Every Physician Search Produces Intelligence. Most Health Systems Throw It Away.

One physician accepts. Several others say no.

The health system keeps a clean record of the one who signed: source, compensation, start date. The requisition closes. The others are reduced to a few disposition codes, if they are recorded at all:

Not interested. Location. Compensation. Accepted another offer.

That is enough to close a file. It is not enough to explain what happened.

Every physician search should produce one of two assets: a well-aligned hire or better intelligence for the next decision. Too often, health systems get neither.

Many health systems have a conversion problem

Most recruiting systems answer process questions. How many candidates entered the funnel? When did someone withdraw? Was the offer accepted?

Those questions matter, but they describe the health system’s activity more than the physician’s decision.

An applicant tracking system can show that a physician withdrew after a site visit. It usually cannot tell you whether the physician lost confidence in local leadership, learned that call was heavier than expected, or received clearer answers from another employer. It can record that a physician accepted another offer but rarely explain why that offer won.

It also cannot see the physician who reviewed the opportunity, found too little useful information, and decided the conversation was not worth starting.

For many health systems, the challenge is not getting enough names or applications. It is converting initial interest into interviews, offers, acceptances, and durable hires.

That distinction matters because weak conversion is not always a recruiting problem.

A gastroenterology search may generate interest until physicians learn the weekend call structure. More outreach will only send more people through the same failure point.

An orthopedic surgeon may take a lower guarantee elsewhere because the competing practice offered reliable block time, stronger APP support, and a more credible growth plan. Recording the decline as compensation or another offer misses the decision.

Weak conversion can come from slow follow-up, poor communication, missing information, low trust, an uncompetitive role, or a legitimate mismatch. Those are different problems. They should not produce the same response.

Even compensation can mean several things: a below-market guarantee, an unrealistic productivity threshold, a confusing formula, or simply the safest explanation for a concern the physician does not want to state directly.

A disposition code is not an explanation. It is often where the investigation stops.

Why the signal disappears

The information is not lost because recruiters do not care. It is lost because the system is centered on completing the current search.

Recruiters need to fill positions. Hiring leaders need coverage. Outside firms may be paid when a placement closes. Physicians need to protect their options and reputations.

Most health systems have an owner for the requisition, the candidate, the approval, and the budget. Few have an owner for what the market taught them.

Physicians also have little reason to be fully candid. Someone who may want to work with an organization later is unlikely to offer blunt feedback about leadership, culture, or the practice environment.

The remaining pieces scatter across the ATS, recruiter notes, email, agency conversations, and hiring manager memory. The decline reason survives. The context disappears.

A real learning system would connect what the physician cared about before a role was presented, what the job and clinical unit are actually like, where interest strengthened or weakened, what the physician ultimately chose, and whether the eventual hire worked as expected.

The goal is not to interrogate every physician who says no. It is to identify patterns while protecting enough privacy for physicians to be candid.

A health system does not need to know that a named physician disliked a particular leader. It does need to know when leadership concerns repeatedly suppress conversion in the same unit.

Why we built Tessellate

I have an obvious stake in this argument.

At Winnow, the company I previously co-founded, we helped health systems identify passive physicians they could recruit. That work taught me that widening the funnel did not solve the deeper problem. An employer could reach more qualified physicians and remain blind to why the opportunity converted or failed.

That is why I started Tessellate.

Physicians use Tessellate privately to define what matters to them across their work, finances, lifestyle, family, and geography. Employers structure the realities of a specific site and role, including call, support, autonomy, workload, compensation design, and team environment.

We connect those inputs with physician questions, engagement decisions, progress through the hiring process, and known outcomes. Physicians can also bring us opportunities they found elsewhere, so the learning is not limited to roles Tessellate sourced.

A physician’s identity is shared only when that physician chooses to engage. Employers receive aggregated signal without access to the private profiles of physicians who say no. Physicians pay nothing, and Tessellate does not charge placement fees. Employers subscribe for participation, measurement, and intelligence.

That structure avoids an obvious conflict: needing a particular placement to close for the interaction to have value. When a physician and role are not aligned, the system should be able to say so. The no should still teach us something.

Tessellate cannot fix a poor call schedule, unstable leadership, or an uncompetitive practice environment. It can make recurring patterns visible while an organization still has time to act.

Health systems can build parts of this capability internally. The strategic question is whether they have any credible mechanism for learning from the physicians they do not hire.

Five questions for the next recruitment review

1. Do we know why physicians rejected this opportunity, or only what someone entered into the ATS?

2. Can we tell whether weak conversion is caused by the opportunity itself, how it is represented, or how the hiring process is run?

3. Do we know what the physicians we lost chose instead and what determined that choice?

4. Which reasons recur by specialty, facility, and clinical unit?

5. What will we change because of what this search taught us?

If the answer to the last question is nothing, the organization does not have a learning system. It has a transaction log.

Not every physician search should end in a hire. Sometimes the physician and organization are simply wrong for each other. Forcing a weak match can cost far more than leaving a position open a little longer.

But every search should leave the organization with something valuable: a well-aligned hire or better intelligence about role competitiveness, conversion barriers, and what needs to change next.

The true failure is not a search that ends without a hire.

It is a search that leaves the organization no smarter than when it began.

About the author

Paul Vernich is the founder and CEO of Tessellate, a physician-first career and hiring intelligence company. He previously co-founded Winnow, a physician recruitment technology company acquired by Aya Healthcare.

Explore Tessellate: www.t8health.com