The next time a patient walks into a hospital outpatient clinic, it may be without having spoken with scheduling staff or having visited the clinic’s website. For that matter, she may not know which health system the clinic is a part of.

Welcome to the world of chatbot-enabled, self-directed care, and perhaps one of the most consequential manifestations of healthcare consumerization yet.

This type of frictionless, on-demand ordering has existed for more than a decade outside of healthcare: Amazon, Uber, DoorDash, and Instacart are examples. For experiences there’s Airbnb, VRBO and TripAdvisor.

Healthcare has yet to make that experience routine. Health systems and health technology companies have spent years trying to create and own the “digital front door” by investing in websites, portals and apps intended to become patients’ starting point for care.

Patients have their own ideas.

“The patients don’t want to be owned,” Zocdoc CEO and Founder Oliver Kharraz said recently in an interview.

Zocdoc is betting its next chapter on that reality. On September 16, it announced its Care Access Network , making the scheduling infrastructure behind its marketplace available to outside platforms. Partners include Google’s Gemini, Amazon Health AI, Yelp, Healthgrades and Blue Shield of California. The business opportunity is to help patients get care wherever they begin, even if that’s not Zocdoc.com.

For health systems, that raises a more immediate question: what happens when the conversation that leads someone to a doctor takes place somewhere the health system has little influence?

Patients Are Already Outside the Front Door

Patients have been researching their health online for years. WebMD and Dr. Google helped people look up symptoms, learn about conditions and find physicians. But research often ended at the same familiar barrier: call the office to make an appointment.

Conversational AI changes what happens before that point. Consumers are already using LLMs to describe their symptoms, ask follow-up questions and discuss possible next steps in one exchange. The assistant may help shape the patient’s expectations about urgency, specialty and cost before a physician’s name ever appears.

Patients also have more financial reason to scrutinize their care decisions. KFF’s 2025 employer survey found that the average deductible for single coverage among workers with a general annual deductible was $1,886, up 43% over a decade.

Given the conversational nature of LLMs, it isn’t a stretch to think consumer expectations are evolving to expect these technologies not just to answer questions about their symptoms and potential costs of care, but to be able to identify available doctors and even help them book appointments.

AI Is Shifting Access Points And Priorities For Health Systems

Adam Rice, Chief Marketing Officer of CommonSpirit Health, is already adjusting to that possibility. For Rice, helping patients find CommonSpirit’s clinicians increasingly means making them available beyond the health system’s own websites and patient portals.

“The idea is shifting from being discoverable to really being accessible, right?” Rice noted in an interview.

CommonSpirit’s relationship with Zocdoc illustrates the distinction. In a recent interview, Rice said patients could book CommonSpirit providers directly through Zocdoc and through Zocdoc integrations on Google business listings. Those appointments flow into CommonSpirit’s existing scheduling systems and workflows. Extending that availability into AI answer engines, per Zocdoc’s announcement, was the next step CommonSpirit was working toward.

For Rice, the change also affects how CommonSpirit communicates its expertise. Health system marketing now has to account for an intermediary that interprets information and answers questions on a patient’s behalf.

“LLMs understand us by the information that we have available, so it’s our job to help by providing information about our clinical capabilities, quality, and what’s at the core of our brand,” Rice explained.

A recognizable brand still matters. But the clinical information behind it must be understandable to the services patients consult, and the recommendation has to lead to an appointment they can actually book.

Building The Access Infrastructure

Displaying an appointment is relatively easy. Determining whether a particular patient can book it, and getting that booking into the practice’s workflow, is much harder.

There is a reason that barrier has been difficult to remove.

A physician’s calendar is governed by rules that a simple list of open slots cannot capture. Is the patient new or established? Is the visit an initial consultation, a follow-up or preparation for surgery? Does the clinician treat that condition, accept that insurance plan and see patients of that age? Practices have their own preferences, and electronic health record and practice management systems have their own ways of recording them.

Zocdoc has spent nearly two decades working on that problem. Building its marketplace meant connecting to providers’ systems, learning their scheduling rules and making availability usable for patients.

The company also had to rebuild its economics. When it shifted its pricing model to charging providers fees for new patient bookings, it had to go through a high stakes regulatory gauntlet, ultimately resulting in a favorable 2019 advisory opinion from the HHS Office of Inspector General.

As a result of the painstaking legwork of developing the connective infrastructure, shifting its business model and building out its network across the country, Zocdoc is betting the time is now to begin aggregating both providers and consumer tech companies.

