Cutting Accessory Costs: What a 6-Year Spend Audit Actually Showed Me

The Problem You Think You Have

Ask anyone on the floor what's wrong with cutting accessories and you'll get the same two answers: they cost too much, and they don't last.

That was my assumption too. In Q1 2025 I pulled six years of tooling and consumable spend out of our purchasing system. Cutting accessories — circular saw blades, drill bits, hole saws, grinding discs — totaled $41,300 for 2025. Up 19% from 2023, on essentially flat production volume.

So I did the obvious thing. Went back to our three main suppliers, pushed on price, spent six weeks on it. Annual savings: $1,900.

Then I kept pulling data, because $1,900 against a $41,300 line item felt like a rounding error. What I found made the renegotiation look pointless. The money wasn't in what we paid for the blades. It was in everything that happened after the truck showed up.

What's Actually Going Wrong

Three things. None of them show up on a price sheet.

1. Nobody owns the specification

Our purchasing records list "blade, 12 inch" for four different applications. Framing lumber. Prefinished plywood. Melamine. Trim.

Those aren't the same cut. Not close.

A 12" x 80-tooth fine finish blade — like the Diablo D1280X 12" x 80 fine finish saw blade — exists for a specific reason: a high tooth count and a fine-finish grind geometry that leaves a clean edge in material where tear-out means scrap. Run that blade on framing lumber and you get a slow cut and a burned edge. Run a 24-tooth ripping blade on prefinished plywood and you get chip-out on the show face.

We were doing both. Not because anyone was careless — because the pick list said "blade, 12 inch" and that's what came off the shelf.

When I compared our Q1 and Q2 blade consumption side by side — same supplier, same volume, wildly different invoice totals — I finally understood why that number moved so much. Q2 had 14 finish cuts on prefinished material. Q1 had four. Same spec going out the door either way.

2. The question nobody asks about hole saw arbors

Here's an outsider blind spot I fell into myself.

Most buyers focus on the hole saw cup and completely miss the arbor.

So, quickly: what is a hole saw arbor? It's the mandrel that connects the cup to your drill chuck — the threaded shaft, the hex shank, and the pilot bit that runs through the center. The cup is just teeth and a shell. The arbor is what makes it turn true.

Why it matters: arbors aren't universal. Thread sizes differ (1/2"-20 and 5/8"-18 are the common ones), shank types differ (3/8" hex, 7/16" hex, SDS-Plus), and the pilot bit length has to match your depth of cut. Buy a cup without checking the arbor and you've bought a paperweight — or you buy a second arbor you already owned twice over.

We owned six arbors. Three were purchased in 2024 because nobody knew we had them. That's $340 in duplicate buys, which isn't huge. But it's the visible tip of a habit: we buy the consumable and think about the interface later.

3. The $40 items that decide whether the $90 items work

Our blade changes were being done with whatever wrench was closest. Usually an adjustable. Sometimes a box end that had been rounded out on something else two years ago.

Two problems with that. An adjustable wrench on an arbor nut rounds the corners, and a rounded arbor nut means the next guy over-tightens it to compensate. And arbor screws and router bit set screws are inch-pound fasteners — not "tighten until it stops" fasteners. Put a torque screwdriver set at the blade change station and you stop warping flanges.

A warped flange on a blade that's supposed to hold tolerance is a blade that cuts out of tolerance. It doesn't announce itself. It shows up three operations later as a piece that doesn't fit.

So: one torque screwdriver set, one box end wrench set that actually fits the arbor nuts instead of the adjustable, both living next to the blades. Under $200 combined.

What It Costs When You Don't Fix It

Let me put real numbers on this, because "we should be more careful" isn't a budget argument.

Across 2025, I traced three cost buckets back to specification and setup problems rather than blade quality.

Rework. Eleven door panels re-cut after chip-out on prefinished faces. Average 3.1 hours per panel between disassembly, re-cut, and refinish. At our loaded labor rate, roughly $9,900.

Changeover downtime. We logged 214 blade changes in 2025. A change takes about 4 minutes when the right blade and the right wrench are there — about 11 when someone has to go find them. The delta on the ones that ran long: around $4,100 in three-person crews standing around.

The cheap-blade loop. I only saw this after I started tagging purchases by job. On two product lines we were buying a lower-cost general-purpose blade and replacing it roughly twice as often as the carbide-tipped blade we'd used the year before. Lower unit price. Cost per cut wasn't even close.

Also worth saying out loud: if a supplier tells you their blade lasts "5x longer," ask for the test data. FTC advertising guidelines require that claims be truthful, not misleading, and substantiated with evidence — see the FTC's Business Guidance on Advertising at ftc.gov. That applies to the vendor you're buying from and, frankly, to the vendor you are.

Total recoverable: somewhere north of $14,000 a year. Against $1,900 from renegotiating price.

I'm not going to pretend price doesn't matter. It means price is the smallest lever in the room.

The Fix Is Boring

I'll keep this short, because the problem is the interesting part.

Give accessories their own cost center. Once cutting tools stopped hiding inside "shop supplies," we could tag spend by job. That single change is what surfaced the cheap-blade loop.

One spec line per material. Material, blade spec, RPM, feed rate. One page, taped to the saw. Not a manual — a page.

Standardize the arbor system before you buy cups. Pick one thread size and one shank type for the shop and buy to it. When you're evaluating a lineup like Diablo tools, the accessory ecosystem matters as much as the cup — arbors, pilot bits, and adapters have to interlock across the range, not just within one SKU.

Torque the fasteners, and keep the right wrenches next to the blade storage. Not in the tool crib. Next to the blades.

Run a test cut on scrap before the production run. This is the whole thing, honestly. Thirty seconds on an offcut of the actual material tells you whether the spec is right. 5 minutes of verification beats 5 days of correction. Every time.

Put a number on it. Before you change anything, count rework hours for one quarter. Once you can see it, the argument makes itself — and it's a much easier conversation than asking for another price cut.

Where My Experience Stops

My numbers come from one shop: about 60 people, millwork and light fabrication, roughly $340,000 in annual MRO and consumable spend. Six years of records, mostly domestic suppliers.

If you're running high-volume metal fabrication, or a service operation where blades live in trucks instead of on a saw bench, the dollar figures won't transfer. The structure probably will — spec ownership, arbor discipline, fastener torque, test cuts. But I can't tell you your split will look like mine.

And one caveat on everything above: I got the renegotiation completely backwards at first. Chased price for six weeks, found the real money by accident in a spreadsheet I'd built for something else entirely.

Should've started with the counts.