The reason your senior role is still open has nothing to do with pay
In this issue
- Why $130k keeps showing up as the number three different clients can’t get past, and why candidates aren’t buying the workaround
- What it really costs to build an engineer instead of waiting for one, from a firm that runs the numbers
- Three reads I found interesting this week
- A ten minute audit that tells you who to promote before you post your next role
Welcome to the first issue
I’ve spent twelve years recruiting in US industrial automation, and I started CandidTalent two years ago because I care about where this industry is going more than I ever expected to.
This is where I write down what I’m seeing in the market and where it looks to be heading, from the hiring side of it rather than the outside looking in.
Stop waiting for the finished engineer. Build them.
Three new roles landed on my desk this month and each of them stopped at $130k on base. Two of those roles are still open, and not for a lack of money in the end. The certified, plug and play engineer everyone writes into the brief is the same person four other firms are chasing, and there just aren’t enough of them to go round.
And the candidates aren’t buying the workaround either. When a client caps the base and leans on an unguaranteed profit share to bridge it, the good ones mark that variable down to almost nothing. One engineer on $122k this month wanted the base held at his number, and another walked away from a role in NC because a soft base propped up by an uncertain bonus wasn’t worth five days a week on site.
Between 2021 and 2025 the average number of automation and control engineers assigned to a single system integrator’s projects rose from 105 to 168, according to Control Engineering’s SI Giants research, which is a 60% jump in the skilled headcount one integrator needs to run its work, and it happened inside four years.
Now set the supply side next to it. Nearly half of automation engineers are 55 or older, and 26% are already past 60, according to Automation World’s 2025 Salary and Career Survey, so this isn’t a distant cliff, it’s this decade. Zoom out and the whole sector is aging at once, with the share of manufacturing employment at firms where a quarter of the workforce is over 55 climbing from 14% in 2000 to over 40% by 2022 (US Census Bureau, December 2025), while Deloitte and the Manufacturing Institute put the shortfall at up to 3.8 million manufacturing roles to fill by 2033, with around 1.9 million likely to go unfilled.
So demand for skilled engineers is up 60% in four years while the most experienced half of the workforce heads for the door, and nearly 58% of firms already say they’re struggling to fill open engineering roles. That squeeze sits underneath every autonomous plant, agentic AI rollout and self optimizing line the industry is currently excited about.
The direction itself isn’t in question, it was all over the feed this fortnight, with Jeff Winter back from Hannover calling industrial agentic AI the defining trend of the show and Vatsal Shah setting decades of OT innovation next to a modern AI box and a billion times the compute. The tools are real and they’re moving quickly.
But a tool doesn’t deploy itself. Someone has to integrate it, commission it and get operators to trust it, and that someone is in far shorter supply than the computing power. So the future doesn’t arrive when the technology is ready, it arrives at the speed of the people who can put it on the floor.
I’m seeing that gap most clearly in one skill at the moment. Ignition came up as the binding constraint in nearly every client call I had this month, and one integrator on the East Coast told me a key customer has promised them hundreds of millions in work on the condition that they have the certified headcount to deliver it. Rather than hold out for finished hires who never arrive, they’ve started taking capable people from adjacent industrial automation backgrounds and brought in a dedicated trainer to close the gap in house.
The economics back that up more than most firms realize. Bill Mueller, who runs the integrator Lucid Automation, put real numbers on it: training a new grad costs his firm twenty to twenty five thousand dollars over a three month internship, and they don’t break even on that engineer for twelve to eighteen months. He also runs about a 70% attrition rate on those interns, and his point is that the number is the model working rather than failing, because the ones who leave move into adjacent automation roles and the ones who stay become the engineers clients later ask for by name. That is a very different way of thinking about cost than leaving a seat open for a hire who never comes.
None of this means the experienced middle has emptied out, and I want to be careful there, because there are plenty of capable engineers in their 30s and 40s and a piece that implies otherwise is just a doom headline dressed up as insight. The narrower and more useful point is that the exact certified platform match most briefs ask for is scarce and getting scarcer, while the merely capable person who could be that engineer inside six months is sitting in the market right now, screened out by a job spec written for a hire you can’t actually make.
