Glossary
The AEO checklist
Updated
Definition
An AEO checklist is the set of on-page checks that make a page retrievable, quotable, and citable by AI answer engines such as ChatGPT, Perplexity, and Google AI Overviews.
Most AEO advice is folklore. This checklist is not: every item below is a factor in an automated scorer that runs nightly against every page on this site, and no page change ships here if it regresses the score. The checks are boring on purpose, because what answer engines reward is not cleverness but retrievability: can a crawler read it, quote it, and defend citing it.
The checklist: 1. Server-rendered text: the full content present in the HTML itself, not assembled by JavaScript after load, because answer-engine crawlers quote what they fetch. 2. A front-loaded answer: the direct answer to the page’s question inside the first 150 words, quotable on its own. 3. Structured data: JSON-LD that tells engines what the page is, a defined term, an article, a FAQ. 4. Question-shaped headings: H2s phrased the way people actually ask, because engines match prompts against headings. 5. Verifiable statistics: concrete numbers with sources, since answers built on stats get cited over answers built on adjectives. 6. Visible dates: a published or updated date in the text and in the structured data, because engines discount undated pages.
The checklist is necessary, not sufficient. Passing it makes a page eligible; being the best answer wins the citation. And the two scoreboards diverge: a page can be cited as a source while the brand behind it goes unnamed, because naming is learned from training data while citation is retrieved from the live web. Measure the two separately, on a schedule, with repeated runs, since single answers vary too much to read.
How this shows up in Waldo
This page practices what it lists: the site’s scorer applies these factors nightly and gates every release. The same measurement discipline is what Waldo sells: scheduled prompt panels that score whether a brand is named and whether its pages are cited, tracked as separate series, engine by engine, because the playbooks barely overlap.
Questions teams ask
How often should you re-run an AEO checklist?
Continuously for the mechanical checks, monthly for the outcome checks. Rendering, structured data, and dates can regress with any deploy, so automate them. Citation share moves slower: a fixed prompt panel re-run on a schedule, several runs per prompt averaged, is the honest cadence.
Do AI engines really care about dates?
Yes, in both directions: fresh, dated content is preferred for retrieval, and undated pages are harder to trust in a sourced answer. The date has to be real; bumping a date without changing the page is the kind of pattern engines learned to discount in classic search.
Is AEO different from SEO?
They overlap heavily and diverge at the scoreboard: SEO optimizes for ranking and clicks, AEO for being retrieved, quoted, and named inside generated answers. Good technical SEO is a prerequisite. The full comparison lives on the AEO vs SEO page.
Related terms and reading
Put Waldo behind your agents
Brand, category, and audience intelligence over 200+ API and MCP endpoints. Sign up, mint a key, and run it against the brands you actually track.