- [ ] **What to check:** Navigate to yourdomain.com/robots.txt and read through every User-agent directive. Look for GPTBot (OpenAI/ChatGPT), PerplexityBot (Perplexity), ClaudeBot or anthropic-ai (Anthropic/Claude), and Google-Extended (Gemini and Google AI features).
- [ ] **Pass criteria:** All four AI crawlers are either explicitly allowed or not mentioned (meaning they fall under a permissive default rule). No `Disallow: /` directives target these crawlers.
- [ ] **Common failure:** A blanket `User-agent: *` followed by `Disallow: /` with only Googlebot explicitly allowed. This blocks every AI crawler by default.
- [ ] **What to check:** Navigate to yourdomain.com/llms.txt. If you get a 404 error, you do not have one.
- [ ] **Pass criteria:** The file exists, loads correctly, contains a clear description of your organization, lists your primary topics and expertise areas, and links to your most important pages.
An AI visibility audit checklist is a structured list of 25 technical, content, schema, and authority checks that determine whether AI search engines like ChatGPT, Perplexity, Claude, and Gemini can find, read, and cite your website. This checklist covers eight technical requirements (robots.txt access for AI crawlers, llms.txt, sitemap status, page speed, mobile-friendliness, SSL, JavaScript rendering, and canonical tags), five schema markup checks, eight content structure points, and four authority signals. If you run through every item on this list and fix what is broken, you will move from invisible to citable in AI search results. You can print this page, work through it with your team, or hand it to your developer as a punch list. Every point is something you can verify today, and most of them take less than ten minutes to check. For teams that want to automate this entire process, GetCited runs all 25 checks programmatically and delivers a scored report, but this free checklist gives you everything you need to do it manually.
The shift from traditional search to AI-generated answers has created a new kind of invisible website. Your site might rank on page one of Google for a dozen keywords and still not appear in a single AI-generated response. That is because AI engines do not scrape a list of ranked results. They read, evaluate, and synthesize content from the sources they trust most, then cite a small number of those sources in their answers. The factors that earn those citations overlap with traditional SEO in some areas and diverge sharply in others. This AI search checklist covers every factor that matters, organized into categories so you can delegate different sections to different team members if needed.
We built this checklist from analyzing the patterns in over 200 AI visibility audits. The sites that consistently earn citations from AI engines share specific technical, structural, and content characteristics. The sites that are invisible almost always have gaps in at least two of the four categories below. Your goal is not perfection on every single point. It is to eliminate the obvious blockers first, then systematically improve the areas that drive the most citation impact.
Let's walk through all 25 points.
How to Use This GEO Audit Checklist
Before you start checking boxes, here is how to get the most out of this list.
Print it or copy it into a spreadsheet. Each item has a checkbox, a description of what to check, and what a passing result looks like. Some items require a developer to fix. Others you can handle yourself in five minutes.
Work through it in order. The categories are arranged by priority. Technical issues come first because they are binary blockers. If AI crawlers cannot access your site, nothing else on this list matters. Schema comes next because it is relatively quick to implement and has outsized impact. Content is the largest section because it is where most of the citation-earning work happens. Authority is last because it takes the longest to build but requires the least ongoing maintenance once established.
Score yourself honestly. For each item, mark it as Pass, Fail, or Partial. A partial score means you have some implementation but it is incomplete. For example, your robots.txt might allow GPTBot but block ClaudeBot. That is a partial pass on the first item.
Prioritize fixes by category. Fix all Technical failures first. Then Schema. Then Content. Then Authority. Within each category, fix the items in the order they are listed. We arranged them that way deliberately.
Now here is the full checklist.
Category 1: Technical Foundation (8 Points)
These eight checks determine whether AI engines can physically access, read, and process your website. Think of this as the front door to your house. If the door is locked, it does not matter how nice the furniture is inside.
Check 1: robots.txt Allows GPTBot, PerplexityBot, ClaudeBot, and Google-Extended
- [ ] What to check: Navigate to yourdomain.com/robots.txt and read through every User-agent directive. Look for GPTBot (OpenAI/ChatGPT), PerplexityBot (Perplexity), ClaudeBot or anthropic-ai (Anthropic/Claude), and Google-Extended (Gemini and Google AI features).
- [ ] Pass criteria: All four AI crawlers are either explicitly allowed or not mentioned (meaning they fall under a permissive default rule). No
Disallow: /directives target these crawlers. - [ ] Common failure: A blanket
User-agent: *followed byDisallow: /with only Googlebot explicitly allowed. This blocks every AI crawler by default.
Our audit data shows that 18.9% of websites are actively blocking at least one AI crawler through their robots.txt configuration. Many of these blocks are unintentional, set up before AI search existed and never revisited. This is the single most common reason websites are invisible to AI search. If you find blocks here, fixing them takes about two minutes and has an immediate impact.
