Traffic from Google Search dropped meaningfully across every client account I touched in Q1 2026. Same content, same authority, same backlinks — fewer clicks. Not because anything was broken on the site. Because the audience shifted.
Half of the queries that used to land on a client blog post now resolve inside a ChatGPT, Claude, or Perplexity answer card. The user gets the answer without ever clicking through. That's not Google's fault and it's not the client's fault. It's a structural shift in how people retrieve information.
The mistake most agencies are making in 2026 is treating AEO like SEO with a new acronym. It is not. SEO optimizes for a ranking — a position on a results page. AEO optimizes for a citation — a sentence with your brand name inside an AI-generated answer. The signals overlap, but the priorities are different.
Five things we ship on every AEO-tuned build: (1) Schema.org markup for the page type — Article, Service, FAQPage, Person — populated with real data, not boilerplate. (2) Direct-answer blocks at the top of every long-form page — a single paragraph, ≤60 words, that answers the page's title question literally. (3) An llms.txt file at the site root listing the canonical URLs and a brief site description. (4) An ai.txt file declaring training-allow policy and attribution requirements. (5) A robots.txt that explicitly allows the AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) you actually want indexing you.
Beyond the technical scaffolding, the writing has to change. Long preambles get ignored. Listicles get parsed but rarely cited verbatim. The blocks that get cited are: definitions, numbered lists with cause-and-effect logic, and direct quotes attributed to a named person.
If you're not seeing your brand name show up in LLM answers eight weeks after publishing, the problem isn't your content. The problem is your schema layer, your llms.txt, or your robots policy is silently blocking the retrieval. Audit those first.