Four implications for an automation business reading this
- Rewrite the spec to the real floor. Put down what good enough to start actually looks like rather than the wish list, because the gap between someone who can do the job today and someone who can be doing it by Q1 is where most of the hires you keep missing are hiding.
- Treat certification as a training cost, not an entry gate. Twenty to twenty five thousand dollars and twelve to eighteen months to break even on a developed engineer is a smaller number than another nine months of an empty seat holding up a live project, and the firms that have done that math are the ones filling roles while everyone else bids against themselves.
- Build the old guard to new guard handover into billable work. The principal in his late 50s who’s about to retire isn’t a documentation exercise to be captured into a wiki nobody opens, he’s the quickest route you have to a deployable junior, but only if you pair them on live jobs while he’s still there rather than after he’s gone.
- Your fastest senior hire might already be on the payroll. The engineer who’s one skill short of the open role, developed on purpose, will usually beat a nine month external search on both time and risk, and most firms look outward first out of pure habit and end up paying for it twice.
The ground is starting to shift on its own, with vocational and community college enrollment up nearly 20% since 2020 and certificate programs growing four years running (National Student Clearinghouse via the MISUMI report). The bottom of the pipeline is rebuilding, just nowhere near fast enough to fill the roles you’re staffing in 2026 and 2027, which is exactly why the firms who build their own now are the ones who’ll have the engineers when the boom actually needs them.
If the perfect certified engineer for your next role doesn’t exist to hire, what would it take to build them in six months instead of hoping the right person is out there?
Worth reading this fortnight
- Jake Hall. The fortnight’s biggest automation post, on the veterans keeping machines running decades past their upgrade date and the knowledge leaving with them, and he’s right on the risk but only half right on the fix, because that knowledge only really sticks in a younger engineer stood next to them, not in a document. LinkedIn, July 2026.
- Connected worker at Hershey and Keurig Dr Pepper. An AI vision system now catches the kind of mismatched barcode error during a changeover that once triggered a 19,000 case recall, which is about the clearest proof going that this technology is a second set of eyes for a skilled operator, not a replacement for one. Control Global, July 2026.
- Control Engineering, SI Giants. Integrator demand for engineers is up 60% since 2021 and “quality of young engineers” is now a named top concern, rising from 13% to 16% of firms year on year, so the scarcity is measured rather than anecdotal. Control Engineering.
Before you post that role, do this
Before you write your next senior job spec, run a promote before you post audit first, because your cheapest and fastest hire might already be sitting on the payroll.
Pull a list of everyone in the business who’s within a single real skill of the role you’re about to advertise, the controls engineer who’s never touched Ignition but has run SCADA for a decade, the process engineer who understands the plant better than any outside hire will manage in a year, and for each name write down the one gap between them and the open role along with an honest guess at the months and the money it would take to close it. Set that against the nine months and the recruiter fee an external search will cost you, mine included, and then take the top two names off the list and have the career conversation this week, before you post the role rather than after they’ve watched you hire over their head.
Two questions to pressure test it:
- Of your last three senior hires, how many could a current employee have grown into with six months of deliberate development?
- Which of your best engineers only finds out you were paying attention when a competitor calls them first?
Over to you
I want issue two to dig into whatever you’re actually stuck on, so here’s one question, and the reply button works: what’s the one hiring or retention problem in your business right now that you can’t get a straight answer to? Rates, retention, where the person is, how to keep them once you’ve got them. Tell me the real one and it shapes what I write next.
On my desk
A new retained search goes live for us at the end of this week, and I wanted to put it here before it lands anywhere else, because it’s a genuinely exciting one. It’s a rare chance for an industrial automation engineer who’s had half an eye on building automation to actually make that move, and to do it on greenfield projects rather than walking into someone else’s legacy system. Those don’t come round often, and the right person for it probably isn’t sat there actively looking, so if that’s you or someone you rate, the reply button gets you a first look before it’s live.
Alongside that I’m looking for Ignition engineers right across the country, so wherever you are, if you know the platform to a real standard or you’re a skill short and want to close that gap, I’d like to talk.
And there’s a strong one live right now, an 800xA or Experion engineer based in the South Central US, on a remote basis with 30 to 40% travel. If that’s your platform, it’s a good seat with a client who knows what they’re doing.
Reply to any of these and it comes straight to me, and I won’t put you near a role that doesn’t fit.
Jon
