The Manufacturers Who Will Still Be Standing in 2036
Manufacturing can be dirty work. And I've lived it.
Unless you sit in aerospace, semiconductor and electronics, automotive, or life sciences and medtech, the disciplines that run a shop floor haven't changed much in decades. The businesses built on top of them haven't either.
Set that against how fast the rest of the economy is moving. Finance has been rebuilt twice over in the last fifteen years. Technology rebuilds itself every eighteen months by design. AI is compressing that cycle further still.
Walk onto most manufacturing shop floors and you're looking at a business model that predates the internet, running inside an economy that no longer resembles the one it was built for. Every time I switch between working with developers and working with manufacturing teams I am struck by the contrast.
That's a pace problem, and it's a real one. But it isn't the only problem, and it isn't confined to the businesses running dated models.
Margins across manufacturing are wafer thin, and in the more sophisticated, highly certified sectors they're engineered to get thinner regardless of how current the operation is. Aerospace supply chains, some of the most advanced manufacturing environments in the world, run on multi-year agreements built around 'should-cost' reviews and productivity commitments, where suppliers are expected to keep finding cost reductions as the price for staying in the contract.
Lean thinking bought a couple of decades of headroom against that. In a rising cost environment, as we have seen particularly in the last five years, lean thinking on its own has limits, and most of that headroom has already been spent, in modern operations as much as old-fashioned ones.
A business model with no real differentiation, a demanding cost structure, and a customer base structurally entitled to expect more for less every year, is not a comfortable place to run a company from, whichever decade its equipment was bought in.
But things still need to be made, whatever happens to the pace of change or the shape of the cost base. Nothing about a compressed manufacturing sector or an accelerating cognitive layer changes that basic fact. Physical goods don't get automated out of existence, they get made by someone, or something, either way.
So this was never really a survival question. It's a market share question. In a demand pool that's shrinking in places and redistributing everywhere, who takes share, and who gives it up.
There's an argument, further down, for why the answer to that question might be better news for British manufacturers than it's been in thirty years. But first, the three things that decide it.
The three things that actually decide it, and why they're all relative
I've written extensively about the Cognitive Automation (AI) Era and the risk of economic and technological disruption, most recently in the article Two Disruptions You Aren't Pricing.
Most manufacturing leaders with margin constraints facing the next economic disruption have picked one story and are holding onto it. Either "we'll automate our way out of the margin problem," or "our customers aren't going anywhere, we've served them for thirty years so AI won't affect us."
Both are half right. Half right is usually how a business dies, just slowly enough that nobody clearly sees the cause until it's too late to change it.
There are three things that matter, and the mistake most leaders make is treating them as absolute measures of their own business. All three are about position relative to everyone else chasing the same customers.
Demand resilience. Not "will demand exist," but what share of a demand pool, possibly shrinking, possibly redistributing across segments, you hold relative to the competitors chasing the same pool. A market can contract by a third and a well-positioned business can still grow, because share is what's actually being contested, not the size of the pie.
Automation pathway. Whether your cost base falls faster than the next business chasing your customers. This is a wider question than most leaders are giving it credit for.
It's not just cognitive automation, the AI layer everyone's talking about. It's physical automation on the shop floor, and it's the repetitive administrative burden sitting underneath both, the cognitive paperwork and the physical handling that nobody's bothered to redesign because it was never anyone's job to.
Automation, though, isn't simply lean thinking taken further. Lean is process redesign, stripping waste out of how work currently gets done, and it can be achieved entirely through discipline, with humans still doing the work. Automation is capital substitution, changing what does the work rather than how carefully it's done.
Lean bought you a couple of decades of headroom. Automation is what you reach for once that headroom is spent, and the business that reaches for it first, while a competitor is still relying on lean, takes their cost base apart from underneath them. There are clear implications for the skill sets your team needs for this, but that's for another discussion.
Protectability. The mechanism that actually lets you hold or grow share within the pool, rather than lose it to a competitor doing the same thing better. This is the one leaders skip past fastest, and it's the one that decides whether resilient demand and a good automation pathway actually convert into share gain, or just get matched by everyone else in the market.
There's a small set of mechanisms that make a manufacturing business hard to take share from. Every business, regardless of sector or size, holds some mix of them. How much they draw on them is another matter.
Process power. Tacit organisational knowledge, built up over years, that's hard to replicate even with identical equipment and a copied process sheet. Usually the strongest mechanism in manufacturing, and the hardest to see from outside the business.
Switching costs. Qualification and certification burden. AS9100 in aerospace, FDA validation in medtech, PPAP in automotive. A customer who's spent eighteen months qualifying a supplier doesn't requalify a competitor over a small price difference.
