Healthcare Provider AEO Case Study: From SEO Decline to AI Citation Authority

⚠️ Composite case study — synthesized from public patterns; not a verified single-company case.

TL;DR

This is a composite case study — a synthesis of public, multi-source patterns from healthcare AEO programs reported by Single Grain, Sagapixel, BrightEdge, Foundation Marketing, MediCloud, and aligned with the FTC's December 2024 health-claims guidance and AAFP's AI-search guidance for clinicians. No single named provider is implied. Numbers represent realistic ranges observed in published AEO reports across multi-specialty US providers. We use this composite framing because (1) public, named provider AEO data is rarely shared at this granularity, and (2) the patterns themselves replicate across the cited reports.

The headline pattern: a multi-specialty US provider with strong pre-2024 SEO authority can lose 30-55% of organic clinical traffic within two quarters once AI Overviews and Google's medical-quality updates compress informational SERPs — and recover within three quarters by re-architecting content and metadata for answer engines. The recovery is not a refresh of old pages; it is a re-platforming of clinical content for citation by ChatGPT, Google AI Overviews, and Perplexity. This case study walks through how, what changed, and what to measure.

Why this case matters now

Healthcare is a leading wave for AEO impact. AI Overviews now appear on roughly 23-25% of all queries (Foundation Marketing, 2025) and disproportionately on health questions. Single Grain's healthcare team reports clients seeing AI Overview citations triple after eight to twelve weeks of focused work, while organic clicks for informational health queries fall 30-40% even when rankings are stable (Single Grain, 2025).

For providers, the consequences cascade beyond traffic:

• Patient-acquisition funnels that depend on "symptom → condition page → booking" break when AI Overviews answer the symptom directly.
• Brand authority erodes as AI Overviews cite competing health publishers (WebMD, Mayo, Cleveland Clinic, NIH) more often than the provider's own clinicians.
• Compliance risk grows when AI engines paraphrase provider content without disclosing context. The FTC's December 2024 health-claims guidance reinforces that providers remain responsible for how their content propagates (FTC, 2024).

The FTC pressure is real: in 2024-2025, Massachusetts courts ordered over $165 million in penalties against health insurers tied to deceptive sales schemes, signaling that regulators are willing to act when health content downstream is misleading. AEO is therefore not optional polish; it is the new floor for healthcare content operations.

The composite provider profile

A five-state, multi-specialty provider (cardiology, orthopedics, primary care, OB/GYN) with:

• 2,400+ pages of clinical and patient-education content built between 2018-2023
• Domain Rating in the high 60s
• Six dedicated content marketers, two SEO contractors, and a VP of Marketing
• Pre-decline organic traffic dominating regional service queries ("cardiologist [city]", "hip replacement recovery")

Between Q4 2024 and Q2 2025, this profile is consistent with publicly reported declines:

• 38-52% drop in organic sessions on informational pages
• 17-24% drop on commercial "find a [specialty]" pages
• AI Overview presence on 31% of tracked clinical keywords by April 2025
• Citation share inside those AI Overviews: under 4% for the provider, vs. 28-41% for major medical publishers (BrightEdge, 2025)

What the audit found

The rebuild started with a four-axis audit:

1. Schema coverage. Only 9% of clinical pages had MedicalCondition or MedicalProcedure schema. Physician pages used generic Person schema, missing medicalSpecialty, hospitalAffiliation, and availableService.
2. E-E-A-T signals. 71% of patient-education articles had no clinical author byline. "Last reviewed" dates were absent on 64% of pages. Citations to peer-reviewed sources averaged under one per 1,500 words (Sagapixel, 2025).
3. Answer-extractability. Long pages buried key answers below 500 words. Tables, FAQ blocks, and definition lists — the formats AI engines preferentially extract — appeared on under 20% of pages (BrightEdge, 2025).
4. Telemetry gap. The team tracked rankings and traffic. They did not track AI Overview presence, citation share inside Overviews, or LLM-referrer sessions. They were blind to the surface that was reshaping their funnel.

The 90-day playbook

Phase 1 (weeks 1-3): Foundation

• Implement MedicalCondition, MedicalProcedure, FAQPage, and HowTo schema across 600 priority pages.
• Upgrade Physician pages to schema.org/Physician with medicalSpecialty, hospitalAffiliation, availableService, and sameAs to NPI registry, hospital systems, and credentialing bodies.
• Add visible clinician-author bylines, credentials, and "medically reviewed by" lines with linked Physician schema (AAFP, 2025).
• Stand up an AEO telemetry stack: Profound or Otterly for AI citation tracking, plus a custom GA4 channel grouping for ChatGPT, Perplexity, Gemini, and Claude referrers.

Phase 2 (weeks 4-9): Clinical-led content rebuild

• Restructure top 200 informational pages around the question hierarchy that ChatGPT and Perplexity actually receive: "What is X?", "What causes X?", "How is X diagnosed?", "How is X treated?", "When should I see a doctor?".
• Move definitive answers to the first 80-120 words. Add a one-sentence "AI-extractable summary" at the top.
• Replace consumerized fluff with clinician-written answers, citing peer-reviewed sources at a target of three references per 1,000 words.
• Add condition-specific FAQ blocks (8-12 questions) with FAQPage schema, drawn from "People also ask," Reddit, and call-center transcripts.
• Reroute boilerplate pages to consolidated cluster hubs to avoid keyword-cannibalization losses that the medic-style updates penalize.

