An analyst at a healthcare-focused investment fund needs to know how a new injectable is actually being prescribed three months after launch, not what the label says, what the doctors on the line at the hospital say.
She opens her firm's expert network portal and posts the request: cardiologist, prescribing this drug class, ideally at a mid-sized community hospital rather than an academic center where prescribing patterns skew unrepresentative. Within four hours, the network's recruiting team has emailed its physician panel, and a cardiologist has replied, passed the compliance screener, and been booked for a call the next morning.
The physician on that call may be exactly the right person. She may also be someone who checked her email first, has an unusually open Tuesday, and prescribes this drug class rarely enough that her impressions are closer to guesswork dressed as clinical experience. Nothing in the recruiting process distinguishes between the two. The screener asks whether she treats patients in this therapeutic area. It does not ask how many, how recently, or whether her institution even carries the drug on formulary.
The fund pays roughly $1,300 for the hour. The cardiologist is paid somewhere between $200 and $500 of it. Neither the price the fund paid nor the answer it received had much to do with whether this particular physician actually knew the answer better than the dozens of others who did not reply in time.
The product being sold is not expertise. It is a hastily verified, promptly available physician, and the fee reflects speed, not certainty.
The industry is large, growing, and heavily healthcare-weighted
Start with the scale, because this is not a boutique curiosity at the edge of medicine.
The expert network industry reached roughly $3 billion in 2025, growing about 11 percent that year and about 16 percent compounded annually over the prior decade, serving roughly 11,200 client firms, according to industry analysis from Inex One. Capital markets account for 42 percent of spend, with consulting close to half of the remainder, and healthcare is consistently one of the industry's densest verticals, meaning a meaningful share of what institutional capital and strategy consultants believe about how medicine is actually practiced is formed in one-hour calls booked by response speed.
Client adoption has been accelerating, not plateauing. The client base grew roughly 150 percent between 2022 and 2025, pulling in a wider set of corporates and smaller funds, many with less mature compliance infrastructure than the large asset managers that built the industry's early norms.
The spread is the business model, and it is enormous
The pricing gap between what a buyer pays and what a physician receives is not incidental to how expert networks operate. It is the mechanism their profitability runs on.
Clients typically pay $1,000 to $1,400 per hour for an expert call, while the expert is paid roughly $200 to $500 per hour, with only the highest-demand healthcare key opinion leaders and C-suite figures clearing $1,000 themselves, per Inex One's 2026 pricing analysis. On a physician hour billed at $1,300 and paid at $300, the network retains roughly $1,000, a spread of more than four times what the physician receives.
Traditional networks compound this further through credit systems requiring $50,000 to $100,000 in annual prepaid client commitments, structures the same analysis describes as leaving "the expert network at liberty of deciding which experts are premium." The physician has essentially no visibility into, and no negotiating leverage over, either the price the client paid or the criteria used to decide her own rate.
Why speed beats expertise as the selection variable
This is the mechanism worth understanding carefully, because it explains why the spread exists and why it is unlikely to close on its own.
Matching quality, whether the physician who takes the call actually has deep, current, hands-on knowledge of the question being asked, is essentially unobservable to both sides at the moment of booking. The buyer cannot verify in advance whether a given cardiologist's prescribing volume, practice setting, and recency of experience genuinely qualify her to answer a specific question about a specific drug in a specific patient population. The physician cannot see what the buyer is willing to pay for a better match, or whether a more qualified colleague exists who simply has not been asked.
In a market where the thing that actually matters cannot be observed, the market optimizes for the thing that can be: speed of response. A network's revenue depends on filling the call, not on finding the best possible physician for it, because the fee is charged per completed call regardless of how well the physician actually answered the underlying question. Physicians self-attest their expertise on a short screener. Practice volume, whether they have actually used the specific product in question, and whether their own institution even purchased it, go unverified against any external source.
The result is a market that reliably delivers an answered call, and only inconsistently delivers an expert answer. Buyers have adapted to this reality in a way that itself reveals the underlying problem: they routinely book three or four calls on the same question, hedging against the likelihood that any single expert may not actually know much, rather than trusting the network's matching to find the right person the first time.
The physicians who take these calls are disproportionately not the ones who should
There is a selection effect on the supply side that compounds the selection effect on the demand side, and it deserves to be stated plainly.
The physicians most likely to reply first to a mass recruiting email and clear a screener quickly are not systematically the physicians with the deepest, most current expertise in a narrow clinical question. Early-career and community physicians, who often have more unstructured time and fewer competing high-value consulting opportunities than senior academic key opinion leaders, take a disproportionate share of these calls. That is not inherently a quality problem; a community clinician's frontline, high-volume experience with a drug or device is frequently more valuable to a buyer trying to understand real-world use than a low-volume academic specialist's. But the current screening process cannot tell the difference between a community physician with genuinely deep, high-volume experience and one who simply has an open afternoon, which means the market cannot reliably reward the former over the latter even when it would benefit from doing so.
The compliance failure is the same failure as the quality failure
This is the structural diagnosis, and the connection between the two problems, mismatched expertise and regulatory risk, is tighter than it first appears.
