AEO 101: Getting Cited in ChatGPT, Perplexity, and Google AI Overviews

For the last 25 years, the model has been: optimize content so Google ranks it, get traffic, convert traffic to leads. That model is being eaten from both ends.

On the user side, an increasing share of searches never reach a results page, the answer appears inline, the user reads it, the user leaves satisfied. Google's AI Overviews now appear on 30%+ of queries and growing. Perplexity served 780M queries in March 2026. ChatGPT search is the default for an entire generation of younger users.

On the publisher side, the rules of who gets credit for those answers are different. Being the #1 organic result no longer guarantees you're cited in the AI answer above it. Being on page 2 doesn't disqualify you. The model is shifting from "where do you rank" to "are you cited."

This is Answer Engine Optimization (AEO). Here's the actual technical playbook, what works in 2026, what doesn't, and why.

What changed

Classic SEO and AEO share most foundational signals. Both reward:

• Topical authority (multiple substantive pieces on a single topic, not one shallow article)
• Structured data (Schema.org JSON-LD on every page)
• E-E-A-T signals (named authors, real bios, citation precedent)
• Clean technical infrastructure (canonicals, sitemap, fast pages)
• Internal linking that forms topic clusters

If you do those things well, you'll win at both. The mistake is treating AEO as a separate workstream that requires different content. It doesn't.

What's actually different is how AI engines decide whose sentence to use when answering a user's question. Three signals drive that decision in our experience:

1. Structural extractability

This is the biggest single lever, and it's the one most agency blogs ignore.

When a user asks Perplexity "what's the difference between SEO and AEO, " Perplexity isn't going to rewrite three paragraphs of your marketing copy into a clean answer. It's going to look for an answer that's already structured as an answer somewhere on the open web, a sentence or short paragraph that directly answers the question, surrounded by enough context to feel authoritative.

So the content patterns that win:

• FAQ sections with literal question-answer pairs. Wrapped in FAQPage JSON-LD schema. Each answer 40-80 words, complete sentences, no marketing-speak.
• Definitional H2s. "What is X" or "How does X work" as a heading, followed by a 2-3 sentence definitional paragraph. The first sentence should be the definition itself.
• Lists where each item is a complete thought. Not five-word bullets that require the surrounding paragraph to make sense.
• Comparison tables. "X vs Y" tables with concrete attributes are heavily extracted.

What loses:

• Marketing copy that hedges. "Most companies find that..." reads as filler to a language model and gets skipped.
• Paragraphs that bury the answer in the third sentence after warmup.
• Adjective-heavy descriptions of how great your service is.

2. Citation precedent

AI engines don't decide who to trust in a vacuum. They use the same authority signals classic SEO uses, most importantly, who else cites you on this topic.

A page on your site that has been linked from three other authoritative sites in the same topic cluster is dramatically more likely to be cited in an AI answer than the same page with zero external citations, even if both rank similarly in classic SERPs. The AI engine is using the link graph as a proxy for "is this source trusted by humans who already evaluated it."

This is why pure on-page optimization isn't enough for AEO. You also need to be the kind of source other people cite. That's a content + outreach strategy, not just a content strategy.

3. Topical authority depth

If your site has one well-written page on a topic, you're a candidate. If your site has eight pages on adjacent angles of the same topic, primary explainer, deep technical post, comparison, case study, beginner guide, advanced playbook, related concepts, FAQ, you're a much stronger candidate.

This is the topic cluster model. AI engines reward depth in a way classic Google SERPs do, but more aggressively. The ranking algorithm might surface your single best page; the AI engine cites you because across your topic cluster you look like the most credible source.

The technical AEO checklist

Concrete things to ship, in priority order:

Foundation (do these first)

• FAQPage JSON-LD on every page that has a Q&A section. Pull questions directly from "People Also Ask" in Google for your target queries, those are queries the engines already think users have. Answer them in 40-80 words each.
• Article / BlogPosting JSON-LD on every blog post. Include author as a real Person schema with sameAs pointing to a LinkedIn profile. Without this, you're publishing anonymously and AI engines treat anonymous content as low-trust.
• Organization JSON-LD on the site root. With sameAs linking to LinkedIn, Twitter, Crunchbase. Builds the entity graph the AI engines use.
• BreadcrumbList JSON-LD on every non-root page. Helps both classic SEO and AI extraction understand site hierarchy.
• Self-canonical on every indexable page. Sounds obvious; constantly broken on the sites we audit.

Content structure (do these next)

• Lead every blog post with a definitional paragraph. If the post is about AEO, paragraph one should be: "Answer Engine Optimization (AEO) is..." Not a hook, not a story, not "in this post we'll cover", a definition.
• Use H2s that are questions. "How does AEO differ from SEO" is better than "AEO vs SEO" because the H2 itself matches a user's likely query.
• Comparison tables. Whenever you mention two related concepts, table them. Pillar cards are great too.
• One canonical "ultimate guide" per topic cluster. This is the page that should rank, get cited, and link out to your eight supporting posts.

Authority signals (do these continuously)

• Get cited on industry publications. HARO / Connectively responses get you quoted in real journalism, which is far more valuable for AEO than a guest post on a low-DR blog.
• Publish on a real cadence. AI engines refresh more often than classic crawl; new content gets surfaced faster than legacy SEO suggested.
• Get listed in relevant directories. G2, Clutch, Trustpilot, industry-specific directories. Each becomes a citation source.
• Optional: llms.txt. Cheap to deploy, modest benefit. Worth doing.

What we don't do

• We don't pay for "AI citation" services. Several have emerged claiming to "guarantee" ChatGPT mentions. They're either content farms (Google penalizes those, AI engines downrank them) or they're submitting your URL to crawler endpoints that have no relationship to citation decisions. Skip.
• We don't keyword-stuff for AI engines. They're better at noticing this than Google's classic algorithm. Write for humans first.
• We don't optimize separately for each AI engine. ChatGPT, Perplexity, Gemini, Copilot, Claude, they have different crawl frequencies and different ranking algorithms, but the underlying signals overlap heavily. Optimize the foundation and you get all of them.

How long until it works

Faster than classic SEO, surprisingly.

We've seen first citations appear in Perplexity within 2-4 weeks of publishing well-structured content on a moderately authoritative domain. The engine retrains continuously and crawls aggressively.

ChatGPT search lags, typically 4-8 weeks. OpenAI's crawl pipeline is less aggressive than Perplexity's.

Google AI Overviews are the slowest, often 8-16 weeks, because they piggyback on Google's classic ranking signals. If you're not already ranking in the top 5 organic results, you're not getting cited in the AI Overview.

For a new domain (DR 0 or low single-digits), expect to see traction in 2-4 months across all engines, with the first citations coming from Perplexity. Compounding gains continue for 12+ months.

The honest version

AEO isn't a separate marketing channel. It's the same content investment, structured better.

The agencies promising "AI citation optimization" as a standalone service are mostly selling foundational SEO with new vocabulary. That's not necessarily a bad thing, most sites need foundational SEO, but it's good to know what you're actually paying for.

The single highest-ROI thing you can do for AEO this quarter: take your existing top-10 blog posts and rewrite the first paragraph of each as a definitional answer to the most common user query about that topic. Add FAQPage JSON-LD. Submit to Google Search Console. That's it. The next time an AI engine crawls and re-trains, you're a candidate.

That's the playbook we run for clients on the SEO & AEO pillar. If you want a free audit of your site's AEO posture, what's structured well, what's invisible to AI engines, and what specific pages to fix first, book a 15-minute call and we'll record a Loom walkthrough you keep regardless of whether you hire us.