- The audit covers 4 dimensions: structure (8 items), content (10 items), technical (8 items), and external signals (6 items). Total: 32.
- Each item is yes/no/partial. Score: <40% critical, 40–70% solid, >70% AI visibility leader candidate.
- Duration: 2–4 hours for one person who knows the site. No platforms or paid tools required.
- Priority: items marked red in the table — those alone deliver 60% of the result.
- Action plan at the bottom: 90 days split into 12 weeks with concrete weekly deliverables.
Who should run this audit?
This audit is built for a marketing lead, founder or growth person — not for a developer. It doesn't need server access, doesn't need technical knowledge beyond opening DevTools in a browser. All we need: your site open in one tab, Google Search Console in another, and ChatGPT in a third. That's it. What this audit is NOT: a technical security audit, a performance audit (PageSpeed Insights does that), a legal audit. This is strictly an audit of AI visibility and structural readiness for GEO.
1. Open your site on desktop and walk through the items in order. Mark each as YES, NO, or PARTIAL.
2. For every NO and PARTIAL write one sentence what's missing and what YES would look like.
3. Sum the score (YES = 1, PARTIAL = 0.5, NO = 0). Maximum 32. Percentage is your AI-readiness score.
4. Look at the „Priority“ column and start with the red items first.
Dimension 1: structure
AI systems look for content they can understand and extract. Structure is what separates a page that gets cited from one that gets skipped.
| Item | What to look for | Priority |
|---|---|---|
| TL;DR or summary at top of articles | The article's first 100 words answer the question the article poses. | High |
| FAQ section on key pages | Service and product pages have an FAQ with 5+ questions customers actually ask. | High |
| Logical H1 → H2 → H3 hierarchy | Every page has exactly one H1. H2 and H3 are semantic, not stylistic. | Medium |
| Schema.org Organization in sitewide | View Source — is there a JSON-LD block with @type Organization. | High |
| FAQPage schema on FAQ blocks | Validate in Google Rich Results test. If it passes — YES. | High |
| Article schema on blog posts | Headline, datePublished, dateModified, author, image — all required. | Medium |
| BreadcrumbList schema | All pages except homepage have breadcrumbs and schema. | Medium |
| An „About what we do“ page with a clear one-line definition | First sentence of that page is a category/service definition. | High |
Dimension 2: content
AI systems reward deep, quotable, original articles. Not length for its own sake, value per word.
| Item | What to look for? | Priority |
|---|---|---|
| 8+ deep „what is / how to choose“ articles | Each 1500+ words, each directly answers one customer question. | High |
| Topical cluster — related topics around one category | Min. 8 articles around the same main topic, cross-linked. | High |
| Original research or data | Min. 1 article with data only you have (internal benchmarking, customer survey). | High |
| Author bio on every article | Photo, name, short bio, LinkedIn link. Person schema. | Medium |
| Publication date and last-updated date | Visible on every article. Schema datePublished + dateModified. | Medium |
| Reading time displayed | Headline, datePublished, dateModified, author, image — all „X min read“ at the top of the article. | Low |
| Tables for comparisons | Any „X vs Y“ or „how to choose“ has a comparison table. | Medium |
| Concrete examples with numbers, not fluff | Statements like „we help brands grow“ → REPLACED with „average +220% organic traffic in 6 mo.“ | High |
| Glossary or „terms“ section for key concepts | If your industry has special terms, one page that defines them. | Low |
| Service pages with clear pricing or pricing range | Any budget signal. „Price on request“ on all — automatic NO. | Medium |
Dimension 3: technical
Technical basics LLMs and AI crawlers check. Without these, all the content in the world won't get cited.