The SEC's case against Dr. Yves Benhamou is the clearest illustration on record. Benhamou sat on the steering committee overseeing a Phase 3 clinical trial while simultaneously consulting for a hedge fund portfolio manager through an expert network arrangement. He was charged with tipping the fund to trial problems before that information became public; the fund avoided at least $30 million in losses, and the stock fell 44 percent on the eventual public announcement.
The case is usually told as a story about insider trading enforcement. It is also a story about a matching failure. A screening process rigorous enough to verify that a physician does not hold a blinded, non-public role directly relevant to the question being asked would have caught this before the call was ever booked, not after regulators reconstructed it years later. The same shallow, self-attested screening that books a peripheral community physician on a question she cannot really answer is the screening that failed to flag a steering-committee member with material non-public information. Bad calls happen when the person on the line does not actually have legitimate expertise and reaches for something they should not share to justify the fee. Verifying expertise properly is not a separate initiative from managing compliance risk; it is the same fix applied to both problems at once.
Why nobody inside the current industry will fix this
Run through the incumbents and the reason each one is structurally unable to close the gap.
GLG, AlphaSights, and Guidepoint cannot publish the spread between what clients pay and what physicians receive without destroying the margin their entire business model depends on. Nor can they verify expertise deeply, checking actual prescribing volume or product exposure against external data, without slowing down the fill rate their revenue is built on filling quickly.
Doximity's revenue comes from pharma and hospital advertising relationships, not from selling its members' expert hours at transparent rates; there is no business reason for it to enter this market and undercut networks that may themselves be Doximity advertisers.
Sermo's anonymity model, useful for candid clinical discussion, is fundamentally incompatible with a named, verified research call, where the buyer needs to know exactly who they are speaking with and why that person is credible.
A specialty society entering securities-adjacent brokerage would take on compliance risk, and the appearance of endorsing paid access to its members, that it has no organizational reason to accept.
The structural feature that keeps this market inefficient, the buyer cannot verify the match and the physician cannot see the price, is also the feature that makes it profitable for the intermediaries who currently run it. There is no incumbent whose business model improves by fixing this.
AI has raised the value of the specific, hands-on answer
One shift is already reshaping demand inside this market, and it points toward what a better version would need to prioritize.
Generative AI has substantially commoditized the generic expert call, the one asking a broad question a well-sourced literature synthesis could answer nearly as well. What AI cannot replicate is verified, current, hands-on operational knowledge: what a specific hospital actually pays for a specific device this quarter, how a specific AI clinical tool is genuinely being used by staff three months after go-live, what the real barriers were to a specific drug's adoption at a specific type of institution. That is precisely the kind of knowledge the current screener-based matching process is worst at finding, because it requires verifying practice specifics that a self-reported questionnaire cannot capture. The market's own evolution is pushing it toward exactly the problem this article describes: the premium is shifting toward verified specificity at the same moment the matching process remains built for speed.
What would actually work
Verify practice parameters before booking, not after. Practice setting, procedure or prescribing volume, and recency of hands-on exposure to the specific product or question are checkable facts, not self-attestations. A verification layer that confirms these before a call is booked would eliminate the majority of mismatches that currently drive buyers to book redundant calls.
Publish both sides of the price. A buyer who can see that a call costs $1,300 and the physician receives $300 has real information about what they are actually paying for intermediation versus expertise, and a physician who can see the client price has the information needed to negotiate rather than accept whatever rate she is quoted.
Screen for conflicts and institutional role at intake, not just compliance training after booking. The Benhamou case is the template for exactly what needs to be blocked before a call is scheduled: any role on a trial steering committee, access to blinded data, or an employer-confidential pricing or contract relationship relevant to the question.
Let peer-visible quality ratings accumulate. A buyer's assessment of whether a physician actually knew the subject, visible to future buyers and to the physician herself, creates the accountability that a one-off, anonymous-to-history call currently lacks entirely.
Route by verified match quality, not response speed. The core fix is structural: change what the market rewards from "who replied to the screener fastest" to "who actually has the deepest, most current, most specific relevant experience," which requires the network to profit from quality rather than volume of completed calls.
Keep hard guardrails explicit and enforced at intake. No material non-public information, no employer-confidential pricing or contract terms, no patient data, no blinded trial results, and no participation from physicians whose institutional or trial roles make the call improper. These cannot be terms-of-service boilerplate; they have to be checked before the call is booked, the way the Benhamou case shows they were not.
What you can do now
If you are a physician considering expert network calls
Check your employer's policy before you accept your first call. Academic medical centers and large systems increasingly require disclosure or pre-approval for paid external consulting, and many physicians take these calls without checking, which is a real professional and, in some cases, contractual exposure.
Ask what the client is actually paying before you accept a rate. You are very likely being offered a fraction of what the buyer is charged, and simply knowing the approximate spread (client rates commonly run $1,000 to $1,400 an hour against physician pay of $200 to $500) gives you a real basis to negotiate rather than accept the first number offered.
Decline any question that touches non-public trial data, blinded results, or your institution's confidential contract terms, without exception. The Benhamou case did not begin with a physician planning to break the law; it began with a physician in a role that made a seemingly ordinary consulting relationship improper from the start.
