A website can be technically accessible to search engines and still struggle to appear in search results or be cited in AI-generated answers. An effective audit looks beyond rankings and checks whether your pages can be discovered, understood, trusted, and used to answer relevant questions.
SEO, GEO, and AEO overlap, but they do not focus on exactly the same outcomes. SEO examines organic search visibility and the technical foundations behind it. Generative Engine Optimization (GEO) considers how content and brand information may be represented in generative search experiences. Answer Engine Optimization (AEO) focuses on making useful, direct answers easy to identify and extract.
This guide provides a practical audit framework for all three. Instead of simply asking whether a website has headings, metadata, or structured data, it identifies what to inspect, what evidence to collect, and what to do when an issue appears.
Can search engines crawl, index, and rank the right pages?
Can generative systems understand and accurately represent your content and brand?
Can answer engines identify concise, relevant answers to users' questions?
1. What Does Each Audit Actually Check?
Start by defining the questions each part of your audit needs to answer. This avoids treating every technical issue as a ranking problem or assuming that appearing in an AI-generated answer can be achieved through one particular optimization technique.
| Audit area | Core question | What to inspect | Evidence to collect |
|---|---|---|---|
| SEO | Can relevant pages be discovered and perform in organic search? | Crawlability, indexing, page experience, content relevance, internal links, and search performance. | Crawl reports, indexation checks, Search Console queries, impressions, clicks, and rankings. |
| GEO | Can generative systems find and accurately interpret the site's information? | Accessible source content, entity clarity, evidence, brand consistency, topical coverage, and external references. | AI search visibility tests, source citations, brand mentions, content checks, and referral analytics where available. |
| AEO | Can a system identify a clear and useful answer to a specific question? | Question coverage, answer placement, concise explanations, page structure, and relevant structured data. | Question-to-page mapping, answer extraction tests, rich result eligibility, and search appearance data. |
2. Audit the Technical Foundation First
Before evaluating individual search or AI visibility opportunities, confirm that the website's important content can be accessed and interpreted. A page that cannot be crawled, is unintentionally excluded from indexing, or returns an error may be unavailable to some search experiences regardless of how well written it is.
URL response and crawl accessibility
Identify URLs that cannot be reliably retrieved or are unnecessarily blocked.
Export URLs returning 4xx or 5xx responses, and separate genuine errors from intentional removals. Investigate internal links pointing to 404 pages and any important pages unexpectedly returning 302 or 307 responses.
Review whether product, service, category, article, and other priority URLs are accidentally blocked. Check that the robots.txt file is available and does not contain conflicting rules that prevent intended crawling.
Compare a normal browser visit with a crawl using the relevant user agent. Where a search or AI system is important to your use case, verify its documented crawler behavior and access rules instead of assuming every bot follows identical instructions.
Trace the final destination of important URLs. Find internal links that point to redirecting URLs, redirect loops, and chains that pass through several intermediate destinations. Update internal links to the final canonical destination where appropriate.
Indexability and canonical signals
Confirm that the pages you want discovered are eligible for indexing and identify duplicate URL variants.
Compare meta robots and X-Robots-Tag directives with the site's intended indexing policy. Pay special attention to recently published pages, migrated sections, and templates that might have inherited a development setting.
Check for missing, malformed, conflicting, or unexpected canonical URLs. Look for canonical tags pointing to redirect destinations, error pages, unrelated content, or a different version of a page than intended.
Related guide: Canonical Tag Issues: Common Problems and Fixes.
Check whether the sitemap includes the intended canonical, indexable URLs and excludes error pages, redirects, and non-canonical duplicates. Validate the sitemap's location and its submission in the relevant search engine tools.
Related guide: XML Sitemap Errors: Common Problems and Fixes.
Compare sitemap URLs, known landing pages, and the crawl's discovered pages. Investigate priority pages with no internal links, especially when they are expected to receive search traffic or support a key conversion journey.
Related guide: Orphan Pages: How to Find and Resolve Them.
