Where hard technology lives or dies
The road.
Deep tech, climate, industrial innovation: it doesn't fail in the lab. The technology has to work, that's the entry fee, but it's almost never the reason these companies die. They die on the way to the market. I've worked that problem from four seats: founder, early-stage investor, operator in the hardest part of the climb, late-stage. Same lesson each time, in a different accent. This is the road, and where I worked it.
Walk it ↓The idea.
I built a company and sold it, and the trap I walked into is the same one that catches hard tech. Every customer wanted a bespoke solution, and bespoke feels like serving them when it's actually how you fail to scale. The discipline I learned the hard way: split it. An MVP everyone can buy, and a bespoke layer that carries the commercial reality on top. Done right, the bespoke customer stops being a customer and becomes a partner, sometimes an investor in the business, and when they've put money in, their conviction runs higher, because now they need you to win. The product was never the hard part. The shape of the deal was.
What kills it here
- no team yet
- founder–market fit unproven
- science is not a company
- nobody who pays identified
- every customer wants something bespoke
Receipt — Co-founder and operator, built and exited.
Early stage.
Once we'd backed the technology, I sat with founders on the commercial side. The technical depth was the reason they were in the room; it was almost never the reason they won. Two things separated the ones that made it. They could name the exact person who'd use it and the exact person who'd decide. And they weren't proving scale on numbers that could disappear, contract revenue that looks like growth right up until the contract doesn't renew. The platform businesses were the seductive ones, brilliant in theory and notorious to scale, especially in B2B. What I learned to want, early, was a structured view of how the thing actually compounds, or at least the founder having a real vision of that structure. Early and wrong looked identical until you forced those questions.
What kills it here
- pilot is not a product
- first-customer risk
- a market instead of a buyer
- no decision-maker named
- scale shown on metrics that can vanish
Receipt — Analyst, hands-on with founders, Wermuth, Plus VC, EU innovation programmes.
The valley.
I left the investor's seat to work this directly, on first-of-a-kind synthetic fuel. We had a regulation-backed alternative, and the obvious lever, the fear of high fines, was the thing that moved them least. The penalty set the stage; it never closed the deal. What closed it was the upside: security of supply and energy independence, not depending on a volatile resource someone else controls, and underneath that, insulation from a spot market that swings, with pricing that rewarded whoever committed first. The thing we thought we were selling and the thing they were actually buying were different things, and you only learn that in the room. The deadlock didn't break on cost or compliance. It broke on structure, the carrots assembled in the right order. Scaling hard technology is never one thing. It's a commercial reality built piece by piece around a molecule that, by then, is the least interesting part.
What kills it here
- the buyer and the financier each waiting for the other to move
- mispricing the long-term value of the customer
- regulation that shifts under the business
- moving targets
- first-plant capex no venture or infra fund is sized for
Receipt — Market Intelligence and AI Lead, eSAF.
Proven scale.
This is where the founder changes job. The deep-tech founder becomes the CEO of a business, and the work stops being the technology and becomes the engine around it. What surprised me most: once real cash flow is coming in, financing always appears. It stops being venture or equity and becomes debt, project finance, structures built on predictable returns, cut the cost, bundle the projects, turn one proven asset into a repeatable one. The game changes from managing an innovation to running a business, and by then it's a flywheel that just needs turning. Boring, predictable, reliable cash. Boring turned out to be exactly what winning looks like, and a measure of how far the start of the road still has to go.
What kills it here
- cost-curve defense
- policy durability
- commoditization
- the next technology
- refinancing risk
Receipt — Investment Associate, late-stage climate PE.
In closing
Four seats, one lesson in four accents: hard technology lives or dies on the commercial reality around it, not the merit inside it. So the question I keep returning to is simple. There's a brilliant technology sitting somewhere on this road right now, stuck at one of these stages, too early or too capital-heavy or too far ahead of its market.
Which stage is it stuck at, and how do I help it through.
© 2026 Anubhav Das · The road
Photography via Unsplash: Barnaby Woodrow, Abhishek R, Leon Seibert, Ozren Cuculic, Leo, Andrey Metelev.