OpenEvidence Has 650,000 Doctors. Now It Needs the Hospitals.
OpenEvidence Has 650,000 Doctors. Now It Needs the Hospitals.
Four years ago, Daniel Nadler built a chatbot that answered medical questions by scanning the literature, gave it away free to any physician with an NPI number, and waited. The wait turned out to be short. OpenEvidence now claims 650,000 active U.S. physicians using it across 27 million clinical encounters in April alone, a number that implies roughly 65% of all American doctors have tried it. That bottom-up wave lifted the company to a $12 billion valuation and $150 million in annualized revenue, up 1,803% from $7.9 million in 2024. Investors including GV, Sequoia, Kleiner Perkins, and Nvidia have poured nearly $700 million into it. STAT News Sacra
This week at the STAT Breakthrough Summit West in San Francisco, Nadler's CTO Zachary Ziegler is making the pitch that comes next: formal contracts with the health systems those doctors already work inside. Sutter Health launched OpenEvidence inside its Epic workflows in February. Mount Sinai went enterprise-wide in March, embedding it across seven hospitals for physicians, nurses, and pharmacists. These are the first publicly announced hospital deployments of this kind. The pitch is no longer "try us free." It is "now sign the contract." Sacra Mount Sinai
Here is the problem Nadler has built for himself. OpenEvidence's entire competitive advantage is that it never needed a hospital's permission to exist. It grew around the procurement committee, past the IT review, beneath the vendor contract. Doctors adopted it because it was faster and better than what their institutions had approved. That bypass is precisely what made it worth $12 billion. And now it needs those same gatekeepers to become paying customers.
The tension is not theoretical. OpenEvidence's revenue comes primarily from pharmaceutical advertising, CPMs reportedly ranging from $70 to over $1,000 per thousand impressions, multiples above typical digital advertising because the audience is verified, licensed clinicians in clinical moments. That model rewards reach and frequency with individual doctors. Hospital contracts are different: they are negotiated, volume-based, often require liability indemnification, and come with procurement cycles that can stretch 18 months or longer. The product that grew by never asking permission has to learn how to ask. Sacra
Epic is the other variable nobody in the room can ignore. The EHR giant is already inside every major health system OpenEvidence wants to enter, and it has been building its own clinical AI tooling. Ziegler was asked at the Summit whether OpenEvidence is a collaboration with Epic or competition with it. His answer, per STAT's reporting: hospitals are choosing OpenEvidence because doctors prefer it to what Epic offers. That may be true. It is also the exact dynamic that makes Epic a potential systemic choke point rather than a neutral distribution channel. STAT News
The Mount Sinai and Sutter deployments are real. They are also the easiest deals. Both are academic medical centers with research missions and cultures that make their doctors unusually likely to adopt new tools on their own terms. The harder question is what happens when OpenEvidence walks into a community hospital network, a rural system with constrained IT budgets, or a health system where Epic's business development team has an existing relationship and pricing leverage. The doctors may already use OpenEvidence. The contract still has to go through the people who approved Cerner's replacement in 2019.
Nadler appears to understand this is not just a sales problem. In recent comments to NBC News, he signaled that the ad-supported model that got the company here may not be the model that gets it to the next tier. "We're not crazy monsters," Ziegler reportedly told the STAT Summit audience, addressing the hospital audience directly. The implicit question underneath that reassurance is what happens to the doctors when the business model changes. Higher CPMs are available in healthcare advertising precisely because the context is clinical and the decisions being influenced carry life-and-death stakes. Whether that market survives a transition to institutional pricing, or whether it has to shrink because hospital legal teams won't accept an ad-supported tool inside their Epic workflow, is the unanswered question at the center of OpenEvidence's next chapter. NBC News STAT News
The $12 billion bet is that the doctors' preference is strong enough to force the institution's hand. That has been true so far. Whether it remains true when the contract shows up for signature is the thing Nadler needs to prove before the next fundraising round, the next valuation reset, or the next competitor that figures out the same bypass.
OpenEvidence is not the first company to discover that disruption and revenue are different sports. It may be the most expensive test case in clinical AI of whether winning the doctors first actually translates into leverage with the people who sign the checks.