web dev//SEO//AEO

2026-03-08

AEO (answer engine optimization) is the practice of preparing web content so that LLMs and AI-powered search (ChatGPT, Perplexity, Google AI Overviews, Claude) can discover it and cite it, and it serves anyone who wants a site to be the source an AI quotes in its answers. It differs from traditional SEO in its consumer: search engines rank pages in a list, while answer engines synthesize information and cite sources, so the aim is to be quoted rather than to appear as a blue link.

Content reaches AI visibility by two paths. Through training data (knowledge cutoff), it is included in the next training snapshot, which is slow (months or years) and requires external mentions and critical mass. Through real-time retrieval, Perplexity, ChatGPT with search and Google AI Overviews fetch the content live, which is faster (days or weeks after indexing).

The technical requirements are few:

The content must be readable by crawlers. AI crawlers (GPTBot, ClaudeBot, PerplexityBot) must be able to read the full text and not only the metadata, so SPAs need server-rendered or edge-injected body content.

llms.txt is a static file that summarizes who the author is and what the site contains, written for LLM consumption (a CLAUDE.md read by Claude Code is the same idea at the scale of one repository).

Structured data in JSON-LD helps AI systems understand the relationships between entities.

robots.txt must not block AI crawlers if the site is meant to be indexed.

The identity must be consistent: the same narrative (name, expertise, topics) repeated across several pages lets LLMs build confidence about who the author is.

The field is new (2024 to 2025) and evolving fast, with no established playbook yet; the closest analogy is SEO around 2005.