Product Cannot Belong to Anyone Else. In Healthcare, too.

I’ve found that healthcare companies have not made up their minds about who runs product.

A health-tech startup needs a product function of its own. Not a person sitting inside marketing. Not something engineering picks up when there is time. Its own seat at the leadership table, reporting to the CEO.

If you work in tech, you are already rolling your eyes. The bros and the sis alike. You have read Marty Cagan, you believe in empowered product teams with full ownership of their work, and you are wondering why this needs an essay at all. I am also biased here. I see everything as a product. A physical device, a software app, a service. This essay is a product itself. Not a very well-tested one.

However, after working with plenty of different people in healthcare, in the UK and the US, over the last decade, I found that it was rare to find people around me who saw the world the same way.

In one conversation after another, I heard the same things. That product belongs under marketing, or under science, but never under engineering, and that a standalone product division has no reason to exist. That the job of technology is simply to work, and since marketing and sales bring the money, marketing and sales should decide what gets built. And from product people themselves, the most honest version of it: that product teams are hired frequently to decide what the product should do, and then spend most of their time fighting with other, typically commercial, stakeholders who are just as opinionated about what the product should do.

So a dichotomy grew inside me. From one side, I had all these experienced people whom I deeply respected telling me that healthcare is different and that all these theories were great, but healthcare is complex and pragmatic. From the other side, I knew firsthand the problems of being a project/delivery manager and saw the gap between how great tech was developed and how medtech was developed. I was going crazy.

I tried to understand each side instead of arguing with it.

Sales will not get their bonus if they miss the quota this quarter. So what they need from technology is simple: unblock the sale and don’t limit their claims. That means no bugs or technical issues that can ruin a demo or a trial. It means not having to say no when a customer asks about a feature that matters to them. Marketing is successful when campaigns land. Technology serves them when it fits the brand narrative and offers a quick time to value, so their focus goes to polishing the features that carry the story and to anything that creates an immediate wow. Science is successful when the product achieves the clinical outcomes, or better outcomes than the competition. So they push the technology towards the best clinical standards, towards results that are repeatable, consistent, and backed by evidence. Engineering is successful when the technology is reliable and novel. Reliable, because that is what delivers the commercial and clinical promises, and because engineers are the ones who catch fire when everyone else is complaining about blockers. Novel, because the technology is often the core IP, and it should not be easy to copy.

The funny part is that none of them is wrong. Which is a recipe for analysis paralysis. So how do most health companies fix that? By being decisive and saying that from now on, a single department will be responsible for product. Or sometimes even worse: they list out all the product management responsibilities on a spreadsheet and say that one function will do some and another will do the others. While cutting the knot is many times a good solution, this problem is fundamental to how a company works and requires deeper thinking.

Regardless of the approach or the industry, a successful product has to be desirable, usable, feasible, and viable. There has to be a market. People have to figure out how to use it. We have to be able to build it with the resources we have. And there has to be a profitable business around it.

Healthcare sharpens each of these. Desirability depends on trust, which is why clinical evidence is not marketing material but a real driver of demand. Usability is not only about figuring out how to use something, but also about using it safely, and human factors are increasingly treated as a safety vulnerability rather than a nice-to-have. Feasibility is harder because of documentation, slow access to feedback, and the fact that you cannot run a fully flexible CI/CD process. Viability often depends on reimbursement, which is a science of its own.

Each function owns one of these dimensions almost perfectly, and is blind to the other three. Sales knows what is commercially desirable for the decision maker, because they sit close to the customer and understand the problem deeply. Marketing knows what is desirable at scale. Neither is trained to judge what is usable, feasible, or viable in a complex technical product, and both lean towards acquisition over retention. Science knows what is clinically desirable and has a strong instinct for usability, but is not used to assessing commercial desirability, feasibility or viability. Engineering knows what is feasible and has the best ideas about what is innovative, because they are closest to new technology, but is not trained to judge what is desirable or usable, and sometimes not viable either.

So the objections I kept hearing were not ignorance. Every one of those people was correctly describing the incentive they live under. That is exactly why someone has to be responsible for all four dimensions at once, and why that someone cannot report into any of the four.

The obvious answer is that this person exists and is called the founder CEO. It is written everywhere in reliable startup literature: in a startup, the founder CEO is the PM. And it makes sense. The founder CEO talks to customers, knows the problem better than anyone, and is usually the one who imagined how a new technology could change something in healthcare. They know what is desirable, and they can work with the right designers, engineers, and business people to make it usable, feasible, and viable.

In tech, this works. In healthcare, I have seen it break quite a few times, so there seems to be a pattern.

Being a good PM in healthcare requires much more depth. The stakeholders are many, and you keep discovering new ones, and each of them has something valuable to say. The domain is four domains: clinical, technical, reimbursement, regulatory. Feedback is slow and often indirect. In most industries, a PM can shop around, try the competitors’ products, feel the problem, sometimes use the product in their own life. In healthcare, you frequently cannot do any of that. And on top of all this comes the PM craft itself: OKRs, roadmaps, discovery, vibe coding prototypes.

Then there is the rest of the founder CEO’s job. Fundraising, grants, operations, finance. Becoming a great PM on top of that is not a matter of effort.

Younger founder CEOs are usually close to the modern product practice. They know the latest tools and can prototype something in an afternoon. What they lack is the clinical, regulatory, and reimbursement experience that tells them which of their ideas will die eighteen months from now for reasons they have never encountered. Older ones have exactly that experience, often earned painfully, and are usually far from how product is practiced today. Each group is missing what the other has. Neither gap closes by trying harder, and both gaps are invisible from the inside.

Which brings us to the part nobody wants to hear. If product cannot belong to sales, or marketing, or science, or engineering, and the founder CEO cannot carry it either, then the seat has to be filled by someone else. On purpose. With a name on it.

That someone can arrive in several shapes. A founding PM, brought in early enough to actually have a say rather than to write tickets. A product cofounder. A first product hire who is senior enough to disagree with the CEO in a room full of people and survive it. And in some cases, a hired CEO, so that the founder can go and do the thing they are genuinely world-class at instead of the thing the org chart handed them.

None of this is about titles. It is about making sure somebody wakes up in the morning owning desirable, usable, feasible, and viable at the same time, and reporting to nobody who owns only one of them. Otherwise, those four get quietly distributed among four departments, each of them right, each of them defending a correct position, and what ships is the average of four correct opinions. Which is not a product. It is a compromise with a launch date.