• AI visibility

Guide · AI search

How does AI search bring a virtual care platform new contracts and more patient visits?

It works on two fronts: it can put your platform on the shortlist when a health system, health plan or employer researches vendors, and it can help the patients those clients serve find the virtual care they already pay for. Both matter, because a platform earns its renewal through visits, not only through the signature. No study yet measures how often virtual care buyers use AI assistants, so treat this as a reasoned case built on public filings and surveys.

By Underneath Team Guide Updated

The short version

  1. One health system contract is worth hundreds of thousands of dollars a year: Amwell’s 2025 annual report (opens in a new tab) puts its average annual contract value at $461 thousand per health system client, on a typical contract term of three years.
  2. The pool of large buyers is small and shrinking: Amwell’s average number of health system clients fell from 129 in 2023 to 88 in 2025, which it attributes to market consolidation and its own strategic shift.
  3. Visits decide what a contract is worth: Teladoc Health ended 2025 with 101.8 million US Integrated Care members and average monthly revenue of $1.29 per member, while visits on the Amwell Platform fell from 5.9 million in 2024 to 4.5 million in 2025.
  4. Employers are a real buyer: KFF’s 2025 employer survey (opens in a new tab) found 30% of firms with 50 or more workers contract for virtual primary care, rising to 45% of firms with 1,000 or more workers.
  5. The patients your clients serve already ask AI: in Rock Health’s survey reported by HIT Consultant (opens in a new tab), 32% of US adults had used an AI chatbot for health information, and 74% of them used general-purpose tools such as ChatGPT rather than a provider’s own bot.

Who buys a virtual care platform, and what is one client worth?

Three kinds of organization buy: health systems, health plans and employers. Each signs multi-year contracts and pays for access and use.

Health systems buy the platform to run their own virtual visits, virtual nursing, tele-stroke and remote sitting under their own brand. The buyer is usually a chief digital officer, CIO or virtual care leader, with clinical leaders and security reviewers holding a veto. Amwell says it powers the digital care programs of “approximately 80 of the nation’s largest health systems,” and it hosts a dedicated instance for each client, “brandable by the client.” That detail matters later: the patient often never sees the platform’s name.

Health plans buy virtual care to offer their members. Amwell reports about 50 health plan clients covering more than 90 million lives, and its largest client, Elevance Health, accounted for 31% of its 2025 revenue. Its top ten clients made up 71% of revenue. In this business a handful of contracts can make or break a year. How plans themselves win members is covered in our guide to health insurers.

Employers buy through benefits teams, consultants and health plans. KFF’s 2025 survey shows 30% of firms with 50 or more workers have a contract for virtual primary care that goes beyond their health plan network. Teladoc, which works through “relationships with health plans, employers, providers, health systems and consumers,” reported $1,579.6 million of Integrated Care revenue in 2025. Our guide to healthtech vendors selling to employers and plans covers that buyer in depth.

What one client is worth depends on the model:

  • Platform subscriptions. Amwell’s average annual contract value per health system client was $461 thousand in 2025, down from $488 thousand in 2024, a drop it attributes to churn. Multiply by a three-year term and a single health system is a seven-figure decision.
  • Per-member access fees. Teladoc earned average monthly revenue of $1.29 per US Integrated Care member in 2025. A plan or employer with hundreds of thousands of members is worth millions of dollars a year.

Where does AI search already sit in a virtual care purchase?

Around the early vendor research and on the patient side; no survey yet isolates virtual care buyers.

The people inside health systems are now regular AI users. The American Medical Association’s 2026 survey (opens in a new tab) found 81% of physicians use AI professionally, and Fierce Healthcare reports (opens in a new tab) that 75% of US health systems use at least one AI application, up from 59% in 2025. Neither survey asks whether those leaders use ChatGPT or Gemini to choose a telehealth vendor.

The closest buyer data point comes from outside healthcare. G2’s 2026 survey (opens in a new tab) of software buyers found 51% start their research with an AI chatbot more often than with Google. G2 runs a review marketplace and the sample is not healthcare-specific, so read it as a direction, not a measurement of your buyers.

A buyer who sticks to Google will probably meet an AI answer anyway. In our AI Overview study of US searches, keywords in B2B software and technology triggered an AI Overview on 96.0% of searches, the highest of the eight industries we sampled.

The patient side is better measured. Rock Health’s survey of 8,000 US adults found 32% had used an AI chatbot for health information, double the 16% of a year earlier. Only 5% of those users turned to a provider-offered bot. When a member wonders whether their plan covers a video visit tonight, a general assistant may answer before your client’s app does.

Which questions do virtual care buyers put to AI assistants?

Questions about fit with their EHR, security, use cases, outcomes and alternatives. We wrote the prompts below ourselves as illustrations; none was collected from a real buyer.