You also want to check whether your robots.txt is blocking access to specific directories that contain your most valuable content. Some sites allow the homepage but block /blog/ or /resources/, which means AI crawlers can see your marketing pages but not the in-depth content that actually earns citations.
Check 2: llms.txt File Exists and Is Well-Structured
- [ ] What to check: Navigate to yourdomain.com/llms.txt. If you get a 404 error, you do not have one.
- [ ] Pass criteria: The file exists, loads correctly, contains a clear description of your organization, lists your primary topics and expertise areas, and links to your most important pages.
- [ ] Common failure: File does not exist at all. Our data shows 92% of websites lack an llms.txt file.
The llms.txt file is a direct communication channel between your website and large language models. It tells AI engines what your site is about, what topics you cover authoritatively, and which pages are your most important resources. Think of it as a cover letter written specifically for AI. Not having one will not get you penalized, but having one gives you a clear advantage over the vast majority of websites that do not.
A good llms.txt file includes your organization name, a one-paragraph description of what you do, your primary content categories, and direct links to your 10 to 20 most important pages. Keep it concise and factual. AI models parse this file to build an understanding of your site's identity and scope.
Check 3: sitemap.xml Is Current and Comprehensive
- [ ] What to check: Navigate to yourdomain.com/sitemap.xml. Verify it loads correctly, includes all your important pages, and shows recent
lastmoddates. - [ ] Pass criteria: Sitemap exists, is valid XML, includes all pages you want AI to find, and shows
lastmoddates within the last 30 days for your active content. - [ ] Common failure: Sitemap exists but has not been updated in months, is missing key pages, or includes pages that return 404 errors.
AI crawlers use your sitemap as a roadmap for understanding your site's structure and identifying new or updated content. A stale sitemap signals to crawlers that your site is not actively maintained, which can reduce crawl frequency. An incomplete sitemap means some of your best content might never get discovered. Check that every page you want cited in AI search results is included in your sitemap, and that the modification dates are accurate.
Check 4: Page Speed Is Under 3 Seconds
- [ ] What to check: Run your key pages through Google PageSpeed Insights or GTmetrix. Record the load time for both mobile and desktop.
- [ ] Pass criteria: Pages load in under 3 seconds on both mobile and desktop. Core Web Vitals are all in the green zone.
- [ ] Common failure: Pages load in 5 or more seconds due to unoptimized images, render-blocking scripts, or slow server response times.
AI crawlers have time budgets. If your page takes too long to load, the crawler may time out, receive incomplete content, or simply deprioritize your site in favor of faster alternatives. Speed also affects the user experience metrics that feed into overall site quality signals. A page that loads in 1.5 seconds gives the crawler everything it needs instantly. A page that takes 6 seconds might lose the crawler entirely.
The most common speed killers are uncompressed images, too many third-party scripts, render-blocking CSS, and slow hosting. Most sites can cut load time in half by compressing images and deferring non-critical JavaScript.
Check 5: Site Is Mobile-Friendly
- [ ] What to check: Run your site through Google's Mobile-Friendly Test. Also manually check your key pages on a phone to verify readability and navigation.
- [ ] Pass criteria: All key pages pass the mobile-friendly test. Content is readable without zooming. Navigation works on touch screens. No horizontal scrolling required.
- [ ] Common failure: Desktop site is fine but mobile version has overlapping elements, tiny text, or broken layouts.
Most AI crawlers index the mobile version of your site, following Google's mobile-first indexing approach. If your mobile experience is broken, the content that AI engines are actually reading may be garbled, incomplete, or improperly formatted. This is not just about user experience. It directly affects what the AI sees when it processes your pages.
Check 6: SSL Is Enabled (HTTPS)
- [ ] What to check: Verify that your site loads with https:// and that the SSL certificate is valid and not expired.
- [ ] Pass criteria: Site loads over HTTPS. Certificate is valid. HTTP requests redirect to HTTPS automatically. No mixed content warnings.
- [ ] Common failure: SSL certificate expired, site still accessible over HTTP without redirect, or mixed content (some resources loading over HTTP on an HTTPS page).
This one should be table stakes in 2026, but we still see sites with expired certificates or incomplete HTTPS implementation in our audits. AI engines treat HTTPS as a baseline trust signal. A site without valid SSL is automatically less trustworthy in the eyes of every major AI model. If your certificate has expired or your site is still serving pages over HTTP, fix this before anything else on this list.
Check 7: No JavaScript Rendering Issues
- [ ] What to check: View your key pages with JavaScript disabled (you can do this in Chrome DevTools under Settings). Compare what you see to the normal version of the page.
- [ ] Pass criteria: All critical content (headings, body text, images, internal links) is visible in the HTML source without requiring JavaScript to render. The page structure is intact without JS.
