Search is splitting in two. There's the classic results page, and there's the AI answer above it — ChatGPT, Perplexity, Google's AI overviews — where more buying decisions start every month. Generic advice gets absorbed and summarized anonymously. The businesses that get named in those answers are the ones whose pages give the engine something specific enough to cite. I restructure sites for exactly this, and this page follows its own advice — here's the working checklist.
Engines assemble answers from pages that already contain answers. "We offer comprehensive digital solutions" contains nothing. "A bilingual five-page site for an LA service business typically lands between $1,500 and $5,000" is a sentence an engine can quote — and attribute. Every question customers actually ask you belongs on a page, phrased the way they ask it, answered the way you'd answer across a table.
Price ranges. Service areas by name. Hours. Process, step by step. What you don't do. Most businesses hide these to "start a conversation" — but AI engines skip vague sources and cite specific ones, so coyness now costs you the exact visibility it used to protect. The businesses being recommended by name in AI answers are overwhelmingly the ones whose pages commit to concrete facts.
Schema markup is how you state facts unambiguously: who you are (Person or LocalBusiness), what you offer (Service), what these questions and answers are (FAQPage), who wrote this and when (Article). It's invisible to visitors and load-bearing for machines. This site carries all four. While you're at it: keep your name, address and phone identical everywhere they appear — your site, Google Business, LinkedIn, directories. Contradictions read as unreliability.
First-person specifics ("when I built the intake flow for a Guadalajara medical practice, WhatsApp beat the contact form ten to one") can't be generated by a model that wasn't there. Engines are getting steadily better at preferring that over interchangeable filler, which means the AI-slop content strategy is a depreciating asset. The durable strategy is publishing things only you can say.
PDFs, image-only menus, content that renders only after five seconds of JavaScript, chat-only information — all invisible or unreliable for engines. If it matters, it lives as real text on a real page. And check your robots.txt isn't blocking AI crawlers while you wonder why AI never mentions you.
Classic SEO asked "what do I rank for?" AEO asks "what am I the citable source for?" — a question about substance, not tactics. Pick the handful of questions where you genuinely know more than the average page on the internet, and answer them better than anyone. That's the whole strategy. The rest is markup.
The foundations overlap — crawlable pages, structured data, real content. The difference is the goal: SEO competes for a ranked position on a results page; AEO competes to be the source an AI answer names and cites. Specificity and verifiability matter more; keyword repetition matters less.
If you sell content itself, maybe. If you sell services, blocking AI crawlers is refusing to be recommended. For a business site I leave them open — being cited by an answer engine is a referral.
Unedited, generic AI output tends to sink because it's identical to a million other pages. Google's guidance targets content made to game search rather than help people. Using AI to accelerate a draft that a human then loads with real prices, photos, and firsthand details is a different thing entirely — the specificity is what gets rewarded.
Structured data and clarity fixes can surface in weeks. Becoming the cited source for a topic is a months-long compounding effect — every specific, honest page raises the odds. There's no instant version; anyone selling one is selling something else.
Have a question these pages don’t answer? Ask me directly — or ask my assistant right now.
Start a Conversation