BuyerConcernIllustrative prompt
Health systemCategory“Which virtual nursing platforms do multi-hospital systems use?”
Health systemEHR fit“Which telehealth platforms embed video visits inside Epic?”
Health systemAlternatives“What are the alternatives to Amwell for a regional health system?”
Health planMember access“Which virtual care vendors run white-label urgent care for health plans?”
EmployerBenefit design“What virtual primary care options suit a self-insured employer with 5,000 staff?”
Any buyerSecurity“Which telehealth platforms hold HITRUST certification?”

The competition in those answers is broad. Amwell’s annual report lists as competitors platform players such as Teladoc and Caregility, consumer telehealth firms, Microsoft, Amazon and Zoom, virtual nursing vendors, and the EHR companies themselves, including Epic and Oracle Health. An assistant asked a simple category question can draw on any of those.

Each question may trigger several searches behind the scenes. Google says AI Overviews and AI Mode may use a “query fan-out” technique (opens in a new tab), issuing multiple related searches, and OpenAI says ChatGPT search typically rewrites a question (opens in a new tab) into one or more targeted queries. For a virtual care question, those hidden searches might cover EHR integration and certifications separately.

How does an AI answer turn into a contract and then into visits?

Through two linked paths: the buyer’s shortlist, then the patient’s first visit, which together decide renewal.

The contract path. As we read the evidence, it runs like this:

  1. A virtual care leader, benefits manager or plan executive asks a category, EHR-fit or alternatives question.
  2. The assistant names a few platforms and links some of the pages it used.
  3. The team checks those names against peers, KLAS and consultants.
  4. A long evaluation follows. Amwell describes a sales cycle “ranging from a few months to a year” that often includes evaluating competitors, and says large clients “often begin to deploy our solution on a limited basis.” Medical device makers face a similar committee review, covered in how devices build demand before value analysis.
  5. A pilot becomes a contract, typically for three years.

The visit path. Once live, the platform earns its renewal through use. Visits on the Amwell Platform fell from 5.9 million in 2024 to 4.5 million in 2025, and Teladoc’s total visits slipped from 17.3 million to 17.1 million. A contract whose members rarely log in is easy to cut at renewal. Here AI search works for your client, not for you: a member who asks an assistant how to see a doctor online should find the plan’s or health system’s own virtual front door, accurately described.

What makes this industry different. The platform is often invisible to the patient, because the service runs under the client’s brand. So a platform has two visibility jobs: its own name in buyer research, and its clients’ branded services in patient questions. We infer that the second job, done well, protects the first at renewal time.

Policy also shapes demand. Congress extended Medicare telehealth flexibilities, including care from home without rural restrictions, through Dec. 31, 2027 (opens in a new tab), the American Academy of Sleep Medicine reports. Buyers weighing a multi-year contract will ask what happens after that date; a platform that explains it clearly in public gives assistants an accurate answer to repeat.

What decides whether an assistant names your platform?

The assistants do not publish their selection rules; studies point to independent sources, and buyers check proof they can verify.

Documented by the platforms. Google and OpenAI confirm that their AI answers run web searches and show links to sources. Neither explains why one telehealth vendor is named and another is not.

Observed in studies. When our hidden-searches study logged ChatGPT’s queries, it looked for reviews in 46.2% of its answers and ran a search aimed at a named publication, ranking or award in 43.8%. On software questions, Chen and colleagues (opens in a new tab) found AI search drew 72.7% of its sources from earned sites such as reviews and independent publications. In virtual care, a reasonable expectation is that those sources include KLAS, health IT trade press and benefits publications.

What virtual care buyers check. These trust factors are specific to the industry:

  • EHR integration. A KLAS patient engagement study (opens in a new tab) of 67 organizations found 55% used Epic, and 70% of those named Epic as the vendor most aligned with their goals. Amwell says building its technology natively into the EHR through channel partners “may lead to a higher win rate.” A platform that is vague about how it works inside Epic or Oracle Health gives buyers and assistants little to go on. Software sold to smaller practices faces similar EHR questions, as our guide to medical practice software shows.
  • Security posture. Amwell publicly lists HITRUST, ISO 27001 and PCI certifications. Health systems and plans will ask about these before any pilot.
  • Access priorities. In the same KLAS study, 53% of respondents named patient access as a leading investment focus in 2025, up from 48% in 2023. Platforms that publish how they improve access give buyers language that matches their goals.
  • Proof of use. Visit volumes, member engagement and outcomes with named clients, published with their permission.

Our inference. Most of what a virtual care buyer verifies could sit on public pages: EHR integrations, certifications, client logos, use cases and results. We expect a platform whose proof is public, current and consistent to be easier for both the buying committee and the assistant to describe correctly. No study has tested this for telehealth vendors.

What does it cost a virtual care platform to be missing?

Fewer seats in evaluations and weaker renewals, though nobody has yet put a figure on the loss.

