Checklist
GEO checklist
The GEO checklist is a 20-step audit for AI citation readiness, covering crawler discovery, answer-first content, FAQ schema, topical authority, and ongoing measurement across ChatGPT, Claude, Perplexity, and Gemini. Tick each item and prioritize the ones you are missing.
Discovery
- ✓ robots.txt allows AI crawlers
- ✓ sitemap.xml is submitted to search engines
- ✓ llms.txt lists primary pages
- ✓ Canonical URLs are set
Content
- ✓ Clear H1 and H2 structure
- ✓ Direct answers near the top of pages
- ✓ FAQ schema on key pages
- ✓ Comparison tables where relevant
- ✓ Original data, benchmarks, or case studies
Authority
- ✓ Topical cluster coverage
- ✓ Internal links between related pages
- ✓ Fresh content with recent dates
- ✓ Authoritative external references
Measurement
- ✓ List of customer prompts to monitor
- ✓ Baseline citation rate tracked
- ✓ Competitor citation comparison
- ✓ Monthly review cadence
Frequently asked questions
What is a GEO checklist?
A working list of the technical, content and measurement conditions that have to be true before an AI assistant can cite you. It is ordered deliberately: discovery items gate everything below them, so a blocked crawler or a JavaScript-only page makes the content work academic.
Which item on the checklist matters most?
Crawler access in robots.txt. It is binary, it takes a minute to change, and every other item is downstream of it. After that, a direct answer near the top of the page — the single thing a model needs in order to have something to quote.
How often should I run through this checklist?
Discovery items when the site changes — a new framework, a robots.txt edit, a migration. Content items per page as you publish. Measurement items monthly, because assistant answers drift between model releases whether or not you changed anything.
Can I check these automatically?
The discovery and structural items, yes — the free GEO audit parses your HTML, robots.txt and llms.txt and scores them deterministically. The authority and measurement items need a human or a tracking tool, because they depend on what the rest of the web says about you and on what the assistants actually answer.
Is FAQ schema still worth adding?
For AI citation, yes — it marks question-and-answer pairs as extractable units, which is exactly the shape a model wants. One rule: the answers must appear in the visible HTML too. Schema whose answers exist only in the JSON-LD gets discarded.