Science-rich, product-poor

The name a field gives itself tends to become the thing it organizes around. Earth observation names what it does: observe the earth. Climate tech names what it serves: the climate. Neither names who would buy it, nor the problem they would buy it to solve. I spent years building climate products inside EO companies, so I mean that from the inside, not as a verdict on anyone else’s.

The science was extraordinary. The product was built by scientists, and what they optimised for was accuracy. Whether a customer could do anything with that accuracy was a different question. The customer sat on the far side of a sales team that lived deep in insurance and government, so the people building the product were walled off from the people using it. The accuracy kept climbing, and the work of connecting what a customer needed to operate to what we built was owned by no one.

When one of these companies stalls, the science is usually still solid. What gives way is the question the name leaves out: who has this problem, how bad is it, what are they already doing about it, and how would the product help them make progress?

That is the product leg of a company: the function that owns the demand question and turns the answer into something a buyer cannot operate without. Without that leg companies are what I call science-rich, product-poor. The science gets the attention, the funding, and the founder’s identity. The leg that asks who is buying and why is owned by no one.

Climate tech does the same thing one level up. A field named after what it does, sitting inside a category named after what it serves, is organised around everything except demand. And the climate framing supplies a story for every symptom of the missing demand. Slow traction becomes an early market. No repeat buyer becomes a market that hasn’t matured yet. That story narrates a gap an ordinary business would be forced to confront, so who is buying stays unasked longer, and the silence costs more.

Then I advised companies in a completely different industry, one whose name gives nothing away. No science in it anywhere. Professional services, deep expertise, paying customers. Same structure. The expertise was hard to replicate, and the system that would have turned it into a scalable product kept getting postponed whenever resources got tight. Customer health lived in people’s heads instead of in any system. A piece of technology the business leaned on had been handed to an outside contractor, with no one inside owning it. No activity in its name, no cause to hide behind, and the demand question was still owned by no one.

That is what convinced me this is not just a deep tech and science problem. It shows up wherever a company is built around a deep capability and treats demand as something that arrives once the capability is good enough. Maybe a better term, then, is capability-rich, demand-blind.

So how do you spot a missing product leg?

The first signature is a conversion gap. Your own work generates demand you can’t convert. I worked with a company whose founder had written the papers that defined its field. Serious institutions read them and reached out, one after another, and every one of those conversations went to the founder personally, and nothing fed back into a repeatable pipeline. The science was manufacturing demand the company had no machine to process. When the interest your expertise creates has nowhere to land but the founder’s inbox, no one owns the product leg.

The second is an everything-roadmap. One strong core capability, aimed at six markets at once, with no owned bet on which one matters. From the inside that breadth feels like optionality. From the outside it is the signature of a segmentation decision nobody has made. Choosing the single buyer who can’t live without you is the core product judgment, and when that seat is empty the roadmap fans out instead of narrowing. I’ve watched a company with a strong technical core describe its addressable market so broadly it was hard to name a buyer it excluded. It looks like ambition, and what it actually is, is an absence of anyone whose job it is to choose.

The third is the permanent side project. In good years the company hires. In lean years it falls back to consulting and project work to make payroll, and product development is the first thing put down. The product becomes the permanent side project. Revenue pressure doesn’t only slow it down, it actually reveals that no one was structurally responsible for protecting it.

The fourth is the hire that can’t land. You try to hire the gap: a commercial director, a head of product, an outside consultant. The hire fails, and everyone reads it as a bad hire. It rarely is. You cannot hire a function the organisation isn’t shaped to hold. I’ve now seen the same failed hire in two deep-tech companies and a professional-services firm, which is how I know it’s the structure and not the person.

This is also why hiring a product manager so often changes nothing. The company wants product thinking and hands the new hire a roadmap, sprints, a backlog, the machinery of execution. That is project management, and mistaking it for product thinking is the whole problem. Product thinking is a different faculty: reading what the customer is trying to do, knowing the domain (or quickly learning it) well enough to see where the problem sits, and the judgment to decide what to build and for whom. You can hire execution. You cannot hire the demand question, or the authority to answer it. Drop a capable product manager into a company organised around its capability instead of its customers, and one of two things happens. They get absorbed into shipping the founders’ existing roadmap faster, or they reach for the demand decision and find the technical core will not give it up. Either way the seat stays empty. You hired someone to run a process around the hole where the function should be.

And you cannot out-build it either. A company can iterate hard and still be product-poor. Improving the thing makes you better at the loop you are on; it cannot tell you the loop was worth choosing. That is the demand question, and it stays unowned no matter how fast you ship.

Underneath all of it is one thing. The people who lead these companies are bound to the capability, not to the customer’s progress. The founder’s identity is the science, or the craft, or the years of expertise. The technical leader’s whole career is the exact thing the company is good at. No one’s identity is the buyer and how their business is changing. So the most important decision the company makes, the cut from capability to demand, has no owner, because owning it would mean filling a seat the founders’ own identities don’t occupy.

And there is a reason this seat stays empty, beyond no one being assigned to it. The answer to the demand question is almost always mundane. The buyer’s real problem is something operational and unglamorous, and solving it doesn’t feel like innovation to people who got into this to do hard science, so it gets passed over for the part that feels like progress. That is the mistake. The boring, mundane problem is usually where the innovation is; the impressive science only matters once it’s pointed at something a customer was stuck on.

Three questions find the empty seat quickly. Would your users notice if the product vanished for a week, or would they quietly go back to the spreadsheet that still works? When the customer’s problem and the science or the tech pull in different directions, which one wins? And who, by name, owns the decision about what to build next? If the honest answer to the last one is “the founder, in the margins of everything else,” you’ve found it.

For investors the implication is direct. Capital underwrites the science, because the science is impressive and shows well in a demo. The demand question is neither, so it goes unfunded and unstaffed, and the company spends its runway making the capability better instead of finding the buyer who can’t live without it.

Science-rich, product-poor shows up in any company built around a deep capability, climate or not, the moment the demand question goes unowned. Most of these companies have a product manager. Far fewer are a product organisation.