a similar structural function:
- Profile A weighs 1 pound per foot and costs $3.00 per pound, or $3.00 per foot.
- Profile B weighs 1.8 pounds per foot and costs $2.20 per pound, or $3.96 per foot.
Profile B has the lower per-pound price but costs 32% more per foot. That does not automatically make it a bad design; the additional metal may provide stiffness or simplify assembly. It does prove that per-pound pricing cannot be the only comparison when the real purchase is a usable profile, finished part, frame, or assembly.
A published range such as $1.35–$3.65 per pound can help establish an initial market expectation, but it combines mill-finish, custom, finished, and differently delivered products. It is a screening range, not a reliable project budget.
Replace the headline with a cost stack
A useful extrusion cost framework treats the quoted number as one layer in a complete cost stack:
Effective cost per accepted pound = (material + conversion + tooling + finishing + fabrication + packaging + freight + duties + inspection + expected rework + carrying cost − recoverable scrap value) ÷ accepted pounds delivered
The calculation does not require a supplier to reveal its internal margins. It gives the buyer a consistent way to normalize different quotes.
Each component answers a different question:
- Material: Which alloy and temper are included? Is the price tied to an aluminum index, a regional premium, or a fixed number?
- Conversion: Does the quote include press time, setup, energy, labor, and the minimum production run?
- Tooling: Is the die included, billed separately, amortized over a forecast, or refundable after a purchase threshold?
- Finishing: Are anodizing, powder coating, color matching, masking, protective film, and finish testing included?
- Fabrication: Are cutting, drilling, CNC machining, bending, deburring, and assembly included?
- Logistics: Does the price cover packaging, inland transportation, ocean freight, brokerage, duties, tariffs, and delivery to the plant?
- Quality: Who pays for inspection, documentation, rejected parts, replacement material, and rework?
- Denominator: Is the price based on billet input, extruded output, shipped weight, finished weight, or accepted usable material?
The denominator deserves as much attention as the numerator. If one supplier charges per pound shipped and another charges per pound accepted after finishing and inspection, their prices are not comparable even when every other line appears identical.
Fixed charges should also be separated from variable charges. A die, first-article inspection, color setup, or special fixture is paid once or once per production campaign. A commodity surcharge or anodizing charge generally scales with weight. Mixing those categories hides how the price will behave when order volume changes.
A $1.58 quote that loses at the dock
Consider an illustrative first order of 2,000 pounds for a custom 6063-T5 profile. The requirement is clear anodizing, cut-to-length service, protective packaging, and delivery to a U.S. plant. The figures below are examples for comparing quote structure, not a universal market rate.
Supplier A provides a partially itemized domestic quote:
- Mill-finish extrusion: $1.85 × 2,000 pounds = $3,700
- Clear anodizing: $0.60 × 2,000 pounds = $1,200
- Cutting and packaging: $0.22 × 2,000 pounds = $440
- Domestic freight: $0.18 × 2,000 pounds = $360
- Custom die: $1,500
- First-article inspection and documentation: $200
The total is $7,400, or $3.70 per pound for the first order.
Supplier B advertises a lower headline price:
- Ex-works extrusion: $1.58 × 2,000 pounds = $3,160
- Clear anodizing: $0.68 × 2,000 pounds = $1,360
- Cutting and packaging: $0.28 × 2,000 pounds = $560
- Freight, brokerage, and duties: $0.75 × 2,000 pounds = $1,500
- Custom die: $2,400
- Inspection and receiving allowance: $200
The total is $9,180, or $4.59 per pound before additional inventory carrying cost or exceptional rework.
Supplier B’s $1.58 headline number is not necessarily dishonest. It may accurately describe the factory-gate extrusion charge. The problem appears when that number is compared with Supplier A’s delivered, finished price as though both cover the same scope.
The reverse can also happen. A supplier may quote $3.20 per pound with tooling, finish, and delivery included, while another quotes $2.70 per pound and lists those items as pending. The higher headline price may be the lower purchase cost once the scope is equalized.
Tooling turns volume into a pricing variable
Custom dies are a classic reason that the first order and the repeat order have different economics. Suppose all recurring costs through delivery total $2.95 per pound and the die, setup, and testing package costs $1,500. The effective price changes with the quantity actually purchased:
- At 500 pounds: $2.95 + ($1,500 ÷ 500) = $5.95 per pound
- At 2,000 pounds: $2.95 + ($1,500 ÷ 2,000) = $3.70 per pound
- At 10,000 pounds: $2.95 + ($1,500 ÷ 10,000) = $3.10 per pound
The die did not become cheaper. Its cost was spread across more accepted material.
This is why an annual forecast is not the same as a committed volume. If a supplier amortizes a $1,500 die over 10,000 pounds but the buyer takes only 2,000 pounds, someone must absorb the missing $1,200 of expected recovery. The commercial agreement should state whether the buyer pays the difference, loses a tooling credit, or receives a higher price on subsequent orders.
Custom tooling can still be the economical choice when it removes other costs. Assume an existing profile costs $2.60 per pound and requires $0.45 per pound of machining and assembly. Its effective system cost is $3.05 per pound. A custom profile costs $2.35 per pound after extrusion but requires a $1,800 die.
The custom option reaches break-even at approximately 2,571 pounds:
$1,800 ÷ ($3.05 − $2.35) = 2,571 pounds
At 2,000 pounds, the custom route costs $3.25 per pound after tooling. At 5,000 pounds, it costs $2.71 per pound. The correct comparison is not standard profile versus custom profile in isolation. It is the total cost of the finished system each option produces.
