Doximity Fiscal 2Q27 Earnings Call - DOXIMITY INC

Financial Performance:

• Q2 2027 revenue was $174.5 million, up 4% year-over-year (YoY), exceeding the high end of guidance by 2%.
• Adjusted EBITDA was $89 million, representing a 51% margin, beating the high end of guidance by 9%.
• Non-GAAP gross margin was 88% (down from 92% in Q1) due to increased AI investment costs, though cost of revenue growth was less than half the rate of prompt volume growth, indicating improved unit economics.
• Non-GAAP diluted EPS was $0.36; GAAP diluted EPS was $0.18.
• Stock-based compensation (SBC) was $41 million (23% of revenue); excluding the AI R&D grant, SBC was ~19% of revenue.
• Free cash flow generated was $91 million, up significantly from $40 million in Q1.
• Balance sheet is strong with $713 million in cash, cash equivalents, and marketable securities, with no debt.
• Net Revenue Retention (NRR) was 103% overall and 106% for the top 20 customers.

Guidance and Future Outlook:

• Q3 2027 revenue guidance raised to $195–$196 million, representing 6% YoY growth.
• Full-year FY27 revenue guidance raised to $678–$683 million, representing 6% YoY growth (midpoint raised by $4.5 million).
• FY27 Adjusted EBITDA guidance raised to $317–$325 million, representing a 47% margin.
• Q4 revenue growth is expected to align with Q3 (6% YoY), though visibility increases after the "upfront" contracting season.
• Management anticipates slight acceleration in H2 revenue growth vs. H1.
• AI Search revenue is expected to ramp gradually throughout calendar 2027 due to medical-legal review processes.

Business Segments and Product Lines:

• AI Search (Pharma): Entered the market with 8-figure bookings; 16 of top 20 pharma companies have purchased, with 3 in contracting. Revenue recognized from July onward; expected to drive incremental growth in a $14B paid search market.
• AI Accuracy & Tools: Launched "DoseCheck," a deterministic medication safety verifier reducing AI drug error rates from 2.3% to near zero in testing. Integrated "UpToDate" content for clinical decision support.
• Health Systems: AI client footprint grew to 190 health systems (including 10 of top 20) from 165 last quarter. New York Presbyterian and UCSD Health are notable new logos.
• Workflow Engagement: Quarterly active workflow prescribers grew 34% YoY to record highs; approximately 50% of all 1.8M US prescribers used the platform last quarter.
• New Features: Launched "Patient Context," allowing AI to access patient charts for specific answers, reducing prompt length by 50%. API access to clinical intelligence is now open.

Market and Competitive Landscape:

• Competitively leading in AI accuracy: Ranked #1 among 16 leading models in independent testing by Fireworks (following Stanford No Harm study results).
• Pharmaceutical marketing sentiment: 81% of 35 surveyed pharma marketers indicated they would spend incremental budget on Doximity, while only 6% would spend more on Google/Paid Search.
• Doximity is transitioning from a pure media company to a "media + search" model, entering the $14B pharma paid search market.
• UpToDate partnership leverages their 2,500+ health system client base, creating synergies with Doximity's 190 AI health system clients.

Risks and Challenges:

• Medical-Legal Review (MLR): Extended timelines for MLR approval for new AI products may delay revenue recognition and ramp-up (expected to take months).
• Data Privacy & Liability: Increasing health system scrutiny regarding AI data leakage and patient privacy (e.g., Johns Hopkins incident) creates a barrier to entry for unapproved AI tools, though Doximity is positioned as a secure, approved alternative.
• AI Cost Efficiency: While improving, AI costs are scaling with engagement; maintaining margins while investing heavily in accuracy and compute remains a focus.
• Market Volatility: Previous quarters faced "perfect storm" market headwinds; recovery is viewed as stable but not yet dramatically accelerating.

Management Commentary and Tone:

• Tone is optimistic and confident, emphasizing "cautious optimism" regarding future growth and the "re-acceleration" expected in H2 FY27.
• Management framed AI as a "generational technology shift" where Doximity is "playing offense" with strong moats in accuracy and distribution.
• CEO Jeff Tangney highlighted that engagement growth (50% prescriber penetration) directly correlates with future monetization opportunities.
• CFO Matt Sonfeld expressed confidence in the business model's ability to generate high incremental margins despite AI investments.

Analyst Questions and Answers:

• AI Search Cannibalization: Management stated AI Search is primarily incremental, as it targets separate search budgets within pharma companies (different from core marketing tools) and captures the $14B paid search market.
• AI Compute Efficiency: Efficiency improved significantly; cost of revenue grew at half the rate of prompt volume in Q2, up from a 1:1 ratio in Q1, driven by multi-model usage and proprietary harness gains.
• UpToDate Partnership: Described as a strategic, multi-year partnership; UpToDate is the most trusted point-of-care tool, and the collaboration brings proprietary expert content not available in open-source models.
• NRR Trends: Overall NRR dipped to 103% (from 106% top 20), attributed to seasonal weakness in Q2 upsell activity; expected to re-accelerate with AI Search and core product growth.
• Monetization Model: AI Search monetization is primarily advertising-driven (pharma side); health system AI is sold as a nominal fee within the broader workflow suite.

Other Key Points:

• Doximity is opening API access to its clinical intelligence layer to third-party developers.
• Share repurchases totaled $54 million in Q2, with $347 million remaining in the program.
• The Medical AI Summit convened 150 physicians and two former Surgeons General to guide responsible AI development.
• Doximity is entering a "post-Wild West" phase in AI governance, where health systems are actively blocking unapproved AI tools, favoring secure partners like Doximity.
• Patient Context feature is already driving usage doubles by reducing the effort required for doctors to query specific patient data.