Rendering and mobile usability
Verify that the page's meaningful content and essential functions are available to visitors and relevant crawlers.
Check whether the main heading, article text, product details, links, and important metadata appear in the rendered page. If content depends on JavaScript, test that the rendering process completes successfully and that key content is not hidden behind failed API calls or client-side errors.
Inspect representative mobile URLs for layout shifts, slow loading, and delayed interactions. Review field data where available, rather than relying exclusively on a single lab test. Check whether intrusive overlays or broken responsive layouts obscure the main content.
For a more detailed walkthrough of these foundations, see Technical SEO Audit: What to Check and How to Start and How to Check Website Crawlability for Google, Bing, and AI Bots.
3. SEO Audit Checklist: What to Check Beyond Technical Errors
Once the technical foundation is understood, examine whether each important page targets a clear purpose and whether search engines have enough information to understand its relevance. A useful SEO audit connects technical findings to page intent, internal linking, and actual search performance.
Titles, headings, and page-level relevance
Identify pages where the search snippet and page structure fail to communicate the page's actual purpose.
Find missing or duplicated titles, titles that describe the wrong topic, and important pages with vague wording. Compare titles against the main content and the queries the page is meant to serve. Review search snippets and performance data to determine which pages deserve rewriting.
See Duplicate Title Tags and Missing Title Tags.
Identify missing main headings, multiple H1s that create ambiguity, headings that do not describe the sections below them, and heading sequences that skip important topics. The goal is to make the document easier to navigate and understand, not to enforce an arbitrary heading count.
Related: Missing H1 Tags and Multiple H1 Tags.
Find missing, duplicated, or irrelevant descriptions. Check whether the copy communicates the page's actual value and matches the likely query intent. Treat length recommendations as writing guidelines rather than fixed ranking requirements, because search engines may generate their own snippets.
See Duplicate Meta Descriptions and Missing Meta Descriptions.
Group pages that target the same intent or contain substantially repeated material. Distinguish legitimate similar pages, such as product variants, from unnecessary duplication. Review whether a page offers unique information or simply repeats what is already covered elsewhere on the site.
Related: Duplicate Content SEO and Thin Content.
Internal linking and content relationships
Find gaps between important pages and the internal paths that help visitors and crawlers discover them.
Identify links pointing to non-existent pages, redirected URLs, and outdated destinations. Prioritize broken links from high-traffic pages, navigation, and pages that support important user journeys.
Read Broken Link Audit and How to Find and Fix Broken Internal Links.
Review which relevant pages link to a priority URL and what the surrounding anchor text communicates. Look for key pages that are buried several clicks deep or have only generic links that provide little context.
Compare the site's informational pages with its service or product pages. Confirm that useful supporting articles link naturally to relevant deeper resources and that visitors can move between introductory, explanatory, and actionable content without encountering dead ends.
4. GEO Audit Checklist: Can Generative Systems Understand Your Website?
GEO auditing focuses on the information that could help a generative system discover, interpret, and potentially use your content. It is not simply a matter of adding an AI-related file or inserting a particular keyword. The audit should examine whether your website provides accessible, accurate, well-organized information and whether your brand is represented consistently.
Content accessibility for AI-related crawlers
Check whether relevant information is technically accessible to the systems you care about.
Identify the AI-related bots relevant to your objectives and inspect the applicable robots.txt rules, server responses, and firewall behavior. Different services may use different crawlers and access policies, so do not assume that allowing one bot enables all AI search systems to retrieve your content.
Check whether important explanatory content, definitions, pricing information, product details, and business information are accessible in the rendered page. Look for critical content hidden behind broken JavaScript, blocked resources, or interaction patterns that prevent access.
If you publish an llms.txt file, check that it is accessible, current, and accurately points to relevant resources. Do not assume that every AI system reads or uses it. It should not replace accessible web pages, useful internal navigation, or the site's established search and crawling infrastructure.
Related article: The Two Sides of the llms.txt Debate.
