You don't need an agency or a paid tool to find out if AI search engines can read your site. Nine steps, ~45 minutes, your browser is enough. This is the same checklist we use as the foundation of every Cited Digital paid audit.
Visit yoursite.com/robots.txt in your browser. Look for any line that starts with Disallow: tied to User-agent: GPTBot, ClaudeBot, CCBot, PerplexityBot, or OAI-SearchBot.
If any of those are blocked, your site is invisible to AI search by definition. The fix is one line: replace Disallow: / with Disallow: (empty value) for those user agents.
Why it happens: a lot of sites added blocks early in 2024 over fears about AI training. The block also stops AI search, which is the part you want.
On your homepage, right-click anywhere → "View page source" (or Cmd/Ctrl+U). Search for application/ld+json. Count how many you find.
You want at minimum: Organization (or LocalBusiness), WebSite, and BreadcrumbList. Content pages should add Article or BlogPosting. Pages with FAQ-style content should add FAQPage. Step-by-step content should add HowTo.
Sites in our 1,168-audit corpus average 0–1 schema types. Citable sites (75+) tend to have 4+.
Paste your URL into search.google.com/test/rich-results. The tool will parse all your schema and flag errors.
Common issues: missing required properties (an Article without author, an Organization without logo), wrong nesting, or schema text that doesn't match the visible page content.
For each top page, count: how many H1s? How many H2s? Is there an FAQ section with question-headed H3s? Is there at least one bulleted list? At least one table?
Target: exactly 1 H1, 3+ H2s, 1+ FAQ section, 1+ list, 1+ table on a major content page. Walls of paragraphs without these structures score poorly because AI engines extract from headings, lists, and tables — not prose.
Every content page (blog post, service page, about page) should have a named human author. Not "by the team," not "by Acme HVAC" — a real person's name with a link to a bio page.
The bio page should include: years of experience, credentials/licenses/certifications, contact info (email at minimum), and at least one external profile link (LinkedIn, GitHub, professional association).
If you don't have this, add it. Even on a one-person business, "by [your name], owner" is dramatically better than nothing.
Every content page should display two dates: the original publish date and the last-updated date. Both should be visible to readers and reflected in article:published_time + article:modified_time meta tags.
AI engines treat undated content as suspect. Dated content gets cited at materially higher rates.
Open ChatGPT or Perplexity. Type a question your business should answer — say, "best HVAC contractor in [your city]" or "[your category] near me." See if your domain appears in the citation list.
If your name doesn't appear at all, you have an AI-visibility problem regardless of your traditional SEO rankings. If your name appears with a low confidence indicator, your authoritativeness signals are weak.
Run your homepage through pagespeed.web.dev. You're looking at three metrics: Largest Contentful Paint (target < 2.5s), Cumulative Layout Shift (target < 0.1), and Interaction to Next Paint (target < 200ms).
Fail any of those on mobile and AI engines deprioritize your content. Mobile is the default test profile because most buyer queries originate on mobile.
After the manual checks, run a comprehensive audit at citeddigital.co/audit. The free version returns an overall AEO score, a 5-category breakdown, top issues, and a citation status check across ChatGPT, Perplexity, and Google AI in 60 seconds. No signup, no card.
The audit catches anything the manual checklist might miss — and gives you a single number you can track over time as you fix things.
About 45 minutes for a 5–10-page site. Bigger sites (50+ pages) take longer because you'll want to spot-check schema and structure on each major template, not just the homepage.
By the 9-step checklist, you want to clear at least 7 of 9. By the AEO score from our audit, 75+ is the citable baseline and 90+ is top-tier. The corpus average across 1,168 sites we've measured is 60.6.
Most of the fixes — adding bylines, dates, contact info, FAQ sections, robots.txt updates — are content changes anyone can make in their CMS. Schema markup is more technical but well within reach for anyone comfortable editing HTML, and tools like Yoast or Rank Math handle it on WordPress automatically.
After every meaningful site change. Quarterly at minimum. AI search engines update their citation graphs every few weeks, so a site that scored well six months ago may have drifted as the rules tightened.
For diagnosis, mostly yes. The 9-step checklist plus our free scan tells you where you stand. The $497 paid audit at Cited Digital adds: page-by-page fix priorities, paste-ready schema blocks tuned to your specific site, AI-citation gap analysis versus competitors, and a Fix Manifest formatted for ChatGPT or Claude Code so a non-developer can implement most of the fixes.
Questions? Contact Connor