Ask ChatGPT "Who is the best plastic surgeon in Miami?" and you will get a specific answer. Not a list of ads. Not a directory. A direct recommendation with reasoning.
But what determines which doctor gets recommended? The answer is not reputation, not patient volume, and not how many years they have been practicing. AI systems evaluate a specific set of technical and content signals that most doctors do not even know exist.
Understanding these signals is the first step toward being the doctor that AI recommends instead of the one it ignores. If you have not yet read about why AI is replacing Google for patient searches, start there for the broader context.
The Four Signals AI Systems Evaluate
Based on published research and analysis of how large language models retrieve and cite sources, AI recommendations rely on four primary signal categories:
1. Entity Verification
AI systems need to confirm that a doctor is who they claim to be. They do this by checking whether consistent, structured information exists across multiple sources. If your website says you are a board-certified dermatologist in Atlanta, the AI checks whether Google Business, medical directories, hospital affiliation pages, and social profiles tell the same story.
When the signals align, the AI gains confidence. When they conflict or are missing entirely, the AI has no basis for a recommendation.
2. Structured Data and Schema Markup
Humans read web pages. AI systems read code. Specifically, they look for structured data, a standardized format (called schema markup) that describes information in a way machines can process.
For a doctor, structured data might include: - Medical specialty and board certifications - Practice location with geographic coordinates - Affiliated hospitals and medical groups - Published research or clinical focus areas
According to Google's own documentation, pages with comprehensive schema markup are significantly more likely to appear in AI Overviews and knowledge panels. The Princeton GEO study found that content structure contributes up to 14% of what determines whether AI cites a particular source.
3. Content Authority
AI systems assess whether a doctor has demonstrated expertise through published content. This does not mean social media posts or patient testimonials. It means substantive medical content -- articles about conditions you treat, guides about procedures you perform, educational material that answers the questions patients actually ask AI.
When a patient asks "What should I know before getting rhinoplasty?", the AI looks for authoritative content that answers that question. If you have a detailed guide on your website addressing exactly that question, you become a candidate for citation.
The SE Ranking study of 129,000 domains found that branded content from official domains gets cited 11 percentage points more than third-party mentions. Publishing on your own website matters more than being mentioned on someone else's.
4. Cross-Platform Consistency
AI systems do not rely on a single source. They cross-reference. If your Google Business profile says you specialize in cardiology, your website says interventional cardiology, and a medical directory lists you as a general practitioner, the AI faces conflicting signals. Conflicting signals reduce confidence. Reduced confidence means no recommendation.
The doctors who get recommended are the ones whose professional identity is unified across every platform: website, Google Maps, Google Business, social media profiles, hospital directories, and medical registries.
Why Traditional Websites Fail This Test
Most doctor websites were designed to look professional and provide a phone number. They accomplish that goal. What they do not accomplish is making the doctor machine-readable.
A well-designed website with no entity markup, no structured data, and no medical authority content is essentially invisible to AI systems. The design is for humans. The data layer is what AI reads. Most medical websites have the first but not the second. For a practical checklist, see 5 things every doctor's website needs to get recommended by AI.
This is not a criticism of web designers. Until recently, there was no reason to build websites this way. The shift to AI-powered patient discovery changed the requirements.
The Competitive Window
As of early 2026, most doctors have not optimized for AI visibility. This means the competitive window is open. The first doctors in a given specialty and geography to build proper AI infrastructure will occupy a position that becomes increasingly difficult to displace.
AI systems favor sources they have cited before. They favor content with established authority signals. They favor consistency over time. This means the advantage of acting now is not just being first -- it is building a compounding position that grows stronger each month.
The question is not whether AI will become a primary way patients find doctors. That is already happening. The question is whether you will be the doctor that AI recommends, or the one it has never heard of. For the full roadmap, read the complete guide to AI visibility for medical practices.