38-Point GEO Checklist: The Complete Technical Audit for AI

Why a GEO Checklist Is Essential in 2026

The digital visibility landscape has fundamentally shifted. With over 800 million active users on ChatGPT and massive adoption of AI-powered answer engines, businesses can no longer rely solely on traditional search optimization. Generative Engine Optimization (GEO) has become a strategic imperative.

But how do you know whether a client's website is truly ready for the AI era? That is exactly what the 38-point GEO checklist addresses. This technical audit tool systematically analyzes every aspect of a brand's online presence that influences how language models (LLMs) perceive, understand, and cite it.

Article updated on September 16, 2026: the checklist now includes 38 points (29 automated checks, 9 expert recommendations). The llms.txt file has been excluded from scoring since June 2026 (detected for information only).

Unlike a traditional audit focused on search result positioning, the GEO checklist evaluates a website's readiness to be referenced by artificial intelligence. It distinguishes 29 automatically verified points and 9 expert recommendations, delivering a comprehensive and actionable assessment.

The 29 Automatically Verified Points

These checks are executed programmatically during each audit. Every point receives a status (compliant, non-compliant, partially compliant) and contributes to the overall visibility score.

1. AI Bot Accessibility via robots.txt

The robots.txt file is the first gateway for AI crawlers. The audit verifies not only that this file exists and is accessible, but also that it does not block major AI bots: GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot, Google-Extended, Bytespider, and Meta-ExternalAgent. A misconfigured robots.txt can render your site completely invisible to language models.

2. AQA Markup (AI Question Answer)

AQA markup (open standard, aqa-spec.org, MIT license) structures question-and-answer content with dedicated AI metadata: date per question, source per answer, and a modification history. AI engines prioritize content with verifiable provenance. In the platform, this item carries a 10% weight in the AEO score.

3. Schema.org Organization

Structured data of the Organization type is fundamental for brand identity. The audit verifies the presence and completeness of JSON-LD markup: name, logo, URL, social profiles, address, phone number, and description. The richer these data, the more accurately LLMs can build a representation of your brand.

4. Schema.org FAQPage

FAQPage markup structures questions and answers so AI can directly leverage them. The audit detects this Schema markup on relevant pages, checks question quality and answer completeness. Well-structured FAQ pages have a 3x higher citation rate in AI responses.

5. Schema.org BreadcrumbList

Structured breadcrumb navigation helps LLMs understand your site hierarchy. The audit verifies BreadcrumbList presence across all pages, level consistency, and associated URLs. This structure facilitates contextual understanding of each page by language models.

6. Open Graph Tags

Open Graph tags (og:title, og:description, og:image, og:type, og:url) are read by numerous AI indexing systems. The audit checks their presence, uniqueness per page, description quality, and image compliance (dimensions, accessibility). Missing or poorly filled OG tags significantly reduce your visibility in contextual AI responses.

7. HTTPS / SSL

A valid SSL certificate is a fundamental prerequisite. The audit verifies certificate validity, automatic HTTP-to-HTTPS redirection, absence of mixed content, and the complete certification chain. LLMs assign significantly higher trust to secure sources.

8. Server-Side Rendering (SSR)

Server-side rendering ensures AI bots access your pages' full content without executing JavaScript. The audit compares server-served HTML with client-side rendered DOM, detects SPA frameworks without SSR, and evaluates the quality of content accessible on first load. A site relying 100% on client-side rendering may be invisible to most AI crawlers.

9. XML Sitemap with hreflang

The XML sitemap is essential for guiding AI bots to your important pages. The audit checks sitemap validity, inclusion of all key URLs, hreflang tags for multilingual content, declared update frequency, and consistency with actually accessible content.

10. Canonical Tags

Canonical tags prevent duplicate content issues that confuse LLMs. The audit verifies per-page canonical presence, consistency between canonical and actual URL, absence of loops or chains, and correct handling of URL parameters.

11. Meta Robots Directives

Meta robots directives (index/noindex, follow/nofollow) control page accessibility to bots. The audit ensures important pages are indexable, directives do not contradict robots.txt, and sensitive pages are properly protected.

12. Structured Data Depth

Beyond mere presence, the audit evaluates structured data depth. An Organization markup with only a name and URL is insufficient. The audit measures the number of populated properties, presence of nested types (PostalAddress, ContactPoint, Offer), and overall semantic richness.