- [ ] Common failure: Content is loaded dynamically through JavaScript frameworks (React, Vue, Angular) and the raw HTML contains little or no readable content.
This is one of the most overlooked items on this AI search checklist. Many modern websites are built with JavaScript frameworks that render content on the client side. When an AI crawler requests the page, it may receive a nearly empty HTML document with a JavaScript bundle that it cannot or does not execute. The result is that the crawler sees a blank page while human visitors see a full website.
Single-page applications (SPAs) are particularly problematic. If your site is built with React, Next.js, Vue, Nuxt, or Angular, check whether your content is available through server-side rendering (SSR) or static site generation (SSG). If all your content relies on client-side rendering, AI crawlers may be reading a blank page when they visit your site.
The fix depends on your tech stack. Server-side rendering is the gold standard. Pre-rendering specific pages for crawlers is an acceptable alternative. At minimum, ensure that your main heading, first paragraph, and key content sections are present in the initial HTML response.
Check 8: Canonical Tags Are Set Correctly
- [ ] What to check: View the source code of your key pages and look for
<link rel="canonical" href="...">in the<head>section. Verify that the canonical URL points to the correct version of each page. - [ ] Pass criteria: Every page has a canonical tag. The canonical URL points to the preferred version of that page (typically the clean URL without tracking parameters or session IDs). No pages have canonical tags pointing to different pages unless that is intentional.
- [ ] Common failure: Missing canonical tags entirely, canonical tags pointing to wrong URLs, or multiple pages with canonical tags all pointing to the same URL (unintentional duplicate content signals).
Canonical tags tell AI crawlers which version of a page is the "real" one. Without them, AI engines might index multiple versions of the same content (the www version, the non-www version, versions with tracking parameters) and dilute the authority of your page. Worse, they might choose the wrong version as the canonical one, which means the URL they cite in their responses might not be the URL you want users to visit.
Category 2: Schema Markup (5 Points)
Schema markup is structured data that tells AI engines exactly what type of content is on your page and how to interpret it. If Category 1 is about whether AI can access your site, Category 2 is about whether AI can understand your site. Schema does not guarantee citations, but it dramatically increases the chances that AI engines will correctly categorize and prioritize your content.
Check 9: Organization Schema on Homepage
- [ ] What to check: Use Google's Rich Results Test or Schema.org's validator to check your homepage for Organization or LocalBusiness schema. You can also view your page source and search for
"@type": "Organization". - [ ] Pass criteria: Homepage includes valid Organization (or LocalBusiness for local businesses) schema with your company name, logo URL, description, contact information, social media profiles, and founding date if applicable.
- [ ] Common failure: No Organization schema at all, or schema that is missing key fields like description and social profiles.
Organization schema is how you formally introduce your business to AI engines in a machine-readable format. When ChatGPT or Perplexity encounters your site, this schema helps the AI quickly understand who you are, what you do, and how to categorize you. Without it, the AI has to infer all of this from your page content, which is slower and less reliable.
Include as many fields as applicable: name, URL, logo, description, foundingDate, contactPoint, sameAs (for social profiles), address, and areaServed. The more complete your Organization schema, the better AI engines can match your site to relevant queries.
Check 10: Article Schema on Blog Posts
- [ ] What to check: Pick three to five of your most important blog posts and check each one for Article schema (or the more specific NewsArticle or BlogPosting schema types).
- [ ] Pass criteria: Every blog post includes Article schema with headline, datePublished, dateModified, author (linked to a Person entity), publisher, description, and image.
- [ ] Common failure: Blog posts have no schema at all, or Article schema is missing the author and dateModified fields.
Article schema is critical for AI citation because it tells the engine exactly when the content was published, when it was last updated, who wrote it, and who published it. AI engines heavily weight recency when choosing which sources to cite. A page with Article schema showing a dateModified of last week will be preferred over an identical page with no schema and no visible date. The author field is equally important because it connects the content to a person entity, which builds trust signals across the web.
Check 11: FAQ Schema on Key Pages
- [ ] What to check: Identify the pages on your site that include FAQ sections. Check each one for FAQPage schema using Google's Rich Results Test.
- [ ] Pass criteria: Every page with an FAQ section includes valid FAQPage schema that mirrors the questions and answers visible on the page.
- [ ] Common failure: FAQ sections exist in the page content but have no corresponding schema markup, or the schema includes different questions than what is displayed on the page.
FAQ schema is one of the most powerful tools in AI visibility because AI engines are fundamentally question-answering machines. When your page includes FAQPage schema, you are packaging your content in the exact format that AI engines are designed to consume: question followed by answer. This makes it trivially easy for the AI to extract and cite your content when a user asks one of those questions.
