Open any digital health company's pitch deck and find the slide near the back. It is titled "Clinical Advisory Board." There are four or five headshots, each with a name, a specialty, and an institutional affiliation that a reader will recognize.
The slide is doing an enormous amount of work. It is telling an investor that serious clinicians have looked at this and approved. It is telling a health system buyer that the product was built with clinical input. It is telling a regulator, implicitly, that adults were in the room.
Now ask the question nobody in that meeting asks: what did those five people actually do?
For a meaningful share of advisory boards, the honest answer is: attended one call, gave some feedback that may or may not have been incorporated, and has not been contacted since.
The arrangement is well described in physician-facing commentary: advisory roles "typically involve a few hours per quarter," and advisors "usually receive equity in the company," commonly in the range of 0.1 to 0.5 percent vesting over two years.
Nobody records what was delivered. The advisor cannot prove which of their roles was substantive. The founder cannot prove theirs was real. And the investor reading the slide has no way to distinguish the two.
It is a trust instrument that carries no information.
The discovery problem: gamed signals all the way down
Start with how advisors are found, because that is where the failure originates.
A founder needing a clinical advisor has three practical sources.
CMS Open Payments, which is the only public database linking physicians to industry relationships. PubMed authorship, which identifies people who publish. Conference programs, which identify people who speak.
Every one of those is a proxy for something other than what the founder needs.
The Open Payments data in particular deserves careful reading, because it is routinely misused. A JAMA analysis covering 2013 to 2022 found:
- 826,313 of 1,445,944 US physicians, or 57.1 percent, received industry payments totaling $12.13 billion.
- The median payment was $48.
- The top 0.1 percent averaged $1,987,862.
Separate JAMA analysis found that physicians receiving over $50,000 constituted 3.4 percent of recipients but 82 percent of total value.
So the distribution is extraordinarily skewed, and a physician's presence in the database is close to meaningless. A median of $48 is a sandwich at a lunch talk. More than half of American physicians appear in this database, which makes appearing in it approximately as informative as having a medical license.
Meanwhile the top of the distribution is a small group receiving enormous sums, who are also, by definition, the most heavily committed and least available people in medicine.
So the primary discovery signal points founders toward exactly the people with the least remaining capacity, and provides no signal at all about the 57 percent in the middle.
The pharma version of this uses more sophisticated tooling, with vendors ranking key opinion leaders by publications, trial participation, and payment history. But as commentary in Nature Reviews Rheumatology observed, KOLs tend to be selected for "prescribing habits, memberships in organizations and contributions to treatment recommendations" rather than for knowledge as such.
The result is a market that misallocates in both directions simultaneously: too many ceremonial roles concentrated at the top, and complete invisibility for capable community clinicians below.
The signal being used is the signal being gamed
Here is the structural problem stated precisely.
Discovery runs on publications and payments. Both are, in the specific sense that matters, endogenous to the market being measured.
A physician who wants advisory work benefits from publishing and from prior industry relationships. Industry finds advisors by looking for publications and prior industry relationships. The measure and the behavior it is supposed to measure feed each other.
Meanwhile the attribute the founder actually needs, which is whether this clinician deeply understands the workflow the product must fit into and will engage substantively, correlates only loosely with either.
The community physician who has run the clinic where this product must work, who knows exactly why the last three attempts failed, and who would give genuinely decisive feedback, appears nowhere in any of the three discovery channels.
Why the market is starting to notice
Two developments are making this less tolerable than it was.
Buyers have stopped being impressed. Industry reporting on 2026 digital health funding describes buyers with "zero appetite for failed technology rollouts." When health systems are burned, the advisor slide moves from an asset to a diligence question: what did these people actually do, and would they say the same thing on a call with me?
Scrutiny of physician-industry relationships is sharpening generally. A JAMA study of physician social media endorsements found that 100 percent of endorsing physicians had received industry payments, with disclosure present in only 47 percent of cases. That finding does not concern advisory boards directly, and it changes the climate in which every physician-company relationship is now read.
The combination means the ceremonial advisory board is becoming a liability rather than a credential. Which creates, for the first time, real demand for the alternative.
Disclosure is not the fix
The instinctive response to any physician-industry problem is more disclosure. It is worth explaining why that does not solve this one.
Open Payments already discloses the money, comprehensively and publicly. It has done so for over a decade. Program year 2025 recorded $14.67 billion in payments across roughly 667,898 physicians with records.
And the problem persists undiminished, because knowing that Dr. Chen received $12,000 from a company tells you nothing about whether Dr. Chen did substantive work for it. The payment is disclosed. The work is not.