13. Page Speed for Bots

Server response speed directly influences AI crawlers' ability to index your content. The audit measures Time to First Byte (TTFB), total response time, and served HTML size. A slow server can cause timeouts during AI crawling, resulting in partial or zero indexation.

14. Mobile Rendering

Many AI bots simulate a mobile user-agent. The audit checks site responsiveness, viewport meta presence, absence of hidden content on mobile, and content parity between desktop and mobile versions.

15. Internal Linking Structure

A robust internal linking structure helps LLMs understand your site's thematic architecture. The audit analyzes navigation depth, internal links per page, orphan pages, and internal PageRank distribution toward strategic pages.

16. Freshness Signals and dateModified

LLMs favor up-to-date sources. The audit checks visible publication and update dates, the dateModified property in JSON-LD markup, lastmod tags in sitemaps, and HTTP Last-Modified headers — the technical signals that prove to AI engines that a site's content is actively maintained.

17. Semantic HTML Structure

Correct use of HTML5 semantic elements (header, nav, main, article, section, aside, footer) improves LLM content comprehension. The audit verifies these elements' presence, consistent usage, and absence of anti-semantic structures (div soup).

18. Heading Hierarchy and Alt Text

A coherent H1-H6 hierarchy and descriptive alt attributes on images are essential. The audit checks H1 uniqueness, logical level progression, absence of skips (H1 directly followed by H4), alt text presence and quality, and language declarations (lang attribute).

19. Meta Title and Meta Description

The meta title is the first signal AI engines read to understand a page's topic; the meta description provides a structured summary. The audit checks their presence, uniqueness per page, and descriptive quality — two straightforward automatic signals that strongly influence source selection.

20. Legal Transparency of the Business

Three regulatory checks grouped into one item: a visible registration number (company ID), direct contact details (email and phone), and identification of the managing director or responsible publisher in the legal notice. Required by the E-Commerce Directive, these transparency signals let AI engines validate a company's existence and legitimacy — a core pillar of E-E-A-T.

21. Absence of the noai Meta Tag

The <meta name="robots" content="noai"> tag opts your content out of AI use. If present, the site is invisible to generative AI engines. The audit detects it and flags it immediately.

22. RSS or Atom Feed

An RSS or Atom feed tells AI crawlers that a site publishes content regularly. Perplexity and OAI-SearchBot use feeds to discover fresh content: content younger than 30 days receives 3.2x more citations.

23. Content Freshness (Under 90 Days)

Pages updated within the last 90 days receive 3.2x more citations from LLMs. The audit reads the dateModified property (Schema.org) and the Last-Modified HTTP header; beyond 90 days, a visible update (recent figures, fresh sources) is recommended. Content left without updates for 13 weeks loses citation eligibility.

24. Answer Density (40-50 Word Paragraphs)

LLMs primarily extract short 40-to-50-word paragraphs that directly answer an implicit question. A text made only of long descriptive paragraphs is not answer-ready. The audit measures the share of paragraphs in answer format: the recommended target is at least 30%.

25. Quotable Blocks (LLM-Preferred Formats)

Blockquotes, definition lists (<dl>), statistics tables, and fact boxes (stat/tldr/fact) are cited with attribution by AI engines. A page written as long continuous paragraphs without these blocks is harder to cite — the audit checks for their presence.

26. Brand Name Consistency (Entity Resolution)

If a brand appears under several variants ('Brand', 'Brand Inc.', 'Brand France'...), LLMs fail to merge these entities and citations get diluted. The audit counts the variants; the target is no more than two distinct variants across the entire site.

27. Top-30 Keyword Density

LLMs build a semantic representation of a brand from the 30 most co-occurring terms across its pages. If these strategic terms are missing or too sparse, the brand effectively does not exist on those topics for AI engines. The audit measures this lexical coverage.

28. Blog or News Section

A blog or news section demonstrates continuous editorial activity. The audit checks that this section exists, is well structured, and is connected to the rest of the site.

29. About Page

A complete About page (story, team, mission, contact details) strengthens AI engines' trust in the entity. The audit checks its presence, its accessibility from the main navigation, and the richness of the information it contains.

The 9 Expert Recommendations

These points require more nuanced qualitative evaluation and are presented as weighted recommendations.

