Ask two suppliers for a price on the same part and you'll get two numbers. Award the lower one and you've made a decision that feels obviously correct — and is, roughly half the time. The other half, the "cheaper" part arrives in a minimum order quantity you didn't want, on incoterms that add freight and duty you didn't quote, with a reject rate that costs you an afternoon of rework per batch. The sticker was lower. The part was more expensive.
That gap between price and cost has a name: total cost of ownership. And for mid-sized manufacturers buying thousands of components, closing it is one of the highest-return things procurement can do — because the savings aren't hiding in tough negotiation, they're hiding in comparing the right number.
Why unit price wins arguments and loses money
Unit price wins because it's the one figure everyone can see, on every quote, in the same box. It's easy to line up in a spreadsheet and easy to defend in a meeting. Every other cost — tooling, freight, quality, the cash tied up in a big MOQ — lives somewhere else, arrives later, and belongs to a different budget line. So it gets left out of the comparison, and the comparison quietly optimizes for the wrong thing.
Unit price is the tip. The decision is made underneath the waterline, where the costs are harder to see.
The five layers under the price
Tooling and set-up. One-time costs — tooling, NRE, first-article inspection — spread across the volume you actually run. A low piece price on a part you order in small batches can be dwarfed by amortized set-up.
Logistics and duties. This is where incoterms quietly decide the winner. EXW and DDP for the "same" price are not the same deal — one leaves freight, insurance and customs on your side of the ledger.
Quality cost. Rejects, rework, sorting, and the line time lost to a bad batch. A part that's 3% cheaper and 3% more likely to fail incoming inspection is not cheaper.
Inventory and financing. A large minimum order quantity looks like a better unit price and is really a decision to tie up cash and warehouse space. Carrying cost is real cost.
Supply risk. A slightly cheaper part from a single, distant, sole-source supplier carries a risk premium a spreadsheet never charges for — until the week it stops a line.
You need a consistent model, not a perfect one
The objection to TCO is always the same: it's too complex to do on every part. And a forensic, to-the-cent model would be. But that's not the goal. The goal is to score every quote on the same axes — landed price, quality cost, inventory impact, risk — so the comparison is fair even when the individual figures are estimates. A consistent rough model beats a precise wrong one, because "cheapest" finally points at the right supplier.
Where the hidden costs actually hide
In practice, the biggest distortions come from four places, and they're the same four that make quotes hard to compare at all: different units, different incoterms, different minimum order quantities, and different payment terms. Normalize those onto one basis and most of the TCO gap closes on its own — the "landed" price you compare is suddenly the price you'll actually pay. (That normalization is exactly what our piece on RFQ normalization is about.)
From TCO to a decision
The reason to do any of this isn't a tidier spreadsheet — it's a better award. When a tender is scored on true landed cost plus delivery reliability, quality and risk, the recommendation your team acts on reflects what the part will actually cost to own, not what it looked like on the quote. That's the whole idea of a decision engine: it does the normalization and the scoring in the background and hands your buyers a ranked answer, with the money attached, instead of a column of sticker prices.
Unit price will always be the easiest number to see. TCO is the one worth deciding on.