The questions in your FAQ schema should match real queries that your target audience is actually asking. Do not fill them with self-promotional questions like "Why is our product the best?" Use the actual questions you see in customer emails, support tickets, and sales calls. Those are the questions people are typing into AI search tools.
Check 12: Author Schema on Articles
- [ ] What to check: Check whether your articles include Person schema for the author, and whether that schema includes the author's name, job title, URL (to their bio page), and sameAs links (to their LinkedIn, Twitter, or other professional profiles).
- [ ] Pass criteria: Every article has Person schema for the author that includes name, jobTitle, url, and at least one sameAs link to an external profile.
- [ ] Common failure: Articles list an author name in the byline but have no corresponding Person schema, or the schema includes only a name with no additional details.
Author schema connects your content to a verified person entity on the web. AI engines use this to assess the credibility of the content. An article written by a named author with a job title, a bio page, and linked professional profiles carries more weight than an article attributed to "Admin" or with no author at all. This is the structured data equivalent of E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness), and AI engines care about it deeply.
Check 13: HowTo Schema on Tutorials and Guides
- [ ] What to check: Identify any tutorial, how-to, or step-by-step content on your site. Check each page for HowTo schema.
- [ ] Pass criteria: Tutorial and how-to content includes valid HowTo schema with clearly defined steps, estimated time, tools/materials needed (if applicable), and images for key steps.
- [ ] Common failure: Step-by-step content exists without any HowTo schema, or HowTo schema is present but missing the step details.
HowTo schema is particularly valuable for AI visibility because how-to queries are one of the most common types of questions people ask AI engines. When someone asks ChatGPT "how do I fix a leaking faucet" or "how do I set up schema markup," the AI is looking for structured, step-by-step content to cite. HowTo schema packages your content in exactly that format. If your competitors have step-by-step content without HowTo schema and you have the same content with proper schema, you have a structural advantage that can tip the citation in your favor.
Category 3: Content Structure (8 Points)
This is the largest category because content is ultimately what AI engines are evaluating when they choose which sources to cite. The technical and schema foundations make sure AI can access and understand your content. But the content itself is what earns or loses the citation. These eight checks cover the structural and qualitative factors that determine whether your content is citation-worthy.
Check 14: First Paragraph Answers the Query Directly
- [ ] What to check: Read the first paragraph of each key page on your site. Ask yourself: If someone asked the question this page is supposed to answer, does the first paragraph actually answer it?
- [ ] Pass criteria: The first paragraph provides a clear, direct, complete answer to the primary query the page targets. No fluff, no throat-clearing, no "In today's fast-paced world" introductions. The answer comes first.
- [ ] Common failure: First paragraphs that set up context, tell a story, or build toward the answer instead of leading with it.
This is arguably the most important content check on this entire GEO audit checklist. AI engines extract answers from web content, and they are strongly biased toward content that front-loads the answer. If your first paragraph directly answers the query with specificity and accuracy, it becomes extremely easy for the AI to extract and cite. If your first paragraph is a generic introduction that takes 200 words to get to the point, the AI will find a source that gets to the point faster.
Look at how AI-generated answers are structured. They lead with the answer, then provide supporting detail. Your content should mirror that pattern. Write the answer first. Provide the context, nuance, and detail afterward.
GetCited's analysis of over 10,000 AI-generated citations found that cited content is 3.2 times more likely to have a direct-answer opening paragraph than non-cited content covering the same topic. This single factor is one of the strongest predictors of whether a page will earn AI citations.
Check 15: Key Pages Have 2,500+ Words
- [ ] What to check: Check the word count of your most important pages (homepage, key service pages, pillar blog posts, comparison pages).
- [ ] Pass criteria: Key pages targeting competitive queries have at least 2,500 words of substantive content. Not filler. Not repetitive padding. Genuine depth that covers the topic thoroughly.
- [ ] Common failure: Important pages have 500 to 800 words that skim the surface of the topic without providing enough depth to be the definitive source.
AI engines are looking for the most comprehensive, authoritative source to cite. Longer content gives you more surface area for the AI to find relevant information to extract. It also signals depth and thoroughness, which are strong authority indicators. Our research shows that pages with 2,500 or more words earn citations at significantly higher rates than shorter pages on the same topics.
This does not mean you should pad your content to hit a word count. Every sentence should add value. The goal is comprehensive coverage, not length for its own sake. If a topic genuinely only needs 1,200 words to cover completely, then 1,200 words is the right length. But for competitive topics where your target audience has complex questions, 2,500 words is typically the minimum threshold for the kind of depth that earns AI citations.
Check 16: 8 or More H2 Headings Per Key Page
- [ ] What to check: View the source code or use a browser extension to count the H2 headings on your key pages. Check that they are descriptive and cover distinct subtopics.
- [ ] Pass criteria: Key pages have eight or more H2 headings, each one clearly describing a distinct subtopic or section. Headings use natural language that mirrors how people actually ask questions.
