The Next Industry to Get Product Management Is Yours

TL;DR

Product management, as a discipline, is about to leave tech. HVAC portfolios, regional banks, school districts, state agencies, and private-equity operating models are beginning to hire their first product operators.
The demand is driven by AI-era software that cannot be adopted by institutions without someone owning the adoption, the evaluation, and the operating-model change. That owner is a PM, whether or not the hiring institution calls them one.
The supply is driven by the tech contraction: the same talent being cut from big-tech payrolls is about to become the lubricant for PM adoption in industries that never had the function before.

For twenty years, product management lived inside tech. The discipline was developed, refined, and promoted inside software companies for software companies. PMs at banks were rare. PMs at state agencies were a category error, not a job title. You'd occasionally find a product-adjacent role in those institutions, but it was rarely the same job, and it was rarely called "product."

That's about to change. Not as a prediction for 2028 or 2030. As an observation of what's already moving in 2026.

The question isn't whether product management will leave tech; it already is. The question is which industry goes next, how fast, and whether the PMs being cut from big-tech payrolls end up seeding those transitions or end up reskilling into entirely different work. My bet: the seeding is about to accelerate, because both sides of the market (demand and supply) are converging at the same time.

Why product management is leaving tech

Three forces are pushing PM discipline out of tech and into non-tech institutions at the same time. They reinforce each other.

AI software cannot be adopted by institutions without a PM owning the adoption. In previous technology waves, institutions bought software and deployed it. The software either worked or didn't. Adoption was binary: turn it on, train users, done. AI software is different. It requires context curation, evaluation harnesses, risk-tiered rollout, and ongoing governance of outputs that change as the models underneath change. None of those are one-off deployments. They are ongoing product management functions, with feedback loops and trade-offs that need someone specifically accountable. That someone is a PM, regardless of what their title says.

Tech is cutting the supply of trained PMs at the same time non-tech is developing the demand. The big-tech talent swap is shedding PMs at scale, specifically PMs whose skills are being repriced inside AI-native operating models. Many of those PMs still have the core methodology: problem shaping, stakeholder management, roadmap reasoning, decision-making under uncertainty. They're not obsolete in absolute terms. They're obsolete relative to the AI-native market in tech. But the discipline they carry is well-matched to institutions that have never had it and are about to need it badly.

The institutional operating models are changing fast enough that they need help. Non-tech institutions have historically run on slow operating models: annual planning cycles, waterfall procurement, vendor-managed IT. Those operating models cannot absorb AI software at the pace the software is evolving. Institutions either re-learn how to operate or they fall behind their competitors. The re-learning requires an operator whose job it is to hold product accountability. That's the PM role, exported.

The industries already hiring

A few examples from the last twelve months, based on conversations with operators who've made these moves or hired for them.

Private equity operating teams. PE funds have been hiring product leaders into their operating teams and sometimes into portfolio company CXO roles. The thesis: portfolio companies need help adopting AI at the operating-model level, and the fund can centrally build the expertise once and deploy it across 10–30 portfolio companies. This is the fastest-moving sub-segment of the dandelion effect, because PE fund economics reward anyone who can uplift portfolio performance by even 2–3% in a repeatable way.

Regional and community banks. Not the big four. The mid-size banks that have under-adopted technology for a decade and are now staring at AI-native fintechs eating their customer base. They've historically lacked product functions; the adoption of AI lending, AI fraud detection, and AI customer interaction requires someone with product accountability, and they're starting to hire for it. Compensation is lower than tech, but the mandate is sharper and the work is visibly impactful.

HVAC, plumbing, and trades-services aggregators. Rollups and franchise-model services companies have been quiet heavy adopters of AI (scheduling, routing, quoting, customer intake via voice agents, predictive maintenance). They need product leadership, but they mostly don't know that's what they're hiring for. The early movers are hiring ex-tech PMs with a 40% pay cut and a title like "Head of Technology" or "VP of Operations Software." The role is a PM role in everything but name.

State and federal agencies. Slower-moving, but real. Agencies are running AI pilots that require someone to own the evaluation, governance, and scale-up decisions. Traditionally this work was absorbed by program managers or IT directors. It is increasingly being hired explicitly as product management roles, especially in jurisdictions with mandates around AI deployment in public services.