Phase 3 (weeks 8-13): Trust and amplification

• Publish a public clinical-review SOP and changelog. Each rebuilt page records reviewer, review date, and next-review date — surfaced both visually and via dateModified in schema.
• Earn citations from regional health publishers, university medical centers, and patient-advocacy organizations. AI engines disproportionately cite domains already cited by other authorities (Foundation Marketing, 2025).
• Build a clinician-thought-leadership lane: each specialty publishes one bylined deep-dive per month with a Q&A appendix. These pages, not service pages, are most often pulled into AI Overviews.

Outcomes

Within 12 weeks of full rollout:

• AI Overview citation share rose from under 4% to 14-18% on the tracked keyword set.
• LLM referrer sessions grew from <0.5% to 6-9% of total content sessions, with ChatGPT and Perplexity leading.
• Organic sessions on rebuilt informational pages recovered to 78-91% of pre-decline baselines, even though zero-click rates rose another 4 points.
• Booked-appointment conversion from informational pages improved 1.4-1.7× vs. pre-decline, attributable to tighter intent matching and clinician trust signals (Sagapixel, 2025).
• Brand-search volume increased 19% as AI Overviews began naming the provider in answer text.

These ranges are deliberately presented as ranges, not point estimates, because the underlying public reports vary in measurement window and methodology. The directional pattern — citation share recovery preceding traffic recovery, which precedes booking recovery — is robust across the cited sources.

What did not work

• Adding generic FAQ schema to thin pages without expanding the content. AI engines deprioritize repetitive, low-evidence FAQs.
• Buying "AI-friendly content" packages from generalist agencies. Pages without clinician review or citations failed to earn AI Overview citations and triggered medic-update sensitivity.
• Programmatic location pages without unique clinical signal. Hundreds of "[Specialty] in [City]" pages were pruned to a curated, clinically distinct subset.
• Optimizing for ranking position alone. Rank stability without citation share inside AI Overviews still produced traffic decline.

Lessons for healthcare marketers

1. AEO and SEO share fundamentals (E-E-A-T, schema, intent matching), but AEO requires answer-extractability and citation-readiness as first-class properties of every page.
2. Clinician-led content is the durable advantage. Medical schema and named, credentialed reviewers are the highest-leverage E-E-A-T moves available to providers.
3. Telemetry must include AI surfaces. Track AI Overview presence, citation share, and LLM referrers weekly. If you cannot measure them, you cannot defend the channel.
4. Pages should be structured as answers first and articles second. Lead with a clinician-reviewed summary, then deepen.
5. Compliance is part of AEO. Align with FTC health-claims guidance and document review chains. AI engines extract more aggressively than search engines did, so the quality floor must be higher (FTC, 2024).

How to apply this playbook

• Map your top 200 clinical pages by booking influence and re-audit them on the four axes above.
• Stand up an AEO telemetry stack before any content change so you can measure recovery in weeks, not quarters.
• Pair every content sprint with a schema sprint; do not let metadata trail content edits.
• Recruit at least one clinician per specialty to a public review board. Their bylines are the cheapest E-E-A-T signal you can buy.
• Re-run AI Overview prompts monthly. Track citation share alongside traffic and bookings.

FAQ

Q: How long does a healthcare AEO rebuild take?

Most reported programs see citation-share gains in 8-12 weeks, traffic recovery in 12-20 weeks, and booking impact in 16-26 weeks. Timeline depends on content volume, clinician availability, and how aggressively schema and FAQ infrastructure is deployed (Sagapixel, 2025; BrightEdge, 2025).

Q: Do healthcare providers need to prioritize ChatGPT, Google AI Overviews, or Perplexity first?

Google AI Overviews drive the highest absolute volume for healthcare informational queries today. ChatGPT drives the highest engagement per session and is the dominant LLM referrer for branded health questions. Perplexity is small in volume but high in citation transparency, which makes it useful for measurement. A focused program optimizes for AI Overviews and ChatGPT in parallel, with Perplexity as the citation-quality canary.

Q: Is medical schema markup enough on its own?

No. Schema is necessary but not sufficient. AI engines combine schema, on-page evidence (citations, clinician authorship), and external authority (links and mentions from medical publishers). The composite case study above shows pages with strong schema but weak content and no clinician byline still failed to earn AI Overview citations (BrightEdge, 2025).

Q: How does AEO interact with HIPAA and PHI?

AEO content is patient-education content, not patient-record content, so HIPAA risk is generally indirect. The bigger concern is FTC health-claims compliance: AI engines paraphrase aggressively, which can amplify weak evidence into stronger claims. Providers should document evidence chains, avoid testimonials without disclosures, and align with FTC December 2024 guidance (FTC, 2024).

Q: What metrics matter most for healthcare AEO?

Four metrics, weekly: AI Overview presence rate on tracked keywords, citation share inside those AI Overviews, LLM referrer sessions, and booked-appointment rate from informational pages. Rankings remain useful but are no longer sufficient as a primary KPI.