Brand and entity clarity
Make it easier to determine which organization, product, person, or service a page is actually about.
Compare the homepage, About page, service descriptions, contact information, and relevant social or business profiles. Identify conflicting brand names, outdated descriptions, inconsistent locations, or different explanations of what the business offers.
Review important claims, statistics, specifications, and statements about services. Where appropriate, cite primary sources, include dates for time-sensitive information, and identify who is responsible for the content. Distinguish original findings from summaries of other sources.
Where appropriate, validate Organization, Product, Article, Breadcrumb, or other applicable schema against the visible page content. Check for incorrect names, obsolete properties, mismatched URLs, and structured data that claims information the page does not actually support.
Search for the brand and its important products or services in relevant publications, directories, industry resources, and publicly accessible sources. Record the source, the wording used, and any discrepancies. Do not assume that every mention carries equal credibility or that mentions alone cause AI visibility.
This is an audit framework, not a guaranteed sequence or ranking formula used by AI systems.
5. AEO Audit Checklist: Can Your Pages Answer Specific Questions?
Answer Engine Optimization is concerned with the relationship between a user's question and the information on your page. The most useful audit is not a generic word-count or formatting check. It tests whether the page answers a particular question clearly, accurately, and in a context that can be understood without guessing.
Question coverage and answer usefulness
Test the actual questions your audience asks and check how well your pages respond.
Collect questions from search queries, customer support conversations, sales discussions, site search, and relevant audience research. Map each question to the page that should answer it. Identify unanswered questions, competing pages, and questions that need a new dedicated resource.
For important question-led pages, read the opening section and the relevant subheading. Can someone identify the main answer without searching through unrelated introductions, promotional language, or multiple unrelated paragraphs? Keep essential qualifications close to the answer rather than hiding them far below.
Replace vague explanations with relevant details, concrete steps, definitions, examples, or conditions. For example, a guide about a 404 audit should explain how to identify broken URLs, trace the internal links pointing to them, and decide whether to fix, redirect, or remove the links.
Check whether a single page mixes definitions, troubleshooting, comparisons, and implementation steps without clear organization. Use descriptive subheadings where they genuinely help visitors find different answers. Create separate pages when different intents require substantial, distinct coverage.
Answer structure, formatting, and structured data
Make relevant information clear without adding markup or formatting solely for a presumed search feature.
Where relevant, ensure a definition states what the subject is, instructions include prerequisites and important steps, and comparisons explain the differences being discussed. Tables should use meaningful headings and accurately reflect the text around them.
Check whether question-and-answer sections address real audience needs rather than repeating keywords. Validate any applicable structured data against the actual visible content and current search engine eligibility rules. Do not promise rich results simply because markup is present.
Test a paragraph or short section in isolation. Does it clearly identify what it is describing? Avoid unexplained pronouns, ambiguous references, and claims that rely entirely on an image or an earlier section to make sense. Keep the surrounding context available where a standalone answer could be misleading.
Review eligible rich result features and the appearance of important pages in search. Use available Search Console reporting and live search checks, remembering that feature availability varies and not every well-structured answer will receive a featured snippet or other enhanced display.
For a deeper introduction to this topic, see Answer Engine Optimization (AEO): Definition and Strategies.
6. Content Quality and Trust Checks Shared by SEO, GEO, and AEO
A page can have a clean crawl report and still provide a poor answer. This part of the audit evaluates whether the content is accurate, distinctive, current, and genuinely helpful to the intended audience.
Evidence and attribution
Check important factual claims, statistics, product specifications, and quotations. Record their sources and verify that the cited evidence supports the wording used.
First-hand value
Identify original research, real examples, demonstrations, expert explanations, and other information that adds something beyond a generic summary.
Freshness and accuracy
Review time-sensitive information, expired offers, outdated instructions, changed products, and old screenshots. Update or qualify information when it is no longer current.
Author and business identity
Check whether readers can understand who created the content, what experience or expertise is relevant, and how to contact the business when appropriate.
