Doximity (NYSE:DOCS): The Pharma Platform I’m Still Sitting With — But Not Comfortably
I’ve mentioned Doximity a few times in my monthly reviews, usually filed under “healthcare drag.” Never gave it a proper post. After digging into it properly over the past few weeks, I think it’s time — partly because the thesis has gotten complicated, and partly because I still haven’t fully made up my mind what to do with it.
This one is me thinking out loud.
What Doximity actually is
Most people call it “LinkedIn for doctors.” That’s not wrong but it misses the point.
The business model is simple once you see it through the right lens. Think PropertyGuru (NYSE: PGRU, delisted Dec 2024) or Beike in China (NYSE: BEKE) — platforms that own the verified professional inventory, where the paying customer is whoever wants access to that audience. In Doximity’s case, the inventory is 85% of US physicians (over 3 million of them), and the paying customer is pharma. Physicians use the platform for free — Health Insurance Portability and Accountability Act (HIPAA)-compliant messaging, digital fax, a telehealth dialer, and more. Pharma pays to reach them through targeted ads while they’re reading medical news or checking drug references. Around 80% of Doximity’s revenue comes from that single relationship.
When you own 85% of a professional audience nobody else can replicate, the market prices you very generously. That’s how you get to $70+ per share.
The question is what happens when that audience’s attention starts going elsewhere — and when the pharma companies paying for it have reasons to pause.
Why it’s at $24
DOCS has fallen roughly 68% from its peak. Three things hit at the same time.
Pharma froze. The US government introduced rules requiring drug manufacturers to charge Americans no more than the lowest price they offer any other country. When you’re a pharma CFO suddenly modelling a potential 20-40% revenue haircut, the first thing you pause is discretionary ad spend. That’s Doximity’s entire revenue model. Q4 FY2026 guidance came in at roughly 4% revenue growth — down from 10-17% in prior quarters. Stock fell over 30% after hours.
The good news here: this looks temporary. Eli Lilly’s Q1 2026 earnings call this week named physician-targeted digital campaigns as their primary launch strategy for a major new drug, with television advertising not starting until Q3. Pharma isn’t retreating — they’re redirecting. And pharma budgets are moving from direct-to-consumer (DTC) TV toward healthcare provider (HCP) channels precisely because the Food and Drug Administration (FDA) is cracking down on consumer drug advertising. That DTC-to-HCP shift actually favours Doximity structurally.
The Electronic Health Records (EHR) threat. Think about when a doctor is most likely to prescribe a drug — when a patient is sitting right in front of them. Pharma has figured out they’d rather reach doctors at that exact moment, inside the software doctors use to manage patients and write prescriptions, than reach them later while they’re browsing a news feed. OpenEvidence is already doing this inside Epic, the dominant hospital software system in the US. Doximity isn’t. Their doctors still have to copy-paste the AI-generated notes manually into the system. That tells you everything about where they stand on this.
The competitive squeeze. On April 22, OpenAI launched ChatGPT for Clinicians — free, for all verified US physicians. The person leading it is Nate Gross, Doximity’s former co-founder and Chief Strategy Officer — more on him shortly. OpenEvidence hit $150M+ in annualised revenue and 18 million clinical consultations a month. They just launched an AI-integrated Doctor Dialer — directly attacking Doximity’s core workflow tools. This week also saw Roon launch, a new physician-only social network — founded specifically because doctors don’t use Doximity for community or peer discussion. When someone builds a competing product around your blind spot, that’s not a compliment.
What’s keeping me in
It’s not blind conviction. Here’s what I keep coming back to.
The platform effects
The 85% registration number is real but misleading on its own. The right question is how many are actively engaged in ways pharma pays for — think of it as GMV rather than registered users.
| Quarter | Active Workflow Prescribers | YoY |
|---|---|---|
| Q1 FY2025 (Jun 2024) | 590,000 | — |
| Q1 FY2026 (Jun 2025) | 630,000 | +7% |
| Q2 FY2026 (Sep 2025) | 650,000 | +10% |
| Q3 FY2026 (Dec 2025) | 720,000 | +22% |
Engagement growing 22% year on year is genuinely good. But 720,000 active workflow users out of 3 million registered is a 24% active rate. Pharma pays for the 720,000, not the 3 million.