Cornered resource. Proprietary tooling, unique material access, specific key people, capability that isn't easily reverse-engineered from outside.
Scale economies. Cost advantages that only appear at volume, dominant in commodity and high-capital sectors like semiconductor manufacturing and automotive tier-one supply.
Branding. Perceived value beyond function, mostly relevant in consumer-facing bespoke or luxury segments.
The only question that actually matters is which of these a given business holds, and whether it's being maintained or slowly eroded.
The middle period nobody's pricing in
The long run is the easy part to theorise about. Relatively wealthy institutions and individuals will always need things made, and eventually the market restratifies around whoever's left holding the spending power. That's the comfortable thing to say in a boardroom. It reassures leaders, and usually that's where the conversation ends.
Back to business.
The problem is the journey from here to there. As the cascading implications of exponentially improving technology and the associated adoption take hold at a pace faster than any previous technological revolution, the divergence between adopters and non-adopters, the dynamic and the not so dynamic, the winners and losers, will be sharp. The implications for the whole of society in the medium term will be a challenge.
Diversifying into government, healthcare, education, and wealthy private clients isn't immunity from this compression, and it shouldn't be sold as such. Government procurement gets squeezed by budget cycles. Healthcare gets squeezed by funding pressure. Wealthy individuals get squeezed by asset price swings. None of those segments is safe on its own.
What diversification actually buys you is that four segments under different, largely uncorrelated pressures are unlikely to compress at the same time, in the same way a concentrated customer base can. That's a share-protection strategy, not a safety net.
The size trap
Large, established manufacturers often assume this question is already answered. Scale and certification came with the size of the business, not from a deliberate decision anyone made, and it's easy to read that as permanent protection rather than a position that's currently eroding.
Certifications commoditise as more competitors qualify. Capital intensity stops being a barrier once automation lowers the entry cost for smaller players too. The mechanisms a large manufacturer has been quietly relying on can be thinner than they look, and a great deal further along in eroding than anyone at the top of the business has had reason to check.
What makes this worse for a large manufacturer than a small one is that there's far less room to move once the erosion is visible. A small specialist can reposition. A large manufacturer with an entrenched model, built around the scale and certifications it's had for decades, has capital, workforce, customer base and certifications all built around the current position. That's a much harder thing to unwind.
It isn't impossible, though, and the businesses that manage it are proof of exactly the opposite conclusion to the one most large manufacturers reach for.
As Jim Collins wrote about in his iconic Good to Great, Kimberly-Clark's Darwin Smith took over a company built on coated paper mills in 1971, decided the traditional business was heading nowhere, and sold the mills, the original namesake plant in Wisconsin included, throwing the proceeds into what became Kleenex and Huggies. The market called it a bad decision at the time. Twenty-five years later the business he built was outperforming Procter & Gamble in most of its major categories.
The lesson isn't that entrenched leaders occasionally get lucky. It's that entrenched infrastructure was never a good reason to stay static, it was just the excuse most leaders reached for instead of doing the harder thing. The bigger the business and the more entrenched the position, the more that leader needs to be challenged on whether the position is actually still worth defending, not less.
Size buys comfort. It was never supposed to buy exemption from the question.
The Rematch
There's a part of this that's actually optimistic, easy to miss underneath everything above.
For a generation, manufacturers who couldn't compete on labour cost lost share to whoever could, wherever in the world that was. That's been the dominant axis of competition for thirty years.
If automation displaces labour cost as the thing that decides who wins the contract, that axis stops mattering in the same way. The businesses that spent a decade losing ground to cheaper labour elsewhere are being handed a rematch, on different terms, where proximity, quality, relationship, and the mechanisms above start to count again.
That's not a guarantee. It's an opening, and openings only convert to share for the businesses positioned to take them, whatever size they happen to be.
The technology decides less of that than most leaders assume. What decides it is whether the person running the business is willing to be challenged on their own position, scale and history notwithstanding, and act before the market forces the question. For the leaders who do, building resilient demand, a real automation pathway, and something that protects it once they've got it, the future here is a good one. For everyone else, someone else takes the share instead.
This is the first in a short series. The next piece looks at what you'd actually need to see, the specific signals and metrics, to know where your own business sits against these three measures today.
_____________________
Owner-managed businesses stagnate quietly, and most leaders don't see it until it's too late. I work with the leaders of those businesses, manufacturing especially, on a dynamic shift through the physical and cognitive automation era: not bolted-on AI, but leadership, culture and strategy rebuilt to withstand it.