Disclosure records the transaction. What is missing is attestation of the contribution.
That distinction is the whole design insight here. We have built an elaborate national infrastructure to record who paid whom, and nothing at all to record what was actually delivered in exchange. For a market where the product being purchased is professional judgment, that is precisely backwards.
What would work
Scoped engagements rather than titles. An advisory relationship defined by hours, deliverables, and compensation form at the outset. "Clinical advisor" is a status. "Reviewed the clinical workflow specification, twelve hours, over two months, compensated in equity" is a fact.
Peer-visible completion records. A record that the engagement happened and what was delivered, visible to the parties who need to evaluate it. This does two things at once: it lets an honest advisor prove their work, and it gives a founder something citable that means something.
A verified badge only for attested engagements. A company should be able to say "clinically advised" only where the engagement is documented and the advisor attests to it. That single rule would deflate the theater immediately, because the ceremonial roles cannot survive the requirement.
Peer attestation, not payment history, as the discovery signal. The question that identifies a good advisor is not "who has industry relationships" but "who has actually done this work well, according to people who worked with them." That information exists among peers and circulates nowhere.
Reputational cost for non-delivery. Currently an advisor who never shows up faces no consequence, which is why the equilibrium is what it is. Any visible record of delivery creates the missing incentive.
And a serious look at the underused supply. Retired and semi-retired clinicians are, for this specific purpose, the highest-quality available advisor pool and among the least used. They have deep operational experience, they are typically free of employer exclusivity constraints, they have time, and they are frequently more interested in the work being good than in the title. Almost nobody is systematically recruiting them.
What a genuinely useful advisory relationship looks like
It is easy to criticize the ceremonial version and harder to describe the real one, so it is worth being concrete, because good clinical advisory work exists and is enormously valuable when it happens.
A clinician doing real advisory work is not reviewing a deck. They are doing some combination of the following, and the difference is visible immediately to anyone who has been in the room.
They describe the workflow as it actually is, not as it is supposed to be. The most valuable thing a practising clinician knows is where the official process and the real process diverge: what gets skipped when the department is busy, which screen everyone ignores, what the workaround is that nobody documented. Product teams building for healthcare are systematically blind to this, because it appears in no specification and no site visit.
They identify the failure mode before it is expensive. An experienced clinician can usually say, within twenty minutes of a demonstration, exactly which part of a product will be abandoned in month four and why. That single observation is worth more than a year of user research and it is exactly the observation a ceremonial advisor never makes, because making it requires having actually thought about the thing.
They tell you who will resist and whether the resistance is correct. Every clinical implementation encounters pushback. Distinguishing obstruction from a legitimate safety objection is a judgment call requiring real experience, and getting it wrong in either direction is fatal.
They say no. The most useful advisor is frequently the one who tells a founder that the problem they have chosen is not a real problem, or is real but not solvable in the way proposed. That advice is the highest-value output available and is structurally discouraged by an arrangement where the advisor holds equity in the company they are advising.
That last point deserves a moment. Equity compensation quietly aligns the advisor with the company's valuation rather than with the truth. It is a reasonable instrument for a genuine long-term contributor and a poor one for the specific function of honest early assessment. Cash for scoped work produces different incentives than equity for a name, and the difference shows up in what the advisor is willing to say.
The supply nobody is recruiting
There is a substantial and largely untouched pool of exactly the people this market needs.
Retired and semi-retired clinicians have, for advisory purposes, an unusual combination of properties. They have decades of operational experience across multiple institutions and multiple technology generations. They are typically free of employer exclusivity constraints and institutional conflict-of-interest restrictions that complicate advisory work for actively employed academics. They have time, which the heavily published, heavily committed top of the payments distribution emphatically does not. And they are frequently motivated more by the work being good than by the title, which is precisely the disposition that produces the honest no.
They are also, in the terms of this article, the population most invisible to every existing discovery channel. Their publication activity has stopped. Their industry payment records have gone quiet. Their institutional affiliation, which is what a founder scrapes and an investor recognizes, has lapsed.
The discovery mechanisms in this market select against the people best suited to do the work, and nobody has built the alternative.
What you can do now
If you are a founder
Define the engagement before you offer the equity. Hours, deliverables, cadence, and what happens if it does not work. This one change filters for people who intend to do the work and gives you recourse when they do not.
Recruit for the workflow, not the credential. The person who has actually run the clinic your product must fit into will give you more decisive feedback than a nationally known name who will attend one call. Community clinicians are undervalued precisely because no discovery channel surfaces them.
