Zocdoc CEO: "Dr. Google is going to be replaced by Dr. AI"

October 20, 2025·1 hr 6 min·5 clips

ZocDoc CEO says Dr. Google will be replaced by Dr. AI, forcing patients to judge when to trust chatbots.

Oliver Kras leads ZocDoc, a platform connecting patients with doctors through booking services now expanded beyond the original app to include insurance partnerships, veteran services, and AI phone agents. ZocDoc reduces average wait times from 30 days with traditional phone booking to typically under 24 hours, representing an order-of-magnitude improvement in healthcare access. The company employs over 1,000 people in a functional structure unified around single revenue and profitability metrics following a major business model transformation. ZocDoc shifted from flat fees to per-patient referral pricing, causing some doctors to pay 100 times more but resulting in only one major customer loss due to the platform's irreplaceable quality and volume. Contrary to pandemic-era predictions, patient data shows 95% of appointments remain in-person, with telehealth adoption limited primarily to mental health services. Kras compares telehealth to "telepizza" - useful except when physical presence is necessary for somatic medicine. The company deployed Zoe, an AI assistant handling phone scheduling with 52% conversion rates, comparable to average human agents but below the 65% achieved by top performers. Kras emphasizes AI works through deterministic orchestration layers rather than relying solely on large language models in context windows. ZocDoc uses AI for patient-doctor matching but refuses to provide diagnostic medical advice, anticipating "Dr. Google" will become "Dr. AI" for self-diagnosis while maintaining human judgment for complex cases. Kras identifies three decision-making categories: hiring senior leadership who know their domains better than he does, absorbing risk to encourage smart experimentation, and triangulating technical feasibility with customer needs and regulatory requirements. He rejects concerns that AI agents like Alexa will disintermediate ZocDoc by automated browsing or voice commands. His argument rests on ZocDoc's 20-year edge case expertise creating advantages competitors cannot quickly replicate, comparing it to measuring England's coastline with increasing granularity. The conversation addresses fundamental tensions in how AI agents access services, contrasting protocol-based frameworks like MCP with automated browser navigation approaches favored by companies like Perplexity. Kras suggests current competitive dynamics among five major AI providers give service platforms more negotiating leverage than they had during Google's search monopoly. He welcomes automated traffic that doesn't drain transaction fees, arguing companies like Uber and Airbnb profit from driver/host fees rather than advertising. The discussion occurs live at Tech Futures Conference in New York, representing Nilay Patel's return from parental leave. The conversation balances technical platform details with strategic business model questions and existential threats from emerging AI agents. Kras maintains an optimistic stance that ZocDoc's operational complexity and integration depth protect against displacement while enabling new AI capabilities that enhance rather than replace the core service.

As heard by us

A grounded Decoder conversation on Zocdoc, AI-assisted doctor matching, and the management choices behind healthcare search.

Zocdoc CEO Oliver Kharraz talks through AI in healthcare in unusually practical terms: not as magic diagnosis software, but as a way to improve the match between patients and doctors.

Read the full review in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Decoder with Nilay Patel