Pull up any climate hardware company's website. Then open their pitch deck. Then their LinkedIn. Then their careers page.

You're looking at four different companies.

Each surface is competent. The copy is adequate. The design is professional. The information is correct. Audit any single one and it passes.

And still something isn't working.

You've probably felt this yourself. You can explain the company perfectly to someone who already gets it. The investor who understands the space nods along. The engineering hire finishes your sentences. The problem shows up with everyone else. The journalist. The enterprise buyer. The partner who needs to sell it internally. Investor conversations stall at the same point. Enterprise buyers can't explain the product to their own colleagues. The careers page attracts the wrong people.

The instinct is always the same. Fix the most visible surface. New website: £30-50K and three months. New agency: six months of onboarding before they understand the product. New deck: a week of the founder's time.

Each helps briefly. The way repainting a room feels like progress on a house with no frame.

Then the drift returns. And nobody can explain why, because the thing that's broken isn't a thing anyone can point at.

The same structural failures keep appearing across technical companies. Not occasionally. Persistently. Three shapes that recur regardless of sector, stage, or how talented the team is. The patterns don't care how good the work is.

Five companies across climate hardware, materials science, computational engineering, robotics, and clean energy. Different sectors, different stages, different ambitions. All credible. All doing real work. Three diagnostic axes. Mechanism clarity: can someone understand

The patterns held across all five. Different sectors, different stages, same structural shapes.

The first shape is the one most companies recognise immediately: inherited narrative.

The story was right once. Seed stage, first launch, first market entry. Then the company grew. New products, new markets, new ambitions. The story didn't grow with it. What remains is a narrative from a previous era. Accurate enough that nobody questions it. Outdated enough that it quietly constrains every conversation.

The Series B company wearing seed-stage clothes.

This is the easiest failure to name. But the more common pattern is harder to detect.

The second shape: coherent but fractured. The story is strong when you look at one surface. The website tells a compelling narrative. The investor deck tells a different one. The careers page tells a third. The homepage says "redefining industrial infrastructure." The investor deck says "47% emissions reduction at lower unit cost." The careers page says "join a team that cares about sustainability." Three promises to three audiences. None of them wrong. None of them the same company.

Nobody inside has the vantage point to see the fracture, because the fracture only exists in the gaps between surfaces. Every team that owns a surface believes their version is the right one.

The deepest version is the one that feels like it shouldn't be a problem at all.

The third shape: mechanism without world.

The company can explain what it builds in extraordinary detail. The engineering is legible. The specifications are precise. You finish reading and you understand the product.

And you feel nothing.

The mechanism is fully articulated. The world it operates in, the stakes, the tension, the future it enables or prevents, is absent. The information is all there. The architecture that makes someone care is not.

These are not creative failures. The writing is often good. The design is often strong. The people are often talented. The failure is structural. An absence of architecture underneath the surfaces.

Structural failure doesn't announce itself. It feels like underperformance without a clear cause. Like the story should be landing but isn't. Like the problem is always somewhere else.

The reason these patterns repeat is not lack of effort.

Narrative architecture is invisible in both directions. When it's present, nobody notices. The story just feels obvious, like it could not have been any other way. When it's absent, nobody notices that either. The company just feels like it's not quite landing. The diagnosis never reaches the layer where the actual problem lives.

Every surface-level audit will return a passing grade. The website is fine. The deck is fine. The LinkedIn is fine. The problem exists in the relationship between them. The structural logic that should connect every surface to a single coherent architecture but doesn't. Because that architecture was never built.

There is no clean prescription for this. Maybe there can't be. But the patterns are more consistent than expected.

Narrative failure in technical companies is not random. It clusters. It repeats. And it's diagnosable, if you know where to look.

Which is precisely the layer nobody thinks to examine.