School districts and higher-education systems. This is the surprising one. The first wave of AI adoption in education was chaotic. The second wave, starting now, requires structured product ownership: which tools are approved, how are outputs evaluated, how is teacher workflow integrated. Districts that hire a single product-trained operator are getting results. Districts that don't are quietly drifting into an AI patchwork that no one owns.

None of these are the destination most tech PMs think of when they think about their career. All of them are about to become real hiring channels for PMs who are willing to see them as such.

What the work actually looks like

The specific daily practice is different from tech PM work. I've watched several ex-tech PMs make this transition and there's a consistent adjustment curve.

The stakeholders are different. Inside tech, your stakeholders are other product people, engineers, designers, and go-to-market. Outside tech, your stakeholders are operations leaders, compliance officers, front-line workers, and the CFO. The political economy is different. Persuading a regional bank's risk committee to approve an AI lending pilot is not the same meeting as persuading a tech exec team to greenlight a new feature. The skill transfers; the posture doesn't.

The cadence is slower, then faster. Non-tech institutions run on slower decision cycles at the top (monthly or quarterly rather than weekly), but once a decision is made, the operational adoption can move faster than tech because the institutions are less over-processed. You wait three weeks for the bank's risk committee to bless the approach, and then you roll out to 400 branches in six weeks because the branch managers actually do what they're told. This takes adjustment for tech operators used to shorter loops and looser execution discipline.

The definition of "shipping" shifts. In tech, shipping a feature is a technical event. In non-tech, shipping a feature is a change-management event. The code going live is the start of the work, not the end. This is uncomfortable for tech PMs who expected the change-management layer to be handled by someone else. It's also where the disproportionate value is created; the operator who can do the change-management work in addition to the product work is absurdly valuable in these institutions.

Evaluation matters more, not less. Non-tech institutions have lower tolerance for AI errors (regulated industries especially) and fewer tools for catching them. The PM who brings eval practices to a regional bank is creating a function that bank didn't know it needed. This is the cleanest edge a tech PM has in the non-tech hiring market: evaluation fluency is genuinely rare outside AI-native tech companies, and it maps to the biggest failure mode these institutions face.

Why the contracting tech workforce is the lubricant

The demand side explains why the industries need PMs. The supply side explains why it's happening now specifically.

The big-tech cull is putting thousands of trained PMs on the market at the same time that non-tech institutions are beginning to adopt AI at scale. The match is imperfect (the compensation gap is real; the ego gap is larger) but it's the match that the market has. Non-tech institutions could not have built this many PMs organically on a ten-year timescale. The tech contraction is compressing that supply into 18–24 months.

A useful way to think about it: the tech industry spent twenty years training product operators for a specific operating model, and that model is now surplus inside tech. The training doesn't evaporate. It's becoming available to institutions that could never have afforded to run that training themselves. Banks, utilities, aggregators, agencies, and education systems are all about to hire PMs who would have been running product teams at 10-year-old SaaS companies a decade ago.

The compensation will re-price. The initial wave of non-tech PM hires are negotiated down heavily because the employer doesn't know how much a trained PM costs and the PM has no outside offers to anchor on. Within two or three years, the second wave will re-price upward as the results from the first wave become visible. The operators who take the initial, undervalued offers are usually making the right skip move, because the skill-stack they build is scarce, and the compensation catch-up follows the results.

What this means for tech PMs considering the move

Three practical observations.

The institution doesn't know what it's hiring for. Most non-tech institutions posting PM roles in 2026 don't have a clear picture of what the role does. The job description will be wrong; the reporting line will be weird; the success criteria will be fuzzy. This is a feature, not a bug. The PM who walks in with a clear operating model gets to define the role around themselves. The PM who expects a well-structured tech-style PM job will be frustrated for six months until they realise the job was always theirs to shape.

The ego cost is higher than the career cost. Going from Senior PM at a recognised tech company to Head of Product at a regional bank no one outside the region has heard of feels regressive. On a skill-stack and compensation trajectory of 3–5 years, it usually isn't. The bank job builds scarcer skills. The regional-bank-to-PE-operating-partner jump, or the HVAC-rollup-to-vertical-SaaS-CEO jump, is real and already happening for the early movers.