  • Each missed evaluation is large. At an average of $461 thousand a year over a three-year term, one lost health system evaluation can mean more than a million dollars of contract value.
  • Fewer buyers raise the stakes. With health system client counts falling and the top ten clients making up 71% of Amwell’s revenue, there are fewer chances to recover a missed opportunity.
  • The EHR fills the gap. Epic and Oracle Health appear on Amwell’s own competitor list. If an assistant cannot place your platform, a reasonable expectation is that the buyer defaults to what the EHR already offers. Newer vendors face this most, as our guide to healthcare startups explains.
  • Answers move. In our consistency study, only 25.2% of the brands ChatGPT named for a question came back in all five runs. Being named once is not a settled position.

How does GEO work for a virtual care platform?

By making your platform and your clients’ services easy for assistants to understand and cite. No mention is guaranteed.

  1. Separate the platform from the brand on the screen. State plainly what you are (an enterprise platform), who buys you, and which clients run services on you, where they agree to be named. Consistency across your site, KLAS profile, EHR marketplace listings and company databases matters. When an assistant confuses you with a consumer telehealth app, our guide to fixing wrong brand information in AI answers sets out the corrections.
  2. Publish integration and security facts. List the EHRs you integrate with and how, your certifications and your data practices on plain web pages, not only in sales decks.
  3. Earn independent coverage. KLAS interviews, health IT and benefits trade press, consultant briefings and conference talks are the kind of earned sources studies associate with AI citations. For a virtual care vendor, the steps for turning that earned record into authority are in how brands build authority for AI search.
  4. Answer each buyer’s question separately. A health system, a health plan and a self-insured employer ask different things. Write honest pages on use cases, comparisons and implementation for each, without clinical claims you cannot support.
  5. Help clients’ front doors get found. Give clients accurate, plain descriptions of their virtual services, coverage and how members start a visit, so a member’s question to an assistant leads to the right place. Commercial health sites that ChatGPT cites tend to show clear review and structure signals, as our summary of what cited health sites have in common explains.
  6. Track buyer and member questions across assistants. Ask them repeatedly in ChatGPT, Google AI Overviews and AI Mode, Gemini, Perplexity, Copilot and Claude; how many prompts to track covers the sample size.

Stay clear of shortcuts. Invented outcomes or planted reviews are a serious risk in a regulated field; see legitimate GEO versus manipulation. Much of the buyer-side work adapts what we describe for healthcare software vendors and for B2B SaaS companies.

Which questions about virtual care and AI search are still open?

Most of them: nobody has yet linked AI visibility to virtual care contracts or visit volumes.

  • No survey of virtual care buyers’ AI use. The physician and health system surveys measure AI adoption in general, not vendor research.
  • Patient data are self-reported. Rock Health’s figures come from a consumer survey, not from logs of what people asked.
  • Filings describe one company each. Amwell’s contract values and client counts are its own and may not match Teladoc, Caregility or EHR-native tools.
  • Policy after 2027 is unknown. The Medicare extension runs through Dec. 31, 2027; demand beyond that depends on Congress.
  • The revenue link is unproven. The general evidence is reviewed in does AI visibility drive business results; for measurement, see how to prove GEO caused sales.

Where should a virtual care platform start?

Ask the questions your buyers and your clients’ members ask, see who gets named, then publish the missing proof.

List what a virtual care leader, a health plan executive and an employer benefits manager would ask at each stage, and what a member would ask when trying to start a visit. Put each question to several assistants more than once. Note which platforms are named, which sources are cited, and whether your EHR integrations, certifications and client services come back accurately.

If you want help with that, talk to us about where AI answers place your platform with health systems, plans and employers. The review shows which buyer questions leave you off the shortlist, where your clients’ virtual front doors are hard to find, and which public proof would most help your pipeline of evaluations and the visit volumes that carry renewals. Our generative engine optimization service page explains how that proof is published and tracked, for buyer questions about EHR fit and security as well as member questions about starting a visit.

Frequently asked questions

Do health system leaders use ChatGPT to pick telehealth vendors?

No published study measures it. What is known: 81% of physicians use AI professionally, and 75% of US health systems use at least one AI application.

Why should a white-label platform care about patient-side AI answers?

Because visits drive renewals. Amwell’s visits fell from 5.9 million to 4.5 million in a year, and members who cannot find a service do not use it.

Does Epic integration matter for AI visibility?

It matters to buyers: 55% of organizations in a KLAS study used Epic. Whether assistants weigh it is untested, but they can only repeat what is published.

Are employers worth targeting separately from health plans?

Yes. KFF found 30% of firms with 50 or more workers contract for virtual primary care, and they research benefits differently from plans.

Does the 2027 Medicare deadline change what we should publish?

Explain clearly how your platform serves clients under current rules and after Dec. 31, 2027, without legal advice. Buyers will ask.

Sources

Free strategy call

Some questions are easier to answer about your own business.

Bring the one that matters most. On a free 30-minute call we’ll take a first look at it and send you a short written read afterward.

Ask about my businessHow we work