Why cost-per-part is the only tool KPI that matters

The metrics most shops actually track

Walk into almost any NZ precision shop and ask the buyer what number they review each month. You'll hear some combination of:

Tool spend (dollars per month).
Unit price on the top-20-by-volume SKUs.
Vendor fill rate or lead time.

All three are worth tracking. None of them is the KPI that should drive decisions.

The KPI that should drive decisions is cost per part produced. Unit price divided by average tool life. A single number, per SKU, that answers the question you're actually trying to answer: what did this tool cost me in the parts I produced with it?

Why the other metrics mislead

Tool spend moves with volume. A 15% increase in tool spend can mean your buyers slipped on brand choice or that the shop had a great quarter. Without normalising by output, the number can't tell you which.

Unit price invites the oldest procurement trap: picking the cheapest sticker. A $39 end-mill that breaks at 44 parts costs you $0.89/part. A $63 end-mill that lasts 110 parts costs you $0.57/part. The expensive tool is the cheaper tool; the unit-price review misses this every time.

Vendor fill rate is an operational hygiene metric, not a procurement one. A vendor who ships on time but whose tools break early is not actually a good vendor — they're a predictable vendor.

The objection every buyer raises

"We can't track cost per part because we don't measure tool life accurately."

Usually true. Also usually overstated. Almost every shop tracks three things already:

TAKE events (tool leaves the crib).
SCRAP events (tool is retired from service).
Production counts (job orders or part stamps, depending on the ERP).

A tool that's taken, produces parts across one or more jobs, and then gets scrapped has a measurable life. The answer isn't perfect — you have to bucket by alloy and machine class for the math to mean something — but it's directionally right from day one. And directionally right is a substantial improvement over whichever number is driving PO decisions in most shops today.

What changes when cost-per-part becomes the default

Three decisions move, in our experience of NZ shops that have made the shift:

Procurement consolidations happen faster. Instead of three SKUs for functionally the same end-mill (one per vendor, each running its own PO cycle), the scorecard shows which one wins and PO volume consolidates within a month.
Vendor conversations get sharper. "Your tool costs 18% more per part than your competitor's on 316 stainless" is a different conversation from "can you come down on unit price." Vendors respect the first question; they're prepared to push back on the second.
Engineering gets a say. When the metric is per part, the CAM engineer's choice of which brand to specify for a particular alloy starts mattering at a commercial level. Procurement and engineering converge on the same language.

The implicit argument

The broader point: the metric you optimise shapes the decisions you make. A shop that tracks tool spend optimises for invoice reduction. A shop that tracks unit price optimises for sticker. A shop that tracks cost-per-part optimises for the thing the owner actually cares about — which is how much each part cost to produce, tooling included.

That's why it's the only tool KPI that matters. The rest are inputs to the calculation, not outputs.

Closing

If your shop can't calculate cost-per-part today, the answer isn't to chase a better spreadsheet. It's to get the three inputs (TAKE, SCRAP, part count) into the same system. The scorecard falls out of the data. Book a discovery call; we'll show you what your own numbers do.