Recognize when you are not the right person for the call, and say so. If your practice volume or recency of experience with the specific question is thin, saying so protects both your professional credibility and the buyer from a call that will mislead more than it informs.
If you buy expert network access
Ask your network what verification, beyond a self-reported screener, was performed before this expert was booked. If the honest answer is "none beyond a questionnaire," treat the resulting call as one data point among several you will need, not a verified fact.
Stop treating redundant bookings as a quality safeguard and start treating them as a symptom. Booking three or four calls on the same question to compensate for unreliable matching is a rational response to a broken market, but it is also direct evidence that the underlying matching process is not working, and it is worth naming that to your own compliance and research leadership.
Push for transparent pricing from your provider. A network that can tell you exactly what the physician was paid, and why this specific physician was matched to your question, is giving you real information about match quality; one that cannot is asking you to trust a process it will not show you.
If you build compliance or research infrastructure
Treat expertise verification and compliance screening as the same system, not two separate ones. The evidence in this article suggests the calls most likely to go wrong on compliance grounds are frequently also the calls where the physician's actual expertise was never properly checked; a verification layer built to solve one problem will substantially help the other.
Frequently asked questions
How much do expert networks pay physicians per hour? Roughly $200 to $500 per hour, while clients typically pay $1,000 to $1,400 per hour for the same call, according to 2026 pricing analysis from Inex One. Only the highest-demand healthcare key opinion leaders and C-suite figures clear $1,000 themselves.
What is an expert network call? A paid, typically one-hour phone consultation arranged by an intermediary firm (such as GLG, AlphaSights, or Guidepoint) connecting an investor, consultant, or corporate client with a professional, in this case a physician, who has relevant knowledge or experience on a specific business or clinical question.
Are expert network calls legal for physicians to take? Generally yes, provided the physician does not share material non-public information, blinded trial data, or employer-confidential terms, and complies with their employer's consulting policies. The SEC's case against Dr. Yves Benhamou, a trial steering-committee member who tipped a hedge fund to unpublished trial problems, resulting in at least $30 million in avoided losses, illustrates exactly where the legal line sits.
Do physicians need employer permission to do expert network consulting? Often yes, though policies vary widely. Many academic medical centers and large health systems require disclosure or pre-approval for paid external consulting, and a substantial number of physicians reportedly take these calls without checking their institution's policy first.
How do expert networks vet their physician experts? Primarily through a short self-reported screener asking about specialty and general practice area, with limited to no independent verification of actual practice volume, product-specific experience, or institutional role, according to industry pricing and process analysis from Inex One.
Why is there such a large gap between what clients pay and what physicians are paid for expert calls? Because matching quality is difficult for either side to observe in advance, so the market optimizes for speed of response rather than verified expertise, and the intermediary's profit comes from the spread between the two. On a $1,300 client-paid hour and a $300 physician payout, the network retains roughly $1,000, or more than four times what the physician receives.
The bottom line
The expert network industry has grown to roughly $3 billion, expanding about 150 percent in client base since 2022, on the strength of a product it markets as expertise and delivers, reliably, as availability. A cardiologist paid $300 for an hour a client paid $1,300 for was selected because she replied to a screener fastest, not because anyone verified she was the physician who actually knew the answer.
That gap is not simply an unfair split of fees. It is the direct cause of two separate failures buyers and regulators both experience: buyers who cannot trust a single call and compensate by booking three or four redundant ones, and a compliance record, most starkly the Benhamou case and its $30 million in avoided losses, that traces back to the same shallow, self-attested screening that mismatches every ordinary call.
Nobody currently profits from fixing this. The intermediaries that dominate the market make their margin from the spread and the speed, not from verified quality, and every credible alternative, Doximity, Sermo, a specialty society, has a structural reason not to enter. AI has, if anything, sharpened the case for a fix, by commoditizing the generic call and pushing the market's remaining value toward exactly the verified, specific, hands-on knowledge that a self-reported screener cannot find.
That analyst is going to keep booking calls with whoever answers first, because nothing currently tells her, or the cardiologist on the other end of the line, that a better match was available. The industry calls the result an expert network. What it has actually built is an answered network, and the difference is worth roughly a thousand dollars an hour to someone.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Department of One
Evidence note: market size, growth, and client concentration figures ($3 billion in 2025, 11 percent 2025 growth, 16 percent decade compound growth, roughly 11,200 client firms, 42 percent capital markets share, 150 percent client base growth 2022 to 2025) are from Inex One's 2026 industry analysis. Pricing figures ($1,000 to $1,400 client rate, $200 to $500 physician rate, $50,000 to $100,000 annual credit commitments) are from Inex One's 2026 pricing analysis; these are industry-reported figures from a firm operating in this space, not an independently audited academic study, and should be read accordingly. The Benhamou case details ($30 million in avoided losses, 44 percent stock decline) are from SEC Litigation Release 21721 (2010), a matter of public regulatory record. This article does not name or describe any specific physician's actual expert network experience; the opening scenario is a composite illustration built from the patterns documented in this evidence, not a specific reported case.