1. E-E-A-T Signals

E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is a major criterion for LLMs. The audit evaluates detailed author pages, expert biographies, qualification mentions, client testimonials, and first-hand experience signals. Read our article on E-E-A-T and AI authority in 2026 for deeper insights.

2. Wikidata / Wikipedia Presence

Language models rely heavily on knowledge graphs. The audit checks whether your entity has a Wikidata entry, a Wikipedia page, and whether information is consistent with your site. Presence on these platforms significantly boosts Entity Health and citation probability.

3. Citation Readiness Score

Citation Readiness measures how easily an LLM can cite your content. The audit evaluates statement clarity, presence of quantified data, question-answer structures, and availability of ready-to-use citations.

4. Backlink Authority Profile

The quality and diversity of backlinks influence the trust LLMs place in your content. The audit analyzes referring domains, inbound link themes, anchor distribution, and overall link profile authority.

5. Content Topical Depth

Topical coverage measures how exhaustively your content covers your expertise areas. The audit evaluates pages per topic, sub-topic granularity, and treatment depth compared to industry standards.

6. Social Proof Signals

Client reviews, testimonials, case studies, and press mentions reinforce LLM-perceived credibility. The audit checks for these elements, their structured markup (Review, AggregateRating), and their freshness.

7. Brand Consistency Across the Web

LLMs cross-reference information from multiple sources. The audit evaluates NAP (Name, Address, Phone) consistency, brand description uniformity, and presence on key industry directories and platforms.

8. Structured FAQ Content

Beyond technical FAQPage markup, the audit evaluates the quality and relevance of Q&As: do they cover real user queries? Are they phrased in natural language? Do they provide complete, self-contained answers?

9. Multi-Platform Presence (3+)

Brands present on at least three platforms beyond their own website (LinkedIn, Wikipedia, Crunchbase, Trustpilot...) are cited significantly more often by conversational AI. Each sameAs link in Schema.org strengthens entity disambiguation. The audit evaluates this external presence and its consistency.

How AI Generates Action Plans from Checklist Results

One of the GEO checklist's greatest strengths is its ability to transform results into concrete actions. After analyzing all 38 points, the system automatically generates a prioritized action plan.

Each identified action is classified into three priority levels:

• High priority (red): blocking actions that directly impact your AI visibility. Examples: robots.txt blocking AI bots, total absence of structured data, no SSR. Estimated implementation time: 1 to 5 days.
• Medium priority (orange): significant improvements that optimize your score. Examples: enriching Schema.org, optimizing Open Graph tags. Estimated time: 1 to 2 weeks.
• Low priority (green): fine-tuning optimizations to maximize your potential. Examples: improving topical coverage, strengthening E-E-A-T signals, adding structured FAQs. Estimated time: 2 to 4 weeks.

Before / After Examples

To illustrate the GEO checklist's concrete impact, here are results observed among our users:

An e-commerce site that corrected the 9 high-priority points identified by the checklist saw its mention rate increase by 34% in 6 weeks. Key actions: unblocking robots.txt for GPTBot, adding enriched Schema.org Product markup, and implementing SSR.

A digital agency using the provider dashboard deployed checklist recommendations across 12 clients simultaneously. Average result: +28% visibility score in 3 months, with particularly strong improvement on the native score.

Integration with Scheduled Audits

The GEO checklist integrates seamlessly with scheduled audits. By configuring recurring audits (weekly or monthly), you can track each checklist point's evolution over time. The dashboard displays trends, identifies regressions, and highlights progress.

For agencies, this feature is particularly valuable: it objectively demonstrates the value of optimizations performed and justifies AEO investments. Every generated PDF report includes the checklist status, facilitating stakeholder communication.

Conclusion: Turning Audits into Competitive Advantage

The 38-point GEO checklist is not just a diagnostic tool: it is a complete methodological framework for AI visibility. By combining automated checks with expert recommendations, it provides a clear roadmap for every audited website. For SEO/GEO agencies and consultants, it is built for fieldwork: audit your clients, prioritize fixes, and demonstrate progress — notably via the provider dashboard.

Also explore our AI visibility audit methodology to understand how these 38 points integrate into a global GEO strategy. And your clients' brands? Test the methodology on a first client to measure the gap concretely.