Doximity Wants to Be Paid Search, Not Pharma Ads
A 54% seventeen-week rally already priced the AI-search pivot — into a quarter whose free-cash-flow margin fell to 3%.
DOCS · Earnings Call · 2026-11-05
A rally that front-ran the print
The tape tells the story before the transcript does. Doximity has climbed roughly 54% over the last 17 weeks, finishing the session before this call within 4.3% of its high — a name being re-underwritten in real time, even though it still sits 67.5% below its 2021 peak of $102. The quarter itself was fine, not spectacular: revenue of $174.5M, up 4% and 2% above the high end of guidance, on an adjusted EBITDA margin of 51%. “Even with over 6x growth year on year in our AI spend, our Q2 adjusted EBITDA margin was 51%, or $89 million, and 9% above the high end of our guidance.” — Jeff Tangney, Co-Founder and Chief Executive Officer · 2026-11-05 The gap between a 4% revenue print and a 54% seventeen-week rally is the entire dossier.The $14 billion question
The bull case is a reclassification. Doximity spent fifteen years as a pharma advertising network; it now wants to be paid search. “It's about $14 billion a year in the United States according to eMarketer just for pharma. So $14 billion a year spent on paid search.” — Jeff Tangney, Co-Founder and Chief Executive Officer · 2026-11-05 That figure is the linchpin of a whole new paid search narrative, and the newest phrase to spike in the company's own keyword set is literally buy AI search. AI Search is now an eight-figure bookings product with 16 of the top 20 pharma companies having purchased and three more in contracting; a biddable/auction model follows next year — the LinkedIn playbook the new CFO keeps invoking. Jeff frames the product as a workflow grab: “They want a suite, not a search bar.” — Jeff Tangney, Co-Founder and Chief Executive Officer · 2026-11-05 Be honest about scale though. Eight-figure bookings against a $678–683M revenue base. Three quarters ago management said there was “no revenue in our forecast for our AI products right now” — Jeffrey Tangney, Co-Founder and CEO · 2026-02-05 (Feb 2026). Today the CFO warns that the medical-legal review can take “months and months” — Jeff Tangney, Co-Founder and Chief Executive Officer · 2026-11-05 — the same caution Perry gave in May about “losing the summer for most of these programs going live” — Perry Gold, Vice President · 2026-05-14 for the prior vertical-video cycle. The $14B is a call option, not a forecast.What management stopped saying
The most revealing signal is what dropped out. Last quarter's top keywords were Clinical AI, drug reference, the NOHARM study and Doximity Ask — a proof-of-accuracy story. This quarter, "study," "Pathway acquisition" and "Scribe" all appear among the biggest declining keywords, and the fresh language is patient context, dose check, UpToDate and the proprietary harness. That is a shift from "our model answers better" (benchmark studies) to "we are the deployment layer inside the chart." UpToDate — the reference set no frontier model can access — is the moat they now lead with; DoseCheck is the safety story replacing the Stanford benchmark. It matters because it is the harder-to-replicate asset and the one an enterprise AI steering committee can actually approve.The bill for playing offense
The part of the story the rally ignores is the income statement. On the last filed (July-quarter) GAAP numbers, gross margin fell 4.3 points to 84.9%, operating margin dropped roughly 18 points to 25.8% — the deepest compression in years — and free cash flow (less SBC) collapsed 87% to $5M, a 3.3% FCF margin. SBC ran $37M, up 68% year over year, while R&D rose 44% to $38M. Management is explicit this is a choice — “we'll continue to play offense very much so.” — Matt Sonefeldt, Chief Financial Officer · 2026-11-05 The genuinely encouraging datapoint is efficiency: “our cost of revenue grew at less than half of our overall prompt volume in Q2” — Matt Sonefeldt, Chief Financial Officer · 2026-11-05, versus roughly one-to-one last quarter. If that compounds, the margin math flips. Note the broader AI data centers complex has been de-rating for months — Doximity's problem is the opposite: it is a consumer of inference, not a builder, so the market rewards cheaper tokens here even as it punishes capex elsewhere.Retiring the scoreboard
Then there is the quiet part. Net revenue retention slipped to 103% overall and 106% at the top 20; customers above $500k grew just 3% and ticked down sequentially on seasonal weakness. The new CFO's response was to question the KPIs themselves.That is either clear-eyed recalibration or a pre-emptive exit from a fading metric. Either way, investors now must judge the business on bookings and AI-search ramp, not the tidy NRR they had grown used to.I think these metrics may not be the right set of metrics to measure our growth going forward... It misses our small and medium growth segment, which has been one of the faster growing segments for several quarters now.
The bet
Two things are true. The AI search product is real and moving — 16 of 20 pharma, an eight-figure book, a biddable product coming — and it opens a TAM well beyond HCP media. And the current financials are ugly: 4% growth, a collapsed cash margin, and a stock that has already re-rated violently from a price to revenue base of ~6x.Maybe. But Google did not spend its own gross margin building the index while the market paid 54% in seventeen weeks for the option. Watch February: the upfront bookings number and one line — whether gross margin stops falling — will tell you if this is a search company or an ad company paying a search tax.it is a new TAM, and I think it may end up being like what Google did to the yellow pages.