Authoritas's 2026 study found Google AI Overviews now appear on 47% of US queries and reduce click-through to organic results by an average of 34.5%. Even cited sources lose clicks compared to pre-AI baselines — but cited sources lose much less than uncited ones.
The strategic question isn't "how do we rank past AI Overviews" — they're not going away. It's "how do we get into the citation slot so we still earn clicks?"
The 5 page archetypes that consistently get cited
- The Definitive Definition page. 1,200–1,800 words on a single concept. TL;DR at top. FAQ at bottom. FAQPage schema. Example: "What is generative engine optimization?"
- The Comparison page. "X vs Y vs Z" with a comparison table. Comparison content earns AI citations because the engine wants to summarise options.
- The Data-Study page. Original research with a methodology section, charts, and downloadable raw data. The hardest archetype to produce but the highest citation rate.
- The Step-by-step Guide. Linear how-to with HowTo schema. AI engines preferentially cite step-based answers.
- The Updated-list page. "Best X tools in 2026" with a visible last-updated date. Recency signals matter a lot to AI engines.
The TL;DR pattern that earns the top citation
Every AI-citable page should open with a 2–4 sentence summary that directly answers the implied user query. AI engines preferentially grab from this slot.
Format: "[Direct answer in 1 sentence]. [One supporting fact with a number]. [What the rest of the page explains]."
Example: "Generative Engine Optimization (GEO) is the discipline of being cited by AI search engines, not just ranked by them. In 2026, 47% of Google queries trigger AI answers — and only the cited sources still earn clicks. This guide covers the schema, content and citation tactics that get you cited."
The schema stack for AI citation
Every page archetype above should include:
- Article or BlogPosting schema with
datePublishedanddateModified - Author Person schema with
jobTitleandsameAs - FAQPage schema if there's a Q&A block
- SpeakableSpecification pointing at the TL;DR selector
- BreadcrumbList
This stack tells the AI engine: who wrote it, what it's about, what the key takeaway is, and how it fits in your site. That's everything an LLM needs to confidently cite.
Measurement — what to track post-AI
- AI Overview impressions (Search Console — limited but improving)
- Manual AI citation captures (monthly screenshot of top 20 queries across ChatGPT / Gemini / Perplexity)
- Brand-mention referral traffic (GA4 source = chat.openai.com, perplexity.ai, etc.)
- Organic click rate decline vs CTR on cited pages (the gap shows your citation value)
"Don't try to rank past AI Overviews. They aren't going away. Get into the citation slot, and you still earn clicks."
— Senior strategist, The Review Makers
