← Matt Stone

07

Somebody is already buying the industrial base

While everyone watched the mines, a public company quietly rolled up the calibration labs and a private operator did five motor-repair deals in a year. Almost nobody in those industries noticed. I found out the hard way. By ranking both categories as wide open, and then checking.

August 2026

Every plan to bring critical minerals production back to America assumes something nobody says out loud: that the businesses around the plants already exist. Somebody has to run the samples. Calibrate the instruments. Machine the parts. Rewind the motors. Haul the hazardous material and file the permits.

Those firms are real, they are mostly owned by one person, and almost nobody writes about them. I spent two months going through twenty of those categories, because I want to buy one.

What I found is more interesting than what I was looking for. Two of the three categories I liked most had already been consolidated. And in both cases the buyer was hiding in plain sight.

What I was doing

Scoring twenty categories on four questions, to find one worth putting money into:

Two of my top three came off the list on the third question. Here's how.

Finding one: Transcat has been buying the calibration industry

The logic looked strong. Every fab, lab and precision shop needs instruments calibrated on a schedule. It's recurring, mandated by quality standards, route-based, and sticky. Searching for a private equity platform in it turned up nothing.

That last part was the error. No PE platform is not the same as no consolidator.

Transcat is a public company, and it has been rolling this up in the open.

A strategic paying 3.8x revenue with a nine-figure credit facility is not white space. It's an auction I'd lose.

I searched for the wrong kind of buyer and concluded nobody was buying.

Finding two: IPS is doing the same thing in motor repair

Same reasoning, same mistake. Motors run everything in a processing facility. Rewind is skilled, local, and recurring. No PE platform surfaced.

Integrated Power Services, out of Greenville, South Carolina, has roughly 2,200 technicians and has completed at least five acquisitions since April 2025:

Five deals in under a year, from an operator with a national technician base. Not white space either.

Why I missed both, and the check that would have caught them

Both failures share one root cause, and it's worth naming precisely because it's a cheap mistake to keep making.

I looked for private equity, found none, and decided nobody was there. Not finding something is not the same as it not existing.

In fragmented industrial services the buyer is often a public company or a large operator, not a fund. Those don't show up where you look for sponsor activity. Transcat files with the SEC. IPS puts out press releases. Neither turns up in a search for new PE platforms, because neither is one.

The corrected check is three searches, not one:

  1. Sponsor platforms and add-ons.
  2. Public company acquisition history in the SIC or NAICS code. Read the 10‑Ks.
  3. Large private operators with a roll-up pattern. Press releases and technician headcount tell you more than deal databases.

I now run all three before a category goes above the line. Both of these would have come off the list in week one.

What I’m not writing about yet

One category made it through. I'm not naming it.

The whole argument for it is that no financial buyer is competing at entry and sellers currently have no bidder. Publishing a detailed case for it is the most efficient possible way to end that condition. I'd be handing a competitor a finished thesis and a head start on my own timeline.

So I'll publish it when I've either bought something or given up. Both outcomes make it worth reading. Neither is worth writing about while it's live.

Two things I still can't evidence, in the category that made it through.

I can show that the category is fragmented. I cannot yet show the size distribution, ownership pattern, or asset profile of the businesses inside it. Those are different claims and I conflated them in my first draft. The target profile in my model is a constructed assumption anchored to an adjacent industry, not an observed median.

I also have no category-specific operating benchmark. Every margin and labor input in my model is borrowed from a proxy industry, while my central argument is that this category behaves unlike that proxy. Both can't be true. That contradiction is unresolved.

Why publish this at all

Two reasons.

The first is selfish. Writing down why a category came off the list is how I stop re-screening it eighteen months later, and how the check that ruled it out becomes permanent.

The second is that a screen with no rejections isn't a screen. Anyone can list categories that look attractive. The work is in the disqualification, and almost nobody shows it. Which means the published version of this kind of research is systematically more optimistic than the real version.

My top-ranked category ranks where it does partly because two others came off the list. That's a weaker endorsement than it looks, and I'd rather say so.

If you own one of these businesses

Two things worth knowing, and they cut in opposite directions.

Your business is probably worth more than it was three years ago. Not because of anything you did. Because defense sourcing rules tighten on 1 January 2027, because a supply chain is being rebuilt around you, and because buyers with real money have started looking at categories they ignored for decades. If you hold AS9100 or ITAR registration, more so.

And the window where you have a choice is narrower than it looks. Once a platform forms in your category, you stop being an owner deciding when to sell and start being a target in someone's add-on pipeline. Calibration shops had that choice in 2023. They mostly don't now.

None of that means you should sell. It means you should know what you're sitting on before someone else tells you.

If you run a calibration lab, a motor shop, or anything else in the industrial services layer around this supply chain, I'd like to talk. Not to pitch you. I'm more interested in what I have wrong than in confirming what I have right, and the fastest way to find out is from someone who has run the P&L.

What a conversation with me looks like  ·  matt@stonehammer.ai