I did not come up through medical devices. I came into the industry the way I have entered every market I have built in, as an outsider trying to build an accurate model of how value actually moves. The single fact that reordered my thinking faster than any other was this: in medical devices, the device is not the product. The payment is.

You can have a technically superior device, a clear clinical benefit, and engaged physicians, and still fail, because the economics of who chooses, who pays, and who profits were never designed before the product was. The clinical story gets the applause at the conference. The economics decide whether there is a company at all. This is the part of the industry that is invisible from the outside, and it is where an enormous share of the value and the risk actually sit.

The usual version of this lesson is that clearance is not coverage. That is true, and everyone in the field will tell you so. The more useful version is more specific. There are four economic truths about how money moves around a device, and each one quietly kills good products built by teams that did not see it coming. What follows is the mental model I wish I had been handed on day one. It is not legal or regulatory advice, and the specifics shift constantly. It is a way of thinking about why so many good devices never become good businesses.

Think of the four truths as a map of where the money sits and where it leaks. Three of them are about who controls the dollars: the user and the buyer, the provider's margin, and the payer's slow machinery. The fourth is about time, the years between approval and payment that decide who is still alive to collect. A team that internalizes the map stops being surprised by the industry and starts designing around it.

The user is not the buyer

Start with a number that reorganizes everything else. In a typical hospital, the items physicians personally choose, the stents, the implants, the tools they reach for, make up something like 40 to 60 percent of the entire supply budget. The physician picks them. The hospital pays for them. The person who falls in love with your product and the entity that has to fund it are two different parties, in two different meetings, answering to two different incentives.

That single split explains most failed device launches. Win the physician and ignore the committee, and you stall in value analysis, where someone who never touched your device decides it is not worth the cost. Win the committee and never earn a clinical champion, and no one inside the room fights for you when the contract comes up for renewal. Neither half is optional, and they are almost never persuaded by the same thing.

So in devices you are always running two sales at once, to a user and to a buyer, and the product that wins is designed for both from the first day. If your plan only accounts for the person who loves the technology, you have built half a business and have not noticed the missing half yet.

The buyer's world has its own machinery that founders rarely see until it blocks them. Most large providers buy through group purchasing organizations and existing contracts, and a value analysis committee weighs any new device against what is already on the shelf at a negotiated price. A superior product with no contract and no committee sponsor is not evaluated on its merits. It waits. Understanding that machinery, and building the economic case it runs on, is as much a part of the product as anything in the device itself.

Covered is not the same as profitable

The second truth is the one that surprises outsiders most. For many procedures, the device itself consumes 30 to 80 percent of the entire payment the hospital receives for doing the case. The implant, in other words, can eat most of the reimbursement before anyone has paid for the staff, the room, or the time.

Read that against the usual assumption and the problem becomes obvious. A device can be FDA-cleared, Medicare-covered, and still lose money for the very people expected to use it. Coverage means a provider is allowed to bill for the procedure. It does not mean the arithmetic works once your device is sitting in the tray. The buyer is running that arithmetic whether or not you are in the room, and if it comes out red, your clearance and your clinical data do not save you.

This is why the reimbursement rate is not a detail you hand to a specialist a few months before launch. It is the product. The builders who win in this field can tell you, without looking it up, the exact margin their device leaves the provider on a real case. If you cannot, that is not a gap in your finance deck. It is the most important open question in your company.

The margin also moves with the site of care, which is why the same device can be a winner in one setting and a loser in another. A procedure paid one way in a hospital outpatient department is often paid quite differently in an ambulatory surgery center or a physician's office, and the share of that payment your device consumes shifts with it. A reimbursement story that works only in the most generous setting is fragile. The durable ones leave the provider whole across the settings where the procedure will actually be done.

None of this is a reason for despair. It is a reason for sequencing. The founders who win do the payment design in parallel with the device, so that by the time the technology is ready, the path to a paid claim is already mapped and partly built. The ones who struggle discover the path only after the device works, when the runway is short and the options are few.

A designation is not a dollar

The third truth is about the badges that feel like progress. By the end of 2025 the FDA had granted more than 1,200 breakthrough device designations. The fast-track Medicare coverage program built specifically for those breakthrough devices has room for roughly five of them a year.

A designation is a wonderful press release. It is not a coverage decision, and the two are separated by years and by a much harder evidence bar. This is precisely where founders get fooled. The badge feels like momentum, so the company quietly files payment under problems for later. Later arrives, the coverage pathway turns out to have room for a handful of products, and the designation is revealed to have promised nothing about dollars.

The discipline is to treat a regulatory milestone as the start of the payment work, not the end of it. In devices, regulatory strategy is reimbursement strategy wearing a different hat. A designation is a to-do list you have not started, not a result you can bank.

The two gatekeepers are also asking genuinely different questions, which is why clearing one tells you so little about the other. Safety and effectiveness, the regulator's bar, is not the same as whether a payer believes the device changes how a patient is managed enough to justify paying more for it. A device can be unmistakably safe, work exactly as designed, and still be judged by a payer as not worth a premium over what already exists. Designing the evidence for the second question, not just the first, is the work that separates a cleared device from a paid one.

The valley is also the moat

The fourth truth reframes the thing everyone complains about. From FDA clearance to Medicare coverage, the median wait for novel devices has run somewhere around five years. The industry has a grim name for the gap: the valley of death. It is where good companies with real products run out of money waiting to get paid.

Everyone curses the slowness. Almost no one notices that the slowness is also a moat. A market that forces every entrant to wait five years for payment is brutal, and it is brutal for your competitors too. The same barrier that nearly kills you is the barrier that keeps the next ten companies out of your category. Slowness is not only a cost. It is a filter.

Which changes the question you should be asking. It is not how to skip the valley, because you cannot. It is whether you are financed and disciplined enough to still be standing on the far side when the slow machinery finally turns in your favor. The prize in this field rarely goes to the fastest. It goes to whoever is built to outlast the wait.

In practice this rewards a specific kind of company. Not the best funded one, but the one that treats every dollar as a way to buy the evidence and proof points that shorten or survive the valley. Capital raised without a plan for the wait simply funds a more expensive failure. Capital raised against a clear path through the gap, milestone by milestone, is what lets a company still be standing when payment finally arrives.

What this means for how you build

Put the four truths together and they say one thing. In medical devices the economics are not downstream of the product. They are the product. The payment path deserves the same rigor, the same early design work, and the same talent you would never hesitate to spend on the device itself.

In practice that means a few concrete habits. Map, on one page, who chooses your device and who pays for it, and design for both. Know the margin you leave the provider on a real case, to the dollar. Treat every regulatory milestone as the opening of the payment work rather than the closing of it. And capitalize the company for the valley you will have to cross, not the one you wish existed.

It also means building the reimbursement function early, not bolting it on before launch. The teams that succeed treat health economics, coding strategy, and payer evidence as first-class workstreams that begin while the device is still taking shape, so the clinical study collects what a payer will later demand and the launch is not the first time anyone asks how the case gets paid.

The companies that endure in this industry are almost never the ones with the most elegant feature. They are the ones that understood, before they built, how money would actually move through the system and who would have to profit along the way for the product to keep being used. So the question I would leave any device founder with is a simple one that is surprisingly hard to answer well: do you know, today, exactly how a paid claim happens for your product, and who has to come out ahead at each step for it to keep happening tomorrow?