- [ ] Common failure: Pages have two or three vague headings, or headings that are clever but not descriptive (like "The Bottom Line" instead of "How Much Does an AI Visibility Audit Cost").
Headings serve as a table of contents for AI engines. When an AI crawler reads your page, it uses H2 headings to understand the page's structure and identify which sections are relevant to specific queries. A page with eight descriptive H2 headings gives the AI eight potential extraction points, eight sections it can cite independently depending on what the user asked. A page with two vague headings gives the AI almost nothing to work with.
Make your headings specific and question-aligned. Instead of "Our Approach," write "How Schema Markup Improves AI Visibility." Instead of "Key Takeaways," write "Five Technical Fixes That Increase AI Citations." Descriptive headings help AI engines match your content sections to specific user queries.
Check 17: Specific Data and Statistics Are Included
- [ ] What to check: Read through your key pages and identify any specific numbers, statistics, percentages, research findings, or data points. Count them.
- [ ] Pass criteria: Key pages include at least five specific data points that support the claims being made. Data is attributed to a source (your own research, a named study, or a reputable organization).
- [ ] Common failure: Pages make claims like "most businesses struggle with AI visibility" without any supporting data, percentages, or specific numbers.
AI engines love data. Specific, attributable statistics make your content more citable because they give the AI something concrete to reference. "18.9% of websites block AI crawlers" is far more citable than "many websites block AI crawlers." The first version is a fact that the AI can extract and present with confidence. The second is a vague claim that the AI has no way to verify or quantify.
Include data from your own research, industry studies, platform-published statistics, and case studies. When you use external data, attribute it clearly. When you use your own data, state the methodology or sample size so the AI can assess its reliability.
Check 18: Comparison Content Exists
- [ ] What to check: Search your site for pages that compare your product, service, or approach against alternatives. These might be "vs" pages, comparison tables, or "alternatives to" content.
- [ ] Pass criteria: You have at least one well-structured comparison page for each major competitor or alternative in your space. Comparison pages include tables, specific feature breakdowns, and honest assessments.
- [ ] Common failure: No comparison content exists at all, or comparison content is overtly biased and lacks specific detail.
Comparison queries are among the most common types of questions people ask AI engines. "X vs Y," "best alternatives to Z," "which is better for small businesses." If you do not have content that addresses these comparison queries, you are ceding those citations entirely to competitors who do.
The key to effective comparison content is specificity and fairness. Compare specific features, pricing, use cases, and outcomes. Be honest about where competitors have advantages. AI engines can detect (and avoid citing) content that is purely promotional without substance. A genuinely useful comparison that helps the reader make an informed decision is exactly the kind of content AI engines want to cite.
Check 19: Content Updated Within 30 Days
- [ ] What to check: Look at the "last updated" or "last modified" dates on your key pages. If you do not display dates, check your CMS for the last edit date.
- [ ] Pass criteria: Key pages have been reviewed and updated within the last 30 days. Updates include new data, expanded sections, corrected information, or additional context based on recent developments.
- [ ] Common failure: Key pages have not been touched in six months or more. Published dates from 2023 or 2024 with no visible update date.
Freshness is a major ranking signal for AI engines. When an AI is choosing between two equally comprehensive sources on the same topic, it will almost always prefer the one that was updated more recently. This is because AI engines need to provide current information, and stale content carries a risk of being outdated.
This does not mean you need to rewrite every page monthly. But you should review and update your key pages at least once a month. Add new statistics, address recent industry changes, update examples, and make sure the information is still accurate. Update the dateModified in your Article schema to reflect the change. This signals to AI crawlers that the content is actively maintained.
Check 20: FAQ Section at Bottom of Key Pages
- [ ] What to check: Scroll to the bottom of your key pages and look for an FAQ section. Verify that the questions are substantive and the answers are thorough.
- [ ] Pass criteria: Key pages have an FAQ section with at least five questions and answers. Questions reflect real queries from your target audience. Answers are detailed (50 to 150 words each) and provide genuine value. FAQPage schema is implemented (cross-reference with Check 11).
- [ ] Common failure: No FAQ section exists, or the FAQ section has two or three generic questions with one-sentence answers.
FAQ sections are citation magnets for AI search. When someone asks ChatGPT a specific question, the AI scans its sources for content that directly matches that question format. An FAQ section literally presents your content as a list of questions with answers. It is the most AI-friendly content format that exists.
The questions should come from real customer interactions. Pull them from support tickets, sales call notes, live chat transcripts, and "People Also Ask" sections in Google. The answers should be thorough enough to stand alone as complete responses. If the AI extracts your FAQ answer and presents it to a user, that answer should be genuinely helpful without needing the rest of the page for context.