In a recent announcement , Zocdoc said providers can join the network for free, emphasizing that its infrastructure lowers the cost of making availability accessible while preserving a way to charge for bringing practices new business.

“We have an incredibly strengthened value proposition to the provider, because now with a single integration to Zocdoc, they get distribution to all these other front doors, right?” Kharraz said.

The strategic bet has three parts. The first is that Zocdoc can help both sides of the market navigate an uncertain environment. Health systems do not know which AI services will become lasting sources of patients. Consumer platforms need appointments they can offer without building connections to thousands of practices themselves. Zocdoc gives each side an existing network and infrastructure to connect to.

The company’s own announcements themselves can help that process. Recognizable distribution partners give providers a reason to participate; more providers give those partners a reason to expand their use of Zocdoc. If those commitments reinforce each other, Zocdoc can turn the fragmentation of patient demand into an advantage. It also protects its relevance as more searches begin somewhere else.

Second, Zocdoc is betting that its supply is ready. Kharraz says the network has reached critical mass with more than 200,000 providers. That is a substantial base, but a national count says little about the choices available to a particular patient. Specialty, clinician type, geography, insurance participation and actual appointment availability all determine how useful the network is.

For Blue Shield of California, for example, the relevant supply is the portion that can meet its members’ covered care needs. Providers elsewhere, or those who do not accept a member’s plan, add little value to that search. The question is whether Zocdoc has sufficient density in the markets and specialties its partners need.

OpenTable offers a useful analogy. In its 2009 IPO filing , it described building a critical mass of restaurants in individual cities and offering reservation inventory comparable to what diners could obtain by phone. Healthcare adds far more constraints, but the marketplace lesson holds: usable local availability makes distribution valuable.

Third, Zocdoc is betting this proposition can help it win more enterprise health system business. The ability to reach patients across outside platforms gives it a way to compete for those relationships while building on the scheduling systems health systems already operate.

CommonSpirit’s approach shows both the appeal and the limits of that proposition. Rice repeatedly described the partnership as non-exclusive, explaining, “We’re not looking exclusively at one service provider or another. We’re really trying to get our availability and access to the places where our consumers and patients are searching.”

Rice described investing in core capabilities that can support multiple services over time, with partners connecting to CommonSpirit’s existing systems and processes. “We made a decision that in this digital or AI gold rush that we weren’t going to be funding other people’s shovel businesses,” he noted.

Zocdoc fits that approach because it extends CommonSpirit’s reach through existing connections and workflows. But the ambitions differ. Zocdoc wants to become essential infrastructure; CommonSpirit wants several effective ways to reach patients.

Zocdoc will have to keep earning its place as those options evolve.

Zocdoc’s attractiveness as a partner for Yelp and others is contingent on its ability to win more health system business. While there are plenty of other technology companies competing in this space, Epic Systems presents a unique challenge.

Epic is the EHR of record for the majority of enterprise health systems. Its position inside health systems gives it established relationships and direct connections to clinical workflows. It is also expanding patient self-service: its Emmie assistant supports conversational scheduling and rescheduling by text, alongside assistance within MyChart.

Zocdoc’s case to a health system must therefore explain the additional patients and distribution it can deliver beyond the capabilities that system already has.

The timing is interesting; Epic faces several lawsuits alleging anticompetitive conduct, and Reuters reported in August that the Federal Trade Commission was investigating how it grants or withholds access to data. Epic denies the anticompetitive allegations, but the public and regulatory scrutiny it faces may impact both its own policies regarding how it works with companies such as Zocdoc and with whom health systems choose to partner for patient access capabilities

A New Chapter For Zocdoc and Consumers

Zocdoc enters this expansion with a different financial profile from the company that spent years rebuilding its model. In a May update , it said it had been profitable annually since 2023, with revenue growing at a 30% compound annual rate over the preceding three years. It also reported a 2025 Rule of 40 score of 51%, combining revenue growth and profit margin.

The same update announced a Khosla Ventures tender offer intended to give eligible employees liquidity at a price implying a $2.3 billion valuation. Zocdoc has also expanded its leadership team , including its first chief marketing officer and chief revenue officer, and a president of health systems.

Such announcements could be signals of a company positioning for an Initial Public Offering (IPO), but Kharraz was quick to temper such expectations.

Regardless, what is clear is that Zocdoc has big plans.

For consumers, whether Zocdoc meets its own big plans is less important than what the move signals: health systems and technology companies are betting big on removing friction to care.

The next consumer who books an appointment at an outpatient clinic may never know Zocdoc helped arrange the visit. For Zocddoc and its partners, that may be a win.