Tooling questions should be answered in writing:
- Is the die owned by the buyer or the extruder?
- Is the tool charge paid upfront or included in the pound price?
- Is a credit available after a specified purchase volume?
- Who pays for die repair, replacement, and storage?
- Does the price assume a minimum order for every future run?
- What happens if the forecast volume is not reached?
The denominator is often wrong
A quote can look favorable because it uses invoiced pounds rather than accepted pounds. Suppose the all-in cost of an order is $7,400 and the invoice covers 2,000 pounds. The apparent price is $3.70 per pound.
If 5% of the material is rejected, damaged, or unusable after cutting, only 1,900 pounds are available. The effective cost becomes:
$7,400 ÷ 1,900 = $3.89 per accepted pound
At a 10% loss, the same order costs:
$7,400 ÷ 1,800 = $4.11 per accepted pound
The supplier may replace defective pieces at no charge, which changes the cash calculation but not necessarily the cost of delay, inspection, line stoppage, or expedited freight. A scrap credit also changes the numerator. The important point is to define whether yield, rejects, and recoverable scrap are part of the comparison.
Fabrication waste can be just as important. Twelve-foot stock may produce a required cut pattern with 94% yield, while a slightly different stock length may produce 98% yield. A lower extrusion price can be erased by poor nesting and offcut disposal. The buyer should compare the cost of usable cut pieces, not just the weight of stock purchased.
Finishing and machining expose the limits of per-pound pricing
Surface treatment frequently includes fixed batch charges that disappear inside a large order but dominate a small one. If an anodizer applies a $350 minimum charge, that minimum adds $1.75 per pound to a 200-pound batch but only $0.175 per pound to a 2,000-pound batch.
Machining creates a similar distortion. A drilling operation is often priced per hole, setup, fixture, or part rather than by weight. A lightweight profile with eight precise holes can cost more to finish than a heavier profile with no secondary operations, even if its extrusion price per pound is lower.
A normalized quote should therefore show both weight-based and event-based charges:
- Price per pound for extrusion and continuous finishing
- Price per piece for cutting, drilling, and inspection
- Setup or fixture charges per production run
- Minimum batch charges for anodizing or coating
- Tooling and testing charges per new profile
This separation reveals where a supplier has a genuine process advantage. It also prevents a buyer from accepting a low per-pound number that leaves labor-intensive operations undefined.
Delivery time has a measurable cost
Landed cost is not limited to the invoice. Lead time affects inventory, cash flow, production scheduling, and the cost of responding to defects.
If a $100,000 order remains in the supply chain for six weeks and the company uses an 18% annual inventory carrying rate, the approximate carrying cost is:
$100,000 × 18% × (6 ÷ 52) = $2,077
That amount is not automatically a supplier surcharge, but it belongs in the sourcing decision. A five-percent factory price advantage can disappear when a longer pipeline requires more inventory or delays a customer shipment.
Quality risk can be modeled in the same way. If there is a 4% chance of a $3,000 rework event, the expected cost of that risk is $120 before considering downtime or customer impact. A pre-shipment inspection fee may cost less than the expected consequence of discovering a dimensional or finish problem after the material reaches the plant.
Imported material also requires a precise delivery definition. Ex-works, port delivery, and delivered-duty-paid pricing assign transportation, customs, and duty responsibilities differently. A quote that does not identify the delivery point and commercial term has not yet stated its true price.
A practical quote-normalization process
A disciplined comparison can be completed with a short set of decisions:
- Freeze the technical scope. Use the same drawing revision, alloy, temper, profile dimensions, length tolerance, finish, packaging, and inspection requirements for every supplier.
- Define the price basis. Ask whether the pound is billet input, extruded output, finished material, shipped material, or accepted material.
- Mark every line as included, excluded, or unknown. An unknown cost should be carried as a range rather than treated as zero.
- Separate fixed and variable charges. Keep dies, setups, fixtures, testing, and minimums separate from recurring per-pound costs.
- Move every quote to the same delivery point. Add freight, brokerage, duties, insurance, and local transportation where necessary.
- Calculate accepted yield. Include cut scrap, rejected pieces, finish damage, and any recoverable scrap credit.
- Convert the result to the real buying unit. Compare cost per accepted part, foot, frame, or assembly in addition to cost per pound.
- Stress-test the assumptions. Recalculate at the minimum order, expected annual volume, a lower-than-forecast volume, and a plausible freight or commodity-price change.
The resulting model does not need to predict every future expense. It needs to make assumptions visible. A supplier that cannot identify what is included in its pound price has supplied a starting point for discussion, not a comparable bid.
The procurement rule that prevents expensive surprises
A higher per-pound quote can be the better deal when it includes the die, finish, inspection, packaging, and delivery that a lower quote leaves open. A custom profile can be cheaper than a catalog shape when it eliminates machining and assembly. An imported price can be attractive at the factory and uneconomical at the receiving dock.
The reliable comparison is therefore not the lowest unexplained number per pound. It is the cost of an accepted, production-ready part delivered to the place where it will be used.
Compare suppliers only after the scope, delivery point, volume assumption, and accepted quantity are identical.
If a quote cannot answer what the pound includes, where risk sits, and how the price changes when actual volume differs from the forecast, it is not yet a price that procurement can safely compare.