Check 21: Internal Links Between Related Pages
- [ ] What to check: Pick five of your key pages and count the internal links on each one. Note whether the links point to related content or are scattered randomly.
- [ ] Pass criteria: Each key page has at least five internal links to related content on your site. Links use descriptive anchor text (not "click here" or "read more"). Links create logical content clusters where related pages reference each other.
- [ ] Common failure: Pages have few or no internal links, or internal links use generic anchor text that provides no context about the linked page.
Internal linking is how AI crawlers understand the relationships between your pages and the overall structure of your expertise. When your blog post about schema markup links to your guide on AI visibility audits, which links to your article about robots.txt configuration, you are creating a content web that demonstrates comprehensive coverage of a topic area. AI engines recognize these clusters and are more likely to view your site as an authoritative source on the topic.
Descriptive anchor text matters because AI engines read the anchor text to understand what the linked page is about. "Learn more about implementing schema markup for AI visibility" tells the AI exactly what it will find on the linked page. "Click here" tells it nothing.
Category 4: Authority Signals (4 Points)
Authority is what separates content that AI engines trust from content they ignore. You can have perfect technical implementation, flawless schema, and well-structured content, and still not earn citations if your site lacks authority signals. These four checks cover the external indicators that tell AI engines your content is worth trusting and citing.
Check 22: Author Bios With Credentials
- [ ] What to check: Check whether your articles and key pages have visible author bylines linked to author bio pages. Verify that bio pages include professional credentials, relevant experience, and links to external profiles.
- [ ] Pass criteria: Every article has a named author with a linked bio page. Bio pages include the author's job title, relevant experience or credentials, a professional headshot, and links to at least two external profiles (LinkedIn, Twitter/X, industry publications, speaking engagements).
- [ ] Common failure: Articles are published under "Admin," "Team," or with no author attribution at all. Or author names are listed but have no linked bio page or credentials.
AI engines evaluate content credibility partly through the author. An article about tax strategy written by a CPA with 15 years of experience carries more weight than the same article published anonymously. AI models can cross-reference author names across the web to build a picture of their expertise. If your author has published on multiple authoritative sites, has a complete LinkedIn profile, and is recognized in their field, their content is more likely to be cited.
Author bios are also a key part of the E-E-A-T signals that Google uses in AI Overviews. Make your authors real, credentialed, and verifiable.
Check 23: Published on Multiple Platforms
- [ ] What to check: Search for your brand name and your authors' names across the web. Check whether your content or expertise appears on platforms beyond your own website.
- [ ] Pass criteria: Your brand or authors appear on at least three external platforms: industry publications, guest posts on authoritative sites, podcast appearances, conference speaking pages, or professional directories.
- [ ] Common failure: All content lives exclusively on your own website with no external presence.
AI engines cross-reference information across sources. When they encounter your brand or your author on multiple reputable platforms, it validates the entity and builds trust. A brand that only exists on its own website is an unknown. A brand that appears on industry publications, has authors contributing to respected outlets, and is mentioned in podcast show notes and conference agendas is an established entity that the AI can trust.
This does not mean you need to be in the New York Times. Industry-specific publications, niche blogs, professional directories, and podcast appearances all count. The key is that your brand and expertise exist beyond your own domain. Even three or four external mentions on relevant platforms can significantly increase your authority signals.
Check 24: Reviews on Third-Party Sites
- [ ] What to check: Search for your business on Google Reviews, G2, Capterra, Trustpilot, Yelp, and any industry-specific review platforms. Note the number of reviews and average rating on each platform.
- [ ] Pass criteria: Your business has reviews on at least two third-party review platforms. Average ratings are above 4.0 stars. Reviews are recent (within the last 90 days). There is a mix of detailed reviews, not just star ratings.
- [ ] Common failure: No reviews exist outside of Google, or all reviews are older than six months.
Third-party reviews are one of the strongest trust signals for AI engines, particularly for local and service-based businesses. When an AI is deciding whether to recommend your business, it checks what real people have said about you across the web. Consistent positive reviews across multiple platforms create a strong trust signal. A complete absence of reviews, or reviews that are exclusively on one platform, is a weaker signal.
Encourage customers to leave reviews on multiple platforms. Respond to reviews thoughtfully. The content in reviews and review responses adds keyword-rich, experience-based text that AI engines can process. A detailed review that says "They helped us improve our AI visibility score from 12 to 67 in three months" is exactly the kind of specific, experience-based content that AI engines find credible.
Check 25: Backlinks From Authoritative Sources
- [ ] What to check: Use a backlink analysis tool (Ahrefs, Moz, SEMrush, or a free alternative like Ubersuggest) to review your backlink profile. Identify links from high-authority domains.
- [ ] Pass criteria: Your site has backlinks from at least 10 unique domains. At least three of those domains are recognized authorities in your industry (Domain Authority 50 or higher). Backlinks are to your content pages, not just your homepage.
