Who's Going to Build It?

Two years ago, a pipe burst in my basement. What followed should have been a simple home improvement project, but what it turned into was an elaborate saga with one central theme: finding the contractors to do the job. The work itself was not complicated, but since every qualified subcontractor in a 50-mile radius was already booked solid – from electricians to HVAC techs, to a decent tile guy – none of them were particularly interested in a residential basement when commercial and industrial jobs paid better and ran longer. With a narrow supply of qualified vendors, they had the luxury to pick their customers to the point that I felt like I had to beg them to give me a quote.

I didn't immediately see the business insight but the more data points I had, the more it started to look less like bad luck and more like an example of a problem a lot of people in the industrial space are about to run into, if they haven't already – just at a much larger scale, and with a lot more capital at stake.

Here's the premise for buy-and-build in industrials: it's the most capital-efficient growth strategy in the middle market. You acquire a platform with a working operational engine already in place – trained crews, customer relationships, a book of recurring business – and you bolt on adjacent companies to gain regional scale, operational capacity, and multiple expansion. That's the pitch, and, importantly, it's not wrong.

It's also missing something. A business that "already runs" is quite a big assumption – it assumes the people who make the platform run today will still be there to run it tomorrow, next year, and three years into your hold period when you're deep into the fourth or fifth add-on. In a meaningful slice of industrials sub-sectors, that assumption is, shall we say, problematic at best, and not in a way that doesn't show up cleanly in a quality-of-earnings report. You can diligence customer concentration, contract terms, and management depth. What's much harder to diligence is: who's physically doing the skilled work, how old are they, and is anyone coming up behind them? If the answer is "eh, let me check," you just might have a big problem on your hands, and it’s getting bigger by the day.

This isn't a one-sector problem, and that's what makes it worth thirty minutes of a Wednesday morning.

Let’s look at construction. Back in 2021, I was driving back from a fantastic concert at Tanglewood. It was a long drive, so I stumbled upon a podcast episode about America's housing shortage. Behind the overly simplistic framing (boomers hold outsized portion of real estate because they're not moving, leaving scant options for millennials) lied a real, foundational problem that stuck with me since: During the 2008 financial crisis, homebuilding collapsed from roughly 2 million new homes a year to as few as 500,000, and stayed depressed for years. That sent a decade-long message to an entire generation of would-be electricians, plumbers, and carpenters that there was no future in the trade. A recession is temporary, but the signal it sent about career choice wasn't. As a result, the average construction worker is now 42, and 1-in-5 is over 55 (my GC and his foreman – squarely in that age bracket). Multiple industry estimates converge on the same uncomfortable truth: roughly 40% of the current workforce could retire within the next 5 years with nobody to replace them.

Utilities taught me this lesson personally, a few years before that. In 2018, I was working with a team at Exelon on what would become Aquify, turning the company’s deep municipal relationships and underground infrastructure management best practices into a standalone analytics business for water utilities. What struck me then was that every conversation we had with Department of Public Works heads referenced how much of the actual operating knowledge lived in the heads of maintenance and inspection crews who'd been doing the same work for decades, with strikingly few people behind them learning it. The Census Bureau's own recent data confirms it at the industry level: the share of utility-sector employment concentrated at firms where a quarter or more of the workforce is over 55 rose from 35% in 2006 to 80% (you read that right) in 2022 – one of the sharpest concentrations of any industry the Bureau tracks. The people who know where the bodies are buried, literally and figuratively, in a 60-year-old substation are aging out faster than almost anywhere else in the economy.

Technical testing is another interesting example of this problem with a sharp twist. RESA Power – an electrical testing and maintenance company acquired by Kohlberg in 2025, and one that's been actively rolling up smaller competitors since – operates in a field governed by a certification body, NETA, that defines a multi-year progression from trainee to senior technician. You can't hire your way out of a shortage here the way you might backfill a general laborer, because the credential itself takes years to earn. NETA's own trade publication has been warning, in plain language, that an entire generation of its certified technicians will be fully within retirement age by 2030. When a platform in a technical space like that talks about buy-and-build, the "build" isn't metaphorical – it's literally hundreds of licensed technicians per acquisition, and there's no fast way to replace them when they retire.

If you’re in manufacturing, you might point to AI as the escape hatch. This is a point worth taking seriously, but just not uncritically. Jeff Bezos is reportedly assembling a ~$100 billion fund, described in his own investor materials as a "manufacturing transformation vehicle," aimed at buying industrial companies in chipmaking, defense, and aerospace. His other venture, Project Prometheus, is applying the AI tools it develops to modernize such companies. Reporting on the fund explicitly frames the opportunity around companies struggling with labor costs and production bottlenecks – which is, in effect, one of the best-capitalized bets in the world that this exact problem is real and severe enough to buy into. That's worth noting for anyone tempted to assume the labor gap will just resolve itself. It's also worth noting how far the actual technology has to go: watch footage from this year's World Humanoid Robot Games in Beijing, where machines that can outrun Usain Bolt also can't reliably stop without bursting into sparks, and you get a pretty good sense of the gap between "impressive demo" and "replaces a licensed electrician inside a five-to-seven-year hold period." The optimism isn't wrong. It's simply too early. And "early" is not a word GPs typically like hearing when it comes to the thesis underlying their next platform acquisition.

So what is an industrials-focused GP to do? The standard answer is: “it's a recruiting problem; we’ll solve it with better wages and marketing to Gen Z”. But every operator in every one of these sectors is already trying that, so you’d be competing for the same (limited) talent. The more interesting plays turn the problem into an asset rather than a cost.

First, treat training infrastructure as a product, not overhead. If your industrial portco has to build serious apprenticeship or certification capacity to solve its own staffing gap, that capacity is sellable – to smaller competitors, to customers, even to other portfolios facing the identical problem. This isn't theoretical; some of the more thoughtful home-services roll-ups have already built internal apprenticeship programs specifically for this reason, and the scale of outside capital now flowing into large-scale trades training tells you serious money doesn't see this as a charity project. Rite Way, a Tucson-based HVAC company that is part of Redwood Services, runs Rite Way University, a paid apprenticeship and training program designed to teach beginners how to become HVAC, plumbing, and electrical service professionals.

Second, extend your most experienced people rather than simply losing them. A remote-support model, wherein one senior technician oversees several junior field crews via camera and structured diagnostics (closer to telemedicine than traditional field service), stretches your scarcest resource further and provides a retention opportunity for your veteran techs whose chief motivation for retirement is field work. It can also become its own service line sold beyond the portco's own jobs. JCI has remote field support technicians, a role that requires a minimum of 10 years of field experience.

Third, capture the knowledge before it retires. The judgment a 20-year technician carries, e.g., what failing equipment sounds like before it happens, which shortcuts are safe and which aren't, is typically undocumented and leaves the company with them. Structured playbooks, video libraries, or an AI model trained on decades of service and inspection records, turn that tacit knowledge into an asset the business owns rather than a risk it's exposed to. Easton Select Group, a portfolio company of Brenton Point Capital Partners, recently launched AskEaston, an internal AI platform that captures the company’s institutional knowledge and makes it available to its customer-facing field techs.

One parting thought: if workforce depth is this central to whether a platform can deliver its growth thesis, it belongs in the diligence process for every add-on, not just the original platform. Sometimes the right acquisition target isn't the one with the best margins - it's the one with the deepest bench.

None of this is a reason to abandon buy-and-build as a strategy. It's one of the better ways to compound value in the middle market. But the diligence question worth adding to every industrials deal memo isn't just what you're buying. It's who's going to build it – and whether that person is still going to be there three years from now.