I got called into a meeting once because I misread a number. It was 2023, and I was managing purchasing for a 180-person precision machining shop. I was looking at a spend line for cutting tools that seemed flat.
Four months later, the shop floor supervisor and I were in a conference room arguing with the operations director and the finance manager about why the same tooling was suddenly "expensive."
He said I was buying "cheap." I pointed out that my choice was the lowest of three quotes. Both of us were right. That's the whole problem.
The problem I thought I had
When I took over this role, my brief was simple: keep the shop tooled, and don't blow the budget.
Tooling was one line item in a bigger purchasing ledger. I tracked it by price per SKU. When I compared suppliers during my first few budget reviews, I was essentially comparing unit prices—putting two quotes side by side and circling the lower one.
If the cheaper brand arrived on time and nothing went drastically wrong, I considered it done.
To me, that was the job. Pick the lowest price. Confirm delivery. Move on. I didn't think harder about it because nobody asked me to.
The problem I actually had (and didn't see)
Here's the thing I didn't understand at first: with cutting tools, the unit price is the smallest part of the cost that actually shows up on a quote.
If you break down a tool's real cost in production, it includes:
- The purchase price of the tool itself
- How long it cuts before needing replacement
- How fast it cuts (feed and speed)
- The surface finish it leaves on the part
- How often it needs changing, and the machine downtime that comes with it
- The scrap rate caused by inconsistent wear
None of those show up on the quote. But all of them hit the same production budget.
I knew, even as a junior administrator, that our machines cost money per hour in terms of lost profit. That means every tool change is a bill nobody sends.
I started thinking about it in concrete terms. If a cheaper end mill adds 30 seconds to a cycle, and that machine runs 80-100 cycles of the same job in a day, that adds up fast. Not in a way that shows up tomorrow. But it shows up.
By 2024, when we started tracking downtime on critical machines, it became clear that unplanned tool changes—and part inspections caused by inconsistent tool wear—were eating a bigger slice of the shift than I wanted to admit. Nobody was logging that under "purchasing."
What it cost us
Two things finally forced me to look deeper.
First, a small thing, but useful. We had a 3-axis milling job producing a moderately complex pocket, and it was consistently running. After a switch to a cheaper tool, cycle time went up by around six minutes. The tooling team thought it was the program or the machine. It took us nearly three weeks to trace it back to the tool geometry and coating difference. In those three weeks we ran maybe 40 extra cycles per week—unplanned, unmeasured.
Second, worse. A tool failure mid-run scrapped a few parts. One of them had already been through a secondary operation. The scrap report went to production and then bounced back to finance, who by then had a list of questions for me.
Finance didn't ask "why did you choose this supplier?" They asked "why is the tooling line item 11% over budget?" I had done every individual purchase correctly on paper.
That was the moment I realized: I was doing each transaction efficiently and still contributing to the more expensive problem.
Anyway, that's when I opened the conversation about cost-per-part versus cost-per-tool.
What I was actually doing wrong
My mistake wasn't that I chose cheaper tools. It's that I treated tooling as a purchasing exercise instead of a machining decision.
When I treated it as a purchasing exercise, these things happened:
- I switched tool models when it was convenient (because another supplier was cheaper that quarter)
- I traded system consistency for unit price
- I never grouped orders to optimize a tooling chain
- I never asked my purchasing team for tool-life data
I was managing this by volume. That was the flaw.
The framework that actually helped
Total cost of ownership changed the angle. It wasn't about ignoring price—it was about looking at the whole picture:
- Cost per part, not cost per tool
- Downtime, expressed in real money (I didn't know the shop's true hourly rate until I asked—I'd recommend any admin buyer do the same)
- Tool change time, factored into the cycle budget
- Scrap rate, treated as a cost rather than an accident
- All components in a tooling cycle, added up
Each of those required input from the shop floor. I couldn't derive it alone. That's the reason tooling purchasing works best when an administrator who owns the purchasing process and a supervisor who owns the machining process sit down together.
What changed in practice (short version)
I went to the sandvik-coromant catalog and started trying to keep machining on one tooling system rather than chasing cheaper individual tools. It cost more on the quote—maybe 8-12%, though I'm not remembering exactly. But we got more consistent cycle times, fewer tool changes, and a much better handle on cycle time predictability. Within a month the shop stopped coming to me with tooling complaints.
In hindsight, matching tools to the machine rather than matching prices between suppliers was the bigger win. And it saved me from chasing the "5% cheaper" supplier, which was a losing game anyway.
It doesn't mean always buying the most expensive tool
I want to be honest about the limits here. My approach wouldn't make sense for every company.
We run steady, predictable high-mix machining where tool wear is a meaningful cost. If you're a job shop where one job runs one tool for a day and never repeats, the cheaper tool might work just fine. If your machines are older and the constraint is elsewhere, tool quality might not be your lever point. I've seen shops overspend on premium tooling they never measure—which is a different problem, not a better one.
I can only speak to what we have. If you're running a smaller or more irregular volume profile, the math probably looks different.
I also couldn't explain why some tool brands consistently outperform others in specific operations—stainless and high-temp alloys especially. My best guess is that it's about substrate, coating, and how well the geometry fits the specific operation, and that's something catalog specs alone won't tell you. You have to test and observe.
Bottom line
If you take away one thing from this, take this: in cutting tools, the easiest cost to calculate is the least important.
I shifted our purchasing process from a "find the lowest quote" administrative routine into a routine where we collect cost data and ask questions that weren't asked before: what's the cost per part, how long is the machine down, how often are we swapping tools.
That—not the lowest quote—is the value the purchasing role actually brings to the shop floor.