- [ ] Common failure: Few or no backlinks from authoritative sources. All backlinks point to the homepage. Backlink profile is dominated by low-quality or irrelevant sites.
Backlinks remain one of the strongest authority signals on the web, and AI engines use them heavily. When an authoritative website links to your content, it is a vote of confidence that AI engines can measure and weight. The quality of the linking domain matters far more than the quantity. One backlink from an authoritative industry publication is worth more than 50 backlinks from random directories.
Focus on earning backlinks to your content pages, not just your homepage. When your guide on schema markup earns a backlink from a respected marketing publication, that signals to AI engines that your schema markup content specifically is authoritative. This makes it more likely to be cited when someone asks an AI about schema markup.
Your Scoring Framework
Now that you have checked all 25 points, here is how to interpret your results.
Count your passes. Each item is worth one point, for a maximum of 25.
- 21 to 25 points: Your site is well-positioned for AI visibility. Focus on maintaining what you have and incrementally improving any partial scores.
- 16 to 20 points: You have a solid foundation but meaningful gaps. Prioritize the failures in the Technical and Schema categories first, as those tend to have the highest impact per fix.
- 11 to 15 points: Significant work needed. Start with the Technical category and work through each category in order. Most sites in this range have two or three high-impact blockers that, once fixed, will shift several other scores.
- 6 to 10 points: Major overhaul required. Consider this checklist a roadmap for a multi-week improvement project. Tackle it category by category, and recheck your scores after each round of fixes.
- 0 to 5 points: Your site is essentially invisible to AI search. The good news is that almost every fix on this list will move the needle. Start with Check 1 (robots.txt) and work forward.
For teams that want to skip the manual process, GetCited automates this entire audit and delivers a scored report with prioritized recommendations. But whether you use a tool or do it manually, the 25 points on this checklist are what matters.
How to Prioritize Your Fixes
Not all 25 items are created equal. Here is a practical prioritization framework based on impact and effort.
High Impact, Low Effort (Fix These Today)
These items can be fixed in minutes and have an immediate effect on your AI visibility:
- robots.txt access (Check 1): Unblocking AI crawlers is a two-minute fix with immediate impact.
- SSL enabled (Check 6): If your certificate is expired, renewing it takes less than an hour.
- Canonical tags (Check 8): A quick template change in most CMS platforms.
- Organization schema (Check 9): A single block of JSON-LD on your homepage.
High Impact, Medium Effort (Fix These This Week)
These take a few hours each but produce significant results:
- llms.txt (Check 2): Writing and publishing this file takes 30 to 60 minutes.
- Article schema on blog posts (Check 10): Can be automated through your CMS once set up.
- First paragraph optimization (Check 14): Rewriting opening paragraphs on your top 10 pages takes a few hours.
- FAQ sections with schema (Checks 11, 20): Adding FAQ sections to existing pages and implementing schema.
High Impact, High Effort (Plan These for This Month)
These require more substantial work but are critical for long-term AI visibility:
- Content depth to 2,500+ words (Check 15): Expanding key pages requires research and writing time.
- Comparison content (Check 18): Creating new comparison pages from scratch.
- Author bios and schema (Checks 12, 22): Building out author profiles with credentials and external links.
- Content freshness (Check 19): Establishing a monthly update cadence for key pages.
Medium Impact, Variable Effort (Ongoing Improvements)
These are important but tend to compound over time rather than providing an immediate boost:
- Backlinks from authoritative sources (Check 25): Link building is a long-term effort.
- Published on multiple platforms (Check 23): Building external presence takes months.
- Third-party reviews (Check 24): Review acquisition is a steady, ongoing process.
Common Mistakes When Running This AI Visibility Audit Checklist
After watching hundreds of businesses work through this checklist (or a version of it), we have seen the same mistakes repeated. Here are the ones to watch for.
Stopping at the Technical category. Some teams fix their robots.txt, add an llms.txt file, clean up their sitemap, and consider themselves done. Technical fixes are necessary but not sufficient. The Technical category opens the door. The Content and Authority categories are what earn the citation. Do not stop at eight points when there are 25.
Treating this as a one-time project. AI visibility is not a set-it-and-forget-it metric. The landscape changes constantly. AI engines update their models, competitors improve their content, and new players enter your space. Plan to re-run this checklist at minimum every quarter, with monthly reviews of the Content category items that require ongoing maintenance.
Ignoring the Content category because it is harder. Schema and technical fixes are satisfying because they are concrete and measurable. Content improvements are harder to quantify in the short term. But content is what AI engines are actually citing. Technical infrastructure and schema make your content accessible and understandable. Content quality is what makes it worth citing.