Content-level verification
Look for the issues that automated crawlers cannot fully judge.
Choose a sample of high-impact pages and trace key factual claims to their source. Flag unsupported statistics, outdated statements, and wording that implies certainty beyond what the evidence supports.
Compare the page's title, introduction, headings, and actual content. A page titled as a complete audit checklist should cover the relevant audit stages rather than provide only a handful of generic tips.
Identify repeated explanations that add no new value, broad claims that combine different concepts, and sections that fail to explain the conditions under which advice applies.
Check that important information in visual assets is also available as readable text where necessary. Provide appropriate alternative text for meaningful images and ensure that diagrams and tables are understandable on smaller screens.
7. Measure the Results of Your Audit
A completed checklist is not proof that a website has improved. Record a baseline before making changes, then monitor the specific indicators related to the problem you are addressing. Use the same definitions, tools, and comparison periods where possible.
Crawl and index health
Errors on priority URLs, unintended noindex directives, internal broken links, and crawl coverage of important sections.
Organic search performance
Impressions, clicks, CTR, query coverage, and relevant landing-page performance over a defined period.
AI search observations
Brand and page mentions, citation examples, referral sessions where measurable, and accuracy of observed answers.
Question coverage
Questions mapped to relevant pages, answer completeness, rich result eligibility, and observed answer presentation.
| Metric or observation | Where to get it | What it can tell you | What it cannot prove alone |
|---|---|---|---|
| Broken internal links | Website crawler | Where internal navigation points to broken destinations. | Whether fixing the link will increase rankings. |
| Indexing and crawl status | Search Console and crawler | Whether important pages appear accessible and how search systems report their status. | Whether a page will rank for a particular query. |
| Organic impressions and clicks | Search Console | Changes in reported organic search exposure and traffic. | Which individual site change caused the movement. |
| AI mentions and citations | Consistent manual testing and suitable analytics | Examples of how a brand or page is represented in selected AI search experiences. | Complete AI visibility across all systems or users. |
| Question coverage and answer quality | Editorial review and query mapping | Whether important audience questions have relevant, complete answers. | Guaranteed appearance in a featured snippet or generated answer. |
AI-generated results can vary by system, query wording, location, language, personalization, and time. Record the test date, platform, exact question, observed response, and cited sources whenever you perform an AI visibility check. Treat a small sample as an observation, not a universal visibility measurement.
8. How to Prioritize Audit Findings
A long audit report can make every issue look equally urgent. Prioritize problems by the affected pages, the seriousness of the failure, the number of users or crawlers potentially affected, and the effort needed to resolve the issue.
Examples: the main service pages return 5xx errors, an important site section is accidentally blocked, or a deployment adds noindex to pages meant for search.
Examples: priority pages are orphaned, internal navigation points to broken URLs, or canonical tags consistently direct search systems away from the intended page.
Examples: overlapping pages target the same intent, titles are misleading, important questions lack direct answers, or structured data does not match visible content.
Examples: small copy edits, non-critical formatting inconsistencies, and cosmetic adjustments to pages that already serve their intended purpose.
9. Recommended SEO, GEO, and AEO Audit Workflow
A repeatable workflow makes it easier to distinguish newly discovered problems from old ones and to evaluate whether fixes have worked. Use the following sequence for a new site audit or a scheduled review.
Define the audit scope
Set the target domain, relevant subdomains, priority directories, important page types, and the search experiences you intend to evaluate.
Start with the homepage, key services or products, major category pages, important articles, and conversion pages. Add the audience questions that these pages should answer.
Capture the baseline
Save the initial crawl, important search performance data, and representative AI visibility observations before changes begin.
Document crawl settings, selected URL scope, date range, relevant search properties, AI platforms tested, and the questions used for manual checks. This makes later comparisons more meaningful.
Run technical and content checks
Identify the issues affecting access, indexing, internal relationships, page relevance, and answer quality.