Ask what they will actually commit to. A clinician who says "I can do two hours a month and I will genuinely read the material" is worth more than one who agrees to everything without hesitation.
Do not use the slide as evidence you have not earned. Sophisticated buyers now ask what advisors did. An unanswerable question is worse than a shorter slide.
If you are a physician considering an advisory role
Ask what is expected, in hours and deliverables. If the answer is vague, the role is probably ceremonial, which is fine if you know that going in and are compensated accordingly.
Keep your own record of what you delivered. Nobody else is keeping it, and it is the only way to distinguish your substantive roles from your nominal ones when it matters.
Understand what your name is being used for. Your affiliation is doing work in fundraising and sales conversations you will never see. That is legitimate when you have contributed and is worth being deliberate about.
Check your Open Payments record. Most physicians have never looked. Given that 57.1 percent of physicians appear in it, yours is probably there, and knowing what it says is basic professional hygiene.
Ask about liability and disclosure obligations. Advisory roles carry disclosure requirements in academic settings, in publications, and in some clinical contexts.
If you are an investor or a buyer
Ask the advisors directly. A fifteen-minute call with a named advisor tells you more about a company than the entire deck. Ask what they reviewed, what they recommended, and whether it was adopted.
Treat an unverifiable advisory board as a neutral signal, not a positive one. Given the prevalence of ceremonial arrangements, names alone should move your assessment approximately not at all.
Ask whether any advisor has used the product in their own practice. The answer is frequently no, and it is the single most revealing question available.
Frequently asked questions
What do clinical advisory board members actually do? It varies enormously and is rarely documented. Physician-facing commentary describes advisory roles as typically involving a few hours per quarter, usually compensated in equity, commonly 0.1 to 0.5 percent vesting over two years. Because no record of delivered work exists, substantive and ceremonial roles are indistinguishable from outside.
How many physicians receive industry payments? A JAMA analysis found 826,313 of 1,445,944 US physicians, or 57.1 percent, received industry payments between 2013 and 2022, totaling $12.13 billion, with a median payment of just $48. The distribution is extreme: the top 0.1 percent averaged $1,987,862, and physicians receiving over $50,000 represented 3.4 percent of recipients but 82 percent of total value.
Is Open Payments useful for finding clinical advisors? Only weakly, and it is frequently misused for this purpose. With more than half of US physicians appearing in the database and a median payment of $48, presence carries almost no signal. The high end of the distribution identifies people who are already heavily committed rather than people best suited to advise a specific product.
How are key opinion leaders selected? Typically by publications, trial participation, conference presence, and payment history, often through commercial KOL-mapping vendors. Commentary in the medical literature has noted that selection tends to reflect prescribing patterns, organizational memberships, and guideline involvement rather than depth of relevant knowledge.
Should physicians accept equity for advisory roles? That is a personal decision involving compensation, disclosure obligations, and potential conflicts, and it warrants professional advice. The practical guidance is to establish the scope in hours and deliverables before accepting, keep your own record of what you delivered, understand how your name will be used, and check your disclosure obligations in academic and clinical contexts.
How can a startup find a genuinely useful clinical advisor? Recruit for the specific workflow the product must fit into rather than for institutional prestige, define the engagement concretely before offering compensation, and prioritize clinicians who have operated in settings comparable to your target customer. Retired and semi-retired clinicians are an unusually underused source of deep operational experience with fewer competing commitments.
The bottom line
Healthcare has built a comprehensive public database recording which physicians received money from industry, covering more than half the profession, at a median of $48.
It has built nothing at all recording what any of them did in exchange.
So the market for clinical expertise runs on two signals, publications and payments, both of which measure something other than the thing being bought, and both of which point toward the small group of people who are already the most committed and least available.
The result is a slide in a pitch deck, with five faces on it, that means whatever the reader wants it to mean.
The physicians most capable of telling a founder what will actually happen when their product meets a real clinic are largely invisible to every discovery channel that exists. And the advisor who genuinely did the work has no way to prove it, which means their contribution is worth exactly the same, to an outside observer, as a headshot and a name.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Coverage Exchange That Agencies Exist to Prevent
Evidence note: payment distribution figures come from JAMA analyses of CMS Open Payments data covering 2013 to 2022 and 2014 to 2018, and from CMS Open Payments program year 2025 data. Advisory role structure and compensation norms come from physician-facing commentary and industry sources and represent typical rather than universal arrangements. The social media endorsement finding comes from JAMA (2024). This article describes market structure and does not allege impropriety by any individual or organization; disclosed industry relationships are lawful and frequently reflect legitimate work.