Doximity’s products
| Product | Proves value? | My read |
|---|---|---|
| DoxGPT | Roadmap, not reality | The Google analogy is right — pharma paying to appear alongside physician drug queries is genuinely high-value. 300,000 monthly users. Zero revenue. Intended model, not current one. |
| AI Scribe | Stickiness yes, revenue no | Strongest daily habit on the platform. 23% of users already cut after-hours paperwork. Physicians don’t go back to manual note-writing. But zero revenue and notes still get copy-pasted into Epic. |
| PeerCheck | Yes for pharma, weakening for physicians | 10,000 physician reviewers vs OpenAI’s 260. Strong brand safety argument for pharma legal teams. But clinical AI answer quality is converging fast across all platforms. |
| RecruitGPT | Underappreciated | Pharma hiring medical experts through the same platform they advertise on is a clean cross-sell. Doximity’s own case studies cite 3x response rates — self-reported, treat with scepticism. Nobody talks about this. |
| DocDynamic | Promising, unproven | Pharma buying multi-module bundles went from 5% to 45% of bookings in one year. Real traction. But the pharma budget freeze hit right at contract renewal season. May 13 tells us if those deals converted. |
The stickiness within that 720,000 is also uneven. The communication and workflow layer — HIPAA-compliant messaging, secure fax, dialer — is genuinely hard to leave. Colleagues need to switch with you. That coordination cost is real. But the clinical AI behaviour is as elastic as a shopper switching from BABA to PDD. A physician can open ChatGPT for Clinicians in a different tab tomorrow with zero friction. Doximity counts the Scribe user. They don’t see the ChatGPT tab open in the background.
The proprietary physician data is the most underexploited asset here. Doximity knows what 300,000 physicians are querying in real time — which drugs, which specialties, which geographies. OpenAI doesn’t have that. It could be a premium intelligence product sold directly to pharma. Instead it’s being used to give better AI answers for free. Most valuable asset. Least monetised.
The products prove physician stickiness. Whether that earns pharma revenue is a different question — which brings me to the management.
The Management question — and it’s the most important one
The network isn’t the problem. The monetisation layer on top of it is — and that’s entirely on the people running the place.
| Person | Role | Background |
|---|---|---|
| Jeff Tangney | CEO & Co-founder | Founded Epocrates (one of first 5 App Store apps, sold for <$300M). Goldman Sachs healthcare analyst prior. |
| Nate Gross | Co-founder & former CSO | Left June 2025 after 15 years. Now VP Health at OpenAI. Built ChatGPT for Clinicians in 10 months. |
| Lisa Greenbaum | CCO (joined Jan 2024) | Ran Medscape for 15 years. Former CCO at Verily (Alphabet’s health company). |
Jeff Tangney is an exceptional product builder and a poor monetiser under competitive pressure — and the pattern runs across both companies. At Doximity the playbook is identical to Epocrates: build until perfect, then charge. He said AI monetisation was coming “this summer” in February 2024. It’s May 2026.
The Nate question. As co-founder, Nate had equal standing with Jeff — not a hired executive who could be overruled. That means the monetisation delay wasn’t Jeff blocking Nate. Both founders agreed to defer it. Which makes his departure harder to read — if they were aligned for 15 years, why did he leave and immediately build the competing product at OpenAI in 10 months? Worth noting: Doximity is headquartered in San Francisco. California has banned non-compete clauses since 1872. There was no legal barrier stopping Nate from walking out and joining a competitor the very next day — which is exactly what happened.
Lisa joining 17 months before Nate left — to me, that signals a planned transition, not a surprise exit. As a listed company you engineer co-founder departures, not scramble through them. Which leans toward opportunism: Nate saw the OpenAI opportunity, set up the handover properly, then left to build what he’d always wanted to build with better resources.
But the 10-month delivery is still the tell. You don’t ship a fully launched, HIPAA-compliant, physician-verified clinical AI in 10 months unless you already knew exactly what needed to be built. Nate didn’t need to figure out the product. He just needed the authority to ship it.
Lisa has the right background and the DocDynamic bundling going from 5% to 45% of bookings in a year probably has her fingerprints on it. But she has no co-founder standing and cannot force Jeff’s hand. She can design the commercial model. She cannot ship it without his decision. Sixteen months in, the commercial AI product still has no name, no price, and no launch date.
Where I land
Doximity is not a broken business. The physician network is real, the margins are real, the data is irreplaceable. But the business model is being challenged from multiple directions and the management team is thinner than it’s ever been.
The question is whether DOCS becomes LinkedIn — which finally monetised a dominant professional network — or WebMD, which had dominant physician traffic, never found a second revenue engine, and slowly faded. None of the competitors doing the disrupting are listed. OpenAI, OpenEvidence, Sermo, Roon — all private. For retail investors, DOCS is the only listed stock with direct exposure to this physician platform space. Which makes the May 13 earnings call the only near-term decision point that matters.
Did the delayed DocDynamic deals actually convert — or did they slip into FY2027, which would make the net revenue retention (NRR) decline look structural rather than cyclical. And does Tangney finally name a commercial AI product with a price and a launch date. I need both. One without the other isn’t enough.
I don’t have a clean answer yet. The earnings call will help me get there.
It is not financial advice. The author may hold positions in securities discussed. Readers should conduct their own due diligence and consult with a qualified financial advisor before making investment decisions.