Over-optimizing FAQ sections with promotional content. Your FAQ sections should answer real questions with genuine, helpful responses. Turning every FAQ answer into a sales pitch undermines the entire purpose. AI engines can detect (and will avoid) content that prioritizes promotion over helpfulness.
Neglecting author credibility. Many businesses pour resources into content quality but publish everything under a generic brand name or "Admin" account. Author credibility is one of the easiest authority signals to build, and one of the most impactful. Put real names on your content, build out their profiles, and connect them to external platforms.
The Printable Version
Here is the condensed version you can print and pin to your wall. Check each box as you verify it.
TECHNICAL FOUNDATION - [ ] 1. robots.txt allows GPTBot, PerplexityBot, ClaudeBot, Google-Extended - [ ] 2. llms.txt file exists and is well-structured - [ ] 3. sitemap.xml is current and comprehensive - [ ] 4. Page speed under 3 seconds (mobile and desktop) - [ ] 5. Site is mobile-friendly - [ ] 6. SSL enabled with valid certificate - [ ] 7. No JavaScript rendering issues blocking content - [ ] 8. Canonical tags set correctly on all pages
SCHEMA MARKUP - [ ] 9. Organization schema on homepage - [ ] 10. Article schema on blog posts - [ ] 11. FAQ schema on key pages - [ ] 12. Author (Person) schema on articles - [ ] 13. HowTo schema on tutorials and guides
CONTENT STRUCTURE - [ ] 14. First paragraph answers the query directly - [ ] 15. 2,500+ words on key pages - [ ] 16. 8+ H2 headings per key page - [ ] 17. Specific data and statistics included - [ ] 18. Comparison content exists - [ ] 19. Content updated within 30 days - [ ] 20. FAQ section at bottom of key pages - [ ] 21. Internal links between related pages (5+ per page)
AUTHORITY SIGNALS - [ ] 22. Author bios with credentials on bio pages - [ ] 23. Published on at least 3 external platforms - [ ] 24. Reviews on 2+ third-party sites - [ ] 25. Backlinks from 10+ unique domains (3+ high authority)
SCORE: _____ / 25
Frequently Asked Questions
How long does it take to complete this full AI visibility audit checklist?
If you are checking every item manually and documenting your findings, expect the full 25-point audit to take four to six hours for a single website. The Technical category is the fastest, usually 30 to 60 minutes. Schema checks take about an hour with a validation tool. Content checks take the longest because they require reading and evaluating your actual content on multiple pages. Authority checks take about an hour and require backlink analysis tools. For teams that need faster results, GetCited runs the equivalent of this audit automatically and delivers a scored report in minutes, but the manual process gives you a deeper understanding of each issue.
Do I need to pass all 25 points to get cited by AI search engines?
No. Very few websites score a perfect 25 out of 25, and you do not need a perfect score to earn citations. What matters more is eliminating the binary blockers (especially in the Technical category) and being stronger than your competitors across the full checklist. If your competitor scores 14 out of 25 and you score 18 out of 25, you have a meaningful advantage. Our data shows that sites scoring 16 or above on this checklist are cited at rates three to four times higher than sites scoring below 10. Focus on getting above that 16-point threshold first, then incrementally improve.
How often should I re-run this audit?
Re-run the full 25-point audit quarterly. Between full audits, do a monthly check of the items that change most frequently: content freshness (Check 19), sitemap status (Check 3), and third-party reviews (Check 24). Also re-check your robots.txt (Check 1) after any website updates or hosting changes, as these sometimes reset crawler access rules without warning. AI engines update their models and ranking factors regularly, so what works today might need adjustment in three months. Treat this as an ongoing process, not a one-time project.
What is the difference between this AI search checklist and a traditional SEO audit?
A traditional SEO audit focuses on factors that influence your ranking in Google's organic search results: keyword density, meta tags, backlink volume, page speed, and domain authority. This AI visibility audit checklist shares some of those elements but adds factors that are unique to AI search: crawler-specific robots.txt rules, llms.txt files, schema types that AI engines prioritize, content structure patterns that earn citations (like direct-answer opening paragraphs), and authority signals that AI engines weight differently than Google does. The biggest difference is the goal. SEO audits aim to get you on page one. This checklist aims to get you cited in AI-generated answers, which is a fundamentally different outcome.
Can I use this checklist for any type of website, or is it specific to certain industries?
This checklist works for any type of website that wants to be visible in AI search results: SaaS companies, e-commerce stores, local businesses, professional services firms, publishers, and personal brands. The 25 points are universal. What changes across industries is which points matter most. Local businesses should weight Check 24 (third-party reviews) more heavily. SaaS companies should prioritize Check 18 (comparison content). Publishers should focus extra attention on Checks 10, 12, and 22 (Article schema, Author schema, and author bios). Use the checklist as a complete framework, but adjust your prioritization based on your business type and the queries your audience is actually asking AI engines.