How I took a high-authority health and wellness resource from fewer than 100 AI citations to over 52,000 in three months — across AI Overviews, ChatGPT, Gemini, Perplexity, and Copilot. Without a single new backlink.
Context
In January 2026, I joined a high-authority mental health and wellness resource as Senior SEO Consultant. The site had strong traditional organic performance and deep topical authority in a YMYL vertical — but zero meaningful presence in AI-driven answer engines. When I ran visibility diagnostics, the site was being cited by AI Overviews, ChatGPT, Perplexity, Gemini, and Copilot fewer than 100 times combined, across the entire domain. For a site of its authority, that wasn’t a visibility problem. It was a structural one.
Challenge
YMYL health content has two competing constraints:
- Editorial trust is non-negotiable. Every piece of content is reviewed by domain experts. Any technical optimization that could compromise accuracy, tone, or authoritativeness was off the table.
- AI answer engines favor content that’s cheap to parse and extractable as discrete passages. This means clean semantic HTML, structured data, direct answers at the top of pages, and self-contained quotable sections — which was the opposite of how most of the site’s content was organized.
The challenge was bridging these two worlds: making the content machine-readable without editorializing it, and optimizing for extractability without sacrificing the trust signals that made the site a health authority in the first place.
Strategy
The work broke into five areas:
1. Fix the technical foundation
Before touching content, I resolved the foundational technical SEO issues that were making the site unnecessarily expensive for crawlers to parse:
- Restructured header hierarchy across hundreds of pages to fix skipped levels and misaligned H1/H2/H3 logic
- Migrated inline CSS to cacheable external files to reduce LCP and improve crawler efficiency
- Cleaned up redirect chains that were wasting crawl budget
- Fixed canonical tag misconfigurations that were causing indexation confusion
These fixes alone improved organic clicks by 46.7% and impressions by 89% within three months — before any AEO-specific work.
2. Implement structured data for AI readability
I rolled out JSON-LD schema across the site: Article, FAQPage, MedicalWebPage where appropriate, Organization, and BreadcrumbList on all nested pages. Structured data is one of the lowest-cost ways for an LLM to extract factual information — no interpretation required, no ambiguity about what’s an author versus a date versus a topic.
3. Restructure content for information gain
I worked with the editorial team to move the most unique, original, or authoritative information to the top of every piece of content. LLMs reward information gain — content that provides something other sources don’t. For health content, this meant surfacing the clinical perspective, the cited studies, and the specific frameworks this resource was known for, rather than burying them under introductions.
4. Optimize content extractability for LLM citation
I redesigned content blocks to be self-contained and quotable:
- Direct answers within the first 100 words
- Clear definition blocks for key terms
- FAQ sections with standalone 2-3 sentence answers
- Clean list and table structures where appropriate
- Semantic HTML5 throughout (
<article>,<section>,<time>,<cite>)
The goal was that any LLM could pull a clean, authoritative 2-3 sentence passage from any page and use it as a direct citation.
5. Reduce crawler rendering cost
Semantic HTML and clean markup aren’t just accessibility best practices — they’re efficiency signals. AI crawlers have compute budgets. Pages that are cheap to parse get crawled more completely and more frequently. We audited every template for DOM bloat, unnecessary JavaScript dependencies, and render-blocking resources.
Results
Over three months:
- AI Overview citations: 0 → 49,700 across 695 pages
- ChatGPT citations: 0 → 1,000 across 208 pages
- Perplexity citations: 0 → 911
- Copilot citations: 0 → 128
- Gemini citations: Trending positive with weekly growth
- Total AI citations: Less than 100 → 52,000+
- Organic clicks: +46.7%
- Organic impressions: +89%
- Average position: 8.4 → 6.9
All of this was achieved without a single new backlink, a content pruning exercise, or a change in editorial voice. The work was entirely structural.
What This Engagement Demonstrates
AEO is not SEO with a coat of paint. It’s a different discipline with different variables. Traditional SEO optimizes for ranking in a ten-blue-links environment. AEO optimizes for being one of three or four sources cited in an AI-generated answer. The variables that matter — information gain, content extractability, structured data coverage, semantic HTML quality, crawler rendering cost — are adjacent to SEO but not the same. A site can rank well in Google and be invisible in AI answer engines. This engagement is the proof.