Most Healthcare AI Strategies Skip the Diagnosis

AI Health Strategist LLC  |  May 20, 2026

In a single seven-day stretch this month, every major healthcare marketing competitor moved on AI.

A 750-person agency rebranded its operating model under a new AI-empowered suite. A boutique digital firm published a defense of paid search in the AI era. A specialty practice shop dropped a tactical playbook for medical groups. A national consultancy started educating health system CMOs that AI is now the first step in the care journey. A major content platform doubled down on AI Overviews structured content. A Tier 2 agency planted a flag squarely in the AI patient acquisition narrative.

Different vendors. Different vocabularies. Same blind spot.

Not one of them is selling measurement.

Every one of them is selling a service that depends on a diagnosis nobody has run.

The Pattern

Healthcare marketing has a long, expensive history of selling tactics as if they were strategy. SEO firms sold rank reporting as a strategy. Programmatic vendors sold impressions as a strategy. Social agencies sold engagement rates as a strategy. The pattern is always the same. A real problem exists: vendors race to claim authority, and the measurement layer that would prove whether any of it is working never gets built.

We are watching it happen again. Faster this time.

The real problem is genuine. ChatGPT, Perplexity, Gemini, Claude, and Grok are now intermediaries in patient care decision-making. AI Overviews appear in most healthcare searches. Patients describe their symptoms in chat, receive recommendations, and choose providers without ever leaving the chat. The top of the funnel has been rewired.

The vendor response is to sell more content, more schema, more SEM, more operating model evolution, and more advisory. All of those have a place. None of them tells you whether AI describes your service line accurately, whether it cites you or your competitor, or how much patient volume is tied to the answer.

You cannot defend volume you have not measured. You cannot measure with a slide deck.

What Measurement Actually Looks Like

AI visibility measurement asks one question. When a patient types a question that should send them to your health system, what does the AI say?

Answering that requires a repeatable methodology. The AHS standard is 75 queries per service line across 5 platforms (ChatGPT, Perplexity, Gemini, Claude, and Grok). That is 375 data points per measurement cycle, per service line. The queries map to the highest-revenue patient intents inside that service line. Every platform is tested on the same queries on the same day. Citations are scored. Accuracy is scored. Competitor displacement is scored.

The output is a benchmark, not a checklist. You know where you stand against peers, which service lines are exposed, which platforms are punishing you, and roughly how much patient volume is at risk. From there, content, schema, structured data, paid media, and operating model decisions actually have a strategic basis.

Without that benchmark, every other piece of work in your AI strategy is a guess.

The CMO Test

Here is a fair test of whether your AI strategy is a strategy or a slide.

Ask three questions in your next leadership meeting:

1. What is our AI citation rate across our top three service lines, by platform?

2. Which competitor is most often cited in our place, and in which platforms?

3. How much net new patient volume per quarter is tied to the queries we are losing?

If the room cannot answer those questions with data, you do not have an AI strategy. You have a content roadmap, a media plan, and vendor relationships. Useful. Not the same thing.

The agencies and consultancies competing for your budget this quarter will tell you the answer is more of what they already sell. That is rational. They are selling solutions. Solutions presuppose a diagnosis. The diagnosis is the work that has not been done.

What To Do This Quarter

Three moves to make in the next 60 days.

Run the diagnostic. Pick your top revenue service line and benchmark your AI visibility across the five major platforms. Compare to peer institutions. Quantify the volume at risk. Whether you do it in-house or with an outside partner, do it before you approve another content budget.

Reset the metric. Stop reporting on Google rank, organic sessions, and click-through rates as your primary AI signal. Those metrics matter. They do not tell you whether AI is recommending you. Citation rate, accuracy rate, and competitor share of voice are the new primary metrics.

Treat AI platforms as distribution. ChatGPT, Perplexity, Gemini, Claude, and Grok are not marketing channels in the traditional sense. They are recommendation systems sitting at the forefront of patient decisions. Health systems that build deliberate strategies to show up well in their systems will defend volume. The ones that keep optimizing for Google ranks will keep losing it without knowing why.

The AI strategy your agency just announced may be helpful. Whether it is or not depends entirely on the diagnosis nobody has run.

Measure first. Then build.

• • •

The AHS National AI Visibility Benchmark Study is the only peer-comparison dataset for health system AI visibility across five platforms and 375 data points per service line per cycle. For a complimentary scan of your top service line, email info@aihealthstrategist.com.