The timing window is relatively narrow. Once the non-tech institutions figure out the value of trained PMs, they will re-price and the compensation will catch up. The arbitrage period where you can get a meaningful role with a 30–40% pay cut is 18–36 months in my estimation. After that, the roles will still exist and will still be real jobs, but the outsized skill-stack return from being early will have been compressed by the broader market.

This is the same pattern that played out in the 2000s when digital marketing left agencies and got absorbed into every Fortune 500 company. The first-movers who took the "in-house digital marketing lead" jobs at 40% pay cuts in 2008–2010 ended up running marketing at the companies that mattered by 2015. The ones who stayed in agencies ended up fine. They didn't end up extraordinary.

The non-tech institutions that will miss this

Not every industry is going to adopt. The ones that won't, or can't, are the ones where AI adoption is gated by cartel or regulatory dynamics that the institution doesn't control. Law firms (partly), healthcare providers (mostly), and some financial services niches (specific ones) are in this position. They will adopt AI eventually but the adoption will move through structures that don't reward product-operator-style ownership.

The industries that will adopt first are the ones where the individual institutions compete directly with each other on operational efficiency and customer experience, without regulatory structures preventing them from adopting faster. Those are the industries that need PMs first, and those are the industries where tech PMs will end up.

Meanwhile, the institutions that think they don't need product management for AI adoption are the same institutions that think they can buy AI like they buy ERP software: a one-time deployment with quarterly support. That's not how AI-era software works, and those institutions will either figure it out or be out-competed by peers who did.

The seed already landed

If you work in a non-tech institution that isn't yet hiring a PM, the seed has probably already landed somewhere in your sector. Someone is making the experiment. Someone hired a former tech PM eighteen months ago and is running circles around their peers by now. The question is whether your institution recognises the pattern and moves while the talent is still in the market, or whether it waits until the pattern is obvious to everyone and the talent has been absorbed.

If you're a tech PM considering the move, the seed is already blowing past your current role. The discipline you spent a decade developing is about to have more value outside tech than inside it. You can take the compensation hit now and capture the skill-stack return, or you can wait for the compensation to re-price and be one of hundreds of PMs competing for the same roles.

The dandelion doesn't decide where the wind takes it. Most of the seeds don't land in good soil. The ones that do root fast and change the shape of the field they landed in. That's the pattern about to run across a dozen industries at once, and the operators who catch the early part of it will quietly run the functions that get built.

Frequently Asked Questions

Is this different from the "PMs should leave tech and go to real industries" advice that's been floating around?

Yes. Previous versions of that advice were ideological (tech is overvalued, real-world industries matter more). This is structural: the specific skill set PMs carry (problem shaping, evaluation, operating-model change, risk-tiered rollout) is exactly what non-tech institutions need to adopt AI, and both the demand and supply are compressing on the same 18–36 month timeline. The driver is skill-set scarcity, not the appeal of "meaningful work."

How do I evaluate a non-tech PM role before taking it?

Three checks. First, does the institution already have any technology decision-making infrastructure (or is the role filling a vacuum)? Both can be fine; the first is easier, the second is more impactful. Second, does the reporting line give the PM access to the operating decisions they'll need to influence (usually this means reporting to the COO or CEO, not the CIO or CTO)? Third, does the compensation include any upside tied to the operational impact, or is it pure base salary? Pure base salary is fine, but a comp package with impact-tied upside is often available and signals that the institution is serious.

Won't AI just replace these PM roles before they become established?

Eventually, partly. The execution layer of product management (writing specs, running standups, drafting briefs) is being absorbed by agents in the same way that's happening inside tech. The decision layer (what to build, how to evaluate, how to sequence adoption) is not being absorbed anywhere. The PM role in non-tech institutions will evolve the same way it's evolving in tech, but the role itself is about to become more important, not less. The execution compression means one PM can cover a bigger scope; the decision volume in an adopting institution means there's more scope to cover.

What if I'm an operator in a non-tech industry, not a tech PM?

You have the inverse advantage: domain depth in an industry that's about to need PM skills. The highest-leverage move is to develop product management fluency deliberately (the core methodology is teachable; reading the product builder handbook is a good start), rather than ceding the incoming role to an ex-tech PM who doesn't know your domain. Domain-plus-PM is scarcer than PM-plus-industry-learning, and the pay scales reflect that in most of the sectors dandelioning right now.