The Technical GEO Audit Checklist
Audit crawl access, indexing, rendering, entities, schema, feeds, performance and monitoring with a risk-based technical GEO checklist.
A technical GEO audit verifies that important content can be discovered, rendered, understood, consolidated, and monitored by the search and AI systems in scope. It extends a normal technical SEO audit with provider-specific crawler controls and answer-level evidence; it does not replace SEO fundamentals.
1. Crawl and Access
- Inventory search, user-triggered, and training crawlers separately.
- Review robots.txt, meta robots, and
X-Robots-Tagrules. - Check CDN, WAF, bot management, authentication, and geography controls.
- Validate important URLs with live requests and server logs.
- Document the business decision behind every allow or block.
Allowing access does not guarantee inclusion. The AI crawler audit explains how to verify requests safely.
2. Indexing and Canonicalization
- Check status codes, indexability, canonical targets, and sitemap inclusion.
- Identify parameter, regional, print, campaign, and syndicated duplicates.
- Ensure internal links point to preferred URLs.
- Map redirects one-to-one and remove chains.
- Compare search-engine-selected canonicals where available.
3. Rendering and Content Parity
- Compare initial HTML with rendered content.
- Test JavaScript failures, lazy loading, accordions, and client-only metadata.
- Make core answers, links, authorship, and structured data reliably accessible.
- Confirm mobile and desktop parity.
Use the JavaScript rendering for AI search guide for deeper tests.
4. Entities and Structured Data
- Verify organization, people, products, services, and locations.
- Use stable identifiers and consistent official URLs.
- Make markup match visible content.
- Validate syntax and monitor template regressions.
- Remove fabricated ratings, credentials, or unsupported types.
5. Feeds and Discovery
Inspect XML sitemaps, RSS/Atom feeds, product feeds, IndexNow implementation where relevant, and update timestamps. Feeds accelerate discovery in supported systems but do not guarantee citation.
6. Performance and Reliability
Monitor availability, time to first byte, render errors, redirect loops, and intermittent blocks. Prioritize failures affecting high-value pages over cosmetic scores.
7. Measurement
Establish:
- crawler log monitoring;
- index and citation baselines;
- page-level AI referral reporting;
- deployment annotations;
- regression tests;
- incident owners and escalation paths.
Severity Model
| Severity | Example | Response |
|---|---|---|
| Critical | Priority site blocked or returning errors | Immediate incident |
| High | Canonical or render failure across a key template | Fix in current sprint |
| Medium | Inconsistent schema or sitemap coverage | Planned remediation |
| Low | Isolated non-critical metadata issue | Backlog and monitor |
Frequently Asked Questions
Is llms.txt required?
No. It is a voluntary proposal and does not replace robots.txt or platform-supported controls.
Should training crawlers always be allowed?
That is a business and legal decision distinct from search retrieval. Document it with security and counsel where necessary.
What is the deliverable?
A reproducible evidence pack, prioritized tickets, owners, acceptance tests, and a monitoring plan—not a generic checklist alone.
