GEO Content Checklist

The GEO content checklist is a pre-publication review that ensures every article is structured, marked up, crawlable, and citation-worthy enough to earn placement in AI answers. It covers structure, frontmatter, schema, links, crawler access, freshness, and citation-worthiness.

TL;DR. Run this checklist on every article before you publish. It groups the highest-leverage GEO checks into eight categories: structure, answer-first format, metadata and frontmatter, schema and structured data, links, AI crawler access, freshness and maintenance, and citation-worthiness. Skipping any one of them is the most common reason a page never gets cited by ChatGPT, Perplexity, or Google AI Overviews.

Why this checklist exists

In AI search, the question is no longer whether your page can be found — it is whether your page gets selected and cited inside a synthesized answer (Microsoft Advertising, October 2025). Selection rewards content that is structurally extractable, factually verifiable, and freshly maintained. This checklist operationalizes those properties into discrete, testable items so you can ship with confidence.

Use it as a release gate. If a page does not pass every High-priority item, do not publish until it does.

1. Structure

Structure is the single biggest determinant of whether an AI re-ranker can extract from your page. See the GEO hub and source selection for the underlying mechanics.

[ ] Single H1 that matches the page topic and the canonical question.
[ ] Logical H2 → H3 hierarchy with no skipped levels.
[ ] Question-format headings for at least the main H2s ("What is X?", "How does X work?", "When should I use X?"). Question headings mirror how users prompt AI engines.
[ ] Self-contained sections — each section answers one question without requiring earlier context.
[ ] Short paragraphs (2-4 sentences) so re-rankers can lift clean chunks.

2. Answer-first format

Independent studies of AI Overviews and ChatGPT citations show that the majority of cited snippets come from the top portion of the source page. Lead with the answer.

[ ] Answer in the first 2 sentences of the page (answer-first pattern).
[ ] Standardized AI summary block as a blockquote starting with AI summary: immediately after the H1.
[ ] TL;DR paragraph within the first 10-20% of the page, written so an AI engine can lift it verbatim.
[ ] Definition before discussion — every key term is defined before it is analyzed.
[ ] Bottom-of-page FAQ with ### Q: headings to give re-rankers extra extractable spans.

3. Metadata and frontmatter

Frontmatter is the contract your page signs with downstream pipelines (search index, JSON-LD generators, llms.txt). Treat it as source of truth.

[ ] Title tag 50-60 characters and front-loaded with the focus keyword.
[ ] Meta description 120-160 characters, snippet-ready.
[ ] Canonical URL explicitly set.
[ ] Full frontmatter schema — identity, canonical knowledge layer (canonical_concept_id, knowledge_domain, concept_type, entities, aliases, related_concepts), taxonomy, SEO, AI readability, lifecycle, relations, i18n, authorship.
[ ] canonical_concept_id populated and consistent across the cluster (used for dedupe).
[ ] llm_summary ≤ 2 factual sentences, no hype, ready to be cited.
[ ] updated_at and last_reviewed_at equal to today on every refresh.
[ ] version bumped on each meaningful rewrite (e.g., 1.0 → 1.1).

4. Schema and structured data

Structured data gives AI re-rankers explicit, machine-readable hooks.

[ ] JSON-LD Article (or TechArticle / HowTo / FAQPage as appropriate).
[ ] FAQPage schema mirroring the on-page FAQ section.
[ ] HowTo schema for any step-by-step content.
[ ] Author and Organization schema with explicit credentials and sameAs links.
[ ] Article.dateModified updated on every refresh.
[ ] mainEntityOfPage pointing at the canonical URL.

5. Internal and external links

Links signal entity relationships and authority.

[ ] Hub link to the section's pillar page (/geo, /aeo, /technical, etc.).
[ ] 3-5 internal links to sibling articles in the same content cluster.
[ ] External links to authoritative sources for any non-trivial claim or statistic.
[ ] related_articles in frontmatter populated with up to 5 sibling slugs.
[ ] Anchor text is descriptive (no "click here").

6. AI crawler access and indexability

If an AI engine cannot fetch the page, it cannot select it. Audit robots.txt explicitly.

[ ] GPTBot (OpenAI training crawler) allowed.
[ ] OAI-SearchBot (ChatGPT Search) allowed.
[ ] PerplexityBot allowed.
[ ] Google-Extended allowed for Gemini and AI Overviews grounding.
[ ] ClaudeBot (Anthropic) allowed.
[ ] Server-rendered HTML — main content visible without JavaScript execution.
[ ] Canonical URL indexable (no accidental noindex).
[ ] Included in sitemap.xml and, where applicable, in llms.txt.

7. Freshness and maintenance

AI engines weight dateModified and on-page recency cues when re-ranking.

[ ] updated_at / last_reviewed_at refreshed on every meaningful change.
[ ] Stats and figures verified against current sources within the last review cycle.
[ ] Internal links validated (no 404s).
[ ] External citations validated and replaced if the source moved or rotted.
[ ] Review cadence documented (default review_cycle_days: 90).

8. Citation-worthiness

Citation-worthiness is what separates content that is retrievable from content that is cited.

[ ] Every strong claim has a source or has been generic-ized to remove the unverifiable specifics.
[ ] Specific entities named — products, standards, people, metrics — not pronouns or vague references.
[ ] Original perspective or data included where possible (first-party research, screenshots, examples).
[ ] No keyword stuffing — the page reads naturally to a human reader.
[ ] No marketing fluff that an AI engine would discard.
[ ] citation_readiness in frontmatter set to reviewed (or verified after editorial sign-off).

How to use this checklist

Author drafts the article and self-checks Sections 1-3.
Editor reviews Sections 4-5 and pushes back on missing schema or hub links.
Technical reviewer verifies Section 6 against the live robots.txt and a fresh crawl.
Refresh owner runs Sections 7-8 every review_cycle_days (default 90).

For a deeper assessment of an existing site, see the longer GEO Audit Checklist: 50-Point Assessment.

FAQ

Q: When during the editorial workflow should I run this checklist?

Run it twice: once when the draft is structurally complete (before line edits) and again immediately before publish. Running it only at the end usually means structural fixes get rushed.

Q: How often should published pages be re-checked?

At least every 90 days for evergreen reference content, and every 30 days for fast-moving topics like model releases or platform changes. Bump last_reviewed_at and version on every pass so AI engines see the freshness signal.

Q: Will blocking AI bots in robots.txt improve my rankings or save bandwidth?

No. Blocking GPTBot, OAI-SearchBot, PerplexityBot, Google-Extended, or ClaudeBot removes you from the corresponding engine's retrieval pool, which makes citation impossible. If you want to be cited, you must be crawlable.

Q: Who should own this checklist on my team?

The content lead is the natural owner because most items are editorial. Technical items in Section 6 are typically co-owned with engineering or DevOps. Treat the checklist as a living document and update it whenever a new platform or signal becomes important.