| Item | What to look for? | Priority |
|---|---|---|
| PageSpeed mobile 90+ | Run pagespeed.web.dev. Mobile score matters more than desktop. | High |
| robots.txt allows AI crawlers | User-agent: GPTBot, ClaudeBot, PerplexityBot, Google-Extended - Allow: / | High |
| sitemap.xml exists and is fresh | All main pages listed, lastmod recent. | High |
| HTTPS everywhere + HSTS header | Open the page, check the URL bar — no „not secure“ warnings. | High |
| Mobile-first responsive at 5 widths | Test at 360, 480, 768, 1024, 1440px. All readable and usable. | Medium |
| Open Graph tags on every page | og:title, og:description, og:image specific per page. | Medium |
| hreflang if the site is multi-language | Bilingual site must have hreflang in . | Medium |
| 404 and 301 redirects properly configured | Old URLs redirect to new ones. 404 page is helpful, not dead-end. | Medium |
Dimension 4: external signals
AI systems look beyond your site. Mentions, profiles, authoritative sources — all build your brand's „weight“ inside LLMs.
| Item | What to look for? | Priority |
|---|---|---|
| Google Business Profile (if local) | Fully filled. Min. 10 reviews. Recent posts. | High |
| Active LinkedIn company page | Min. 1 post per month. Founders and team active from personal profiles. | Medium |
| Wikipedia entry or your category mentions you | If your category has an article — are you in the „notable companies“ list? | High |
| 3+ industry media mentions in last 12 months | Interviews, quotes, guest posts. Linked, not just „logo wall“. | High |
| Active profiles on key platforms | GitHub (tech), Crunchbase, Pitchbook, Trustpilot — whatever's relevant. | Medium |
| Reviews on 2+ platforms | Min. 10 reviews on 2 different platforms (Google + relevant industry). | Medium |
Your score, what it means?
| Percentage | Level | What now? |
|---|---|---|
| 0-40% | Critical | Your site isn't competitive in the AI era. Bigger interventions needed — but good news: you're starting from a low base and progress is fast. Start the action plan from week 1. |
| 41-70% | Solid | You have the foundations. Focus on NO and PARTIAL items marked HIGH priority — that's 60% of the effort for 80% of the effect. |
| 70-90% | Strong | You're better than most. Time for original research (item 11) and topical cluster (item 10) if not there yet. |
| 91-100% | Leader | Ready for an AI visibility leader position in your category. Now it's all about consistency and ongoing measurement. |
90-day plan: what to do week by week
The plan is built for someone with 4 - 6 hours per week to work on this. If you have less, spread it over a longer period. If more, compress it.
Weeks 1 - 4: Foundations
Week 1: run this audit. Write down the score. Open a Notion/Trello board with 32 items.
Week 2: implement schema markup (items 4, 5, 7). Validate in Rich Results test.
Week 3: write TL;DR summaries for the 5 most important pages (item 1).
Week 4: add FAQ sections to all service and product pages (item 2). Min. 5 questions per section.
Weeks 5 - 8: Content
Week 5: write article #1 of your topical cluster (items 9, 10).
Week 6: write article #2. Link to #1.
Week 7: prepare original research (item 11) — internal benchmarking or a 20–50 customer survey.
Week 8: publish the research. Outreach to 5 industry publications.
Weeks 9 - 12: External signals and measurement
Week 9: update LinkedIn company page and 3 team personal profiles. Set a cadence (1 post per week minimum).
Week 10: check Wikipedia entry for your category (item 29). If you're not there, propose an editor.
Week 11: write articles #3 and #4 of the cluster. Outreach for podcasts/interviews.
Week 12: repeat the audit. Compare the score. Set up a monthly Citation Report rhythm.
When to ask for help
This audit and plan are designed for a marketing team to run themselves. Outside help is needed when: (1) the score is <40% and the team lacks 12 weeks of capacity, (2) you need original research with an analytical component outside your skill set, (3) the site's technical infrastructure doesn't allow clean schema implementation (legacy WordPress with 17 plugins, custom enterprise CMS without flexibility), or (4) you want a monthly AI Citation Report that automatically tracks 100+ questions.
Most SMB and scaleup companies can reach 70–80% on their own in 90 days. The last 20–30% (original research, external outreach, multi-language cluster) is where agencies like ours add value. But 80% is achievable solo, which is good news, you don't have to wait for an agency budget to start.