Keep the source URL, destination URL where relevant, observed problem, evidence, recommended action, and priority together. This prevents an issue list from becoming a disconnected collection of counts.
Fix and validate
Resolve the highest-priority issues and verify that the actual problem has been corrected.
After deployment, verify the updated status codes, canonical tags, directives, internal links, rendered content, or page copy. Check for unintended side effects before closing the finding.
Compare results and plan the next review
Separate completed technical fixes from changes in actual search and user outcomes.
Use consistent crawl scope and settings where possible. Compare the count and identity of affected URLs, then evaluate longer-term search performance and selected AI observations separately. Keep unresolved issues visible rather than removing them from the record.
10. Use SiteAuditLint to Review and Compare Technical Findings
A website crawler can help turn the technical portion of this checklist into a repeatable process. For example, a crawl can identify broken internal links, missing or duplicated metadata, heading issues, redirect problems, and other page-level technical findings across a site.
SiteAuditLint is designed for desktop website crawling and technical SEO auditing on Windows. Use it to collect a baseline, review the URLs affected by technical issues, and compare later crawls to see which problems remain and which have been resolved. Editorial assessments, search performance analysis, and manual AI answer testing still require their own evidence and review.
| Technical issue | Baseline crawl | Follow-up crawl | Difference |
|---|---|---|---|
| Broken internal links | 42 | 8 | −34 |
| Missing titles | 25 | 5 | −20 |
| Missing descriptions | 30 | 12 | −18 |
| Redirect errors | 16 | 4 | −12 |
| Total listed issues | 113 | 29 | −84 |
Example comparison only. Issue counts are not equivalent to ranking improvements, and different issue types may have different levels of impact.
Explore the SiteAuditLint features or read the desktop SEO crawler comparison guide to learn more about technical crawl workflows.
Frequently Asked Questions
Are SEO, GEO, and AEO audits the same thing?
No. They share technical and content foundations, but their evaluation questions differ. SEO examines organic search visibility and its supporting signals. GEO examines how content and entities may be interpreted and represented in generative search experiences. AEO checks how effectively pages address specific questions and provide clear answers.
Can one technical SEO crawler perform a complete GEO and AEO audit?
A crawler can help with technical accessibility, metadata, page structure, and internal linking. It cannot fully evaluate factual accuracy, answer quality, brand representation, or visibility across AI systems by itself. A complete audit combines crawl data with content review, search performance data, and relevant manual testing.
Does adding llms.txt make a website visible in AI search?
Not by itself. An llms.txt file may provide a curated list of resources for systems that choose to use it, but it is not a universal requirement or a guarantee of inclusion in AI-generated answers. Keep the site's actual content accessible, useful, and clearly organized.
How often should a website undergo an SEO, GEO, and AEO audit?
Frequency depends on how often the website changes, its size, the importance of its search traffic, and the risks of technical errors. A site undergoing a migration or major redesign may need audits before and after deployment. Other sites can establish a regular schedule for technical crawls and review search and AI visibility observations at appropriate intervals.
What should I fix first after an audit?
Start with issues that prevent important pages from being accessed, crawled, or indexed as intended. Next, address serious internal linking, canonical, and content problems affecting high-priority URLs. Then work through content clarity, question coverage, and other improvements according to the evidence and resources available.
How do I know whether my GEO and AEO improvements are working?
Use a documented baseline and consistent tests. Track relevant AI mentions and citations, referral activity where measurable, and the accuracy of observed representations. For AEO, review question coverage, answer quality, and applicable search appearances. Keep these measures distinct from traditional organic search metrics and avoid drawing conclusions from a few isolated observations.
Start with a Technical Audit You Can Repeat
Before optimizing for more search experiences, find out whether your important pages are accessible, correctly structured, and connected. Build a baseline, fix the issues that matter, and compare the next crawl to see what changed.
Explore SiteAuditLintThis checklist is intended as a practical audit framework. Search engine and AI system behavior can change, and no single technical setting, content format, or audit result guarantees rankings, citations, or inclusion in generated answers.