What an SEO audit should accomplish
An SEO audit is a structured investigation of how well a website helps search engines discover, understand, and evaluate its pages while helping visitors complete their goals. It should produce more than a list of warnings from a tool. A useful audit connects evidence to business priorities, identifies root causes, distinguishes confirmed defects from hypotheses, and gives the team a practical order of work. The scope may include technical accessibility, indexation, site architecture, content quality, on-page signals, internal links, performance, structured data, and measurement. The outcome should be a prioritised plan with owners and verification steps.
Start with goals and scope
Before running tools, define what the website is trying to achieve. A service business may care about qualified enquiries and local discovery, while a publisher may care about relevant readership and content visibility. Record key conversions, priority services or categories, important landing pages, target markets, and known site changes. Agree on whether the audit covers the whole domain, a subdomain, a directory, a migration, or a specific page template. A clear scope prevents teams from spending hours on low-impact warnings while overlooking defects on revenue-driving pages.
Build an inventory of URLs
Create an initial URL inventory using the XML sitemap, internal crawl, analytics landing pages, Search Console exports, and known legacy URLs. These sources overlap but are not identical. Sitemaps show URLs the site declares important for discovery; crawlers show reachable links; analytics reveals URLs visitors use; Search Console provides search-related observations; legacy exports help identify redirects and old content. Combine them carefully, deduplicate variants, and preserve source labels so you know why each URL is in the audit. Do not assume a single source represents the entire site.
Crawl the site methodically
Use a crawler to record status codes, redirect chains, titles, meta descriptions, headings, canonicals, indexability directives, internal links, image attributes, and template patterns. Configure crawl speed responsibly so the audit does not overload production. Crawl both the main domain and relevant subdomains when they are in scope. For JavaScript-heavy websites, compare raw HTML with rendered output. Save the crawl settings and date so the same approach can be repeated after fixes. A crawl is a snapshot, not a complete explanation of why each issue exists.
Validate what the crawler reports
SEO tools can misinterpret intentional behaviour, incomplete data, or site-specific architecture. Verify important findings with direct requests, browser developer tools, server logs, Search Console, and a human review of the page. For example, a missing title on a blocked route may be less urgent than a canonical mismatch on a high-value landing page. A crawler may flag duplicate text that is appropriate for a product specification. Treat tool output as evidence to investigate, not an automatic instruction to change the website.
Check robots.txt and crawl access
Review the site’s robots.txt file and test whether its rules align with the intended crawl strategy. Check for accidentally blocked important directories, CSS, JavaScript, images, or routes. Remember that robots.txt controls crawling, not guaranteed removal from search results. If a page needs to be deindexed, choose a suitable indexation control and ensure the crawler can process it. Inspect environment-specific rules so staging restrictions are not accidentally copied into production or removed from private staging environments.
Review XML sitemaps
Sitemaps should generally list canonical URLs that return successful responses and are intended for discovery. Look for redirects, 404s, 5xx responses, noindex pages, duplicate variants, and URLs that canonicalise elsewhere. Check whether sitemap indexes point to accessible child files and whether lastmod values represent meaningful updates. A sitemap can support discovery, but it does not guarantee crawling or indexing. Compare sitemap entries with the crawl and the site’s content inventory to identify missing important pages and unnecessary URL noise.
Assess HTTP status codes
Important pages should return the intended response consistently. Check for unexpected 404s, soft 404s, redirect chains, loops, and persistent server errors. A genuine missing URL may correctly return 404 or 410; it is not necessary to redirect every removed page. Where a page has moved permanently, send visitors to the closest relevant replacement and update internal links to the final URL. For 5xx issues, investigate logs and infrastructure rather than repeatedly recrawling without addressing the underlying cause.
Canonicalisation and duplicate URLs
Inspect canonical tags for missing, contradictory, or invalid targets. Confirm that canonical destinations are accessible, relevant, and consistent with internal links and sitemaps. Duplicate URLs may be caused by tracking parameters, sorting, filters, alternate hostnames, trailing slashes, or content-management settings. Decide which variants should be consolidated, redirected, canonicalised, or left distinct based on user value and content differences. A canonical tag is a hint, not a guaranteed command, and it does not replace redirect behaviour when a URL has permanently moved.
Indexation coverage and page intent
Use Search Console and representative URL inspection to understand which important pages are indexed, excluded, or selected with an unexpected canonical. Separate pages that are intentionally excluded from pages that should be discoverable. Investigate patterns across templates rather than treating every excluded URL as a defect. A page may be crawled but not indexed because it is duplicative, low-value, or not sufficiently useful for the query space. Improve the page’s purpose and supporting architecture where appropriate instead of assuming every indexing issue needs a technical directive.
Site architecture and click depth
Review how users and crawlers reach priority pages. Important services, categories, and resources should be linked from relevant navigation, category pages, breadcrumbs, or contextual content. Look for orphan pages, confusing taxonomy, unnecessarily deep paths, and broad navigation menus that obscure key destinations. Click depth is a diagnostic clue, not an absolute ranking rule. Improve architecture based on user tasks and topical relationships. Avoid adding links indiscriminately; the goal is clear navigation and meaningful relationships, not the largest possible number of links.
Internal links and anchor text
Find internal links that point to redirects, error pages, non-canonical variants, or irrelevant destinations. Repair links at their source, including menus, breadcrumbs, related-content modules, footer links, and editorial copy. Anchor text should describe the destination naturally, helping users understand what they will find. Avoid repeating exact-match phrases mechanically or using generic labels everywhere. Check for broken links introduced by slug changes and content migrations. When fixing many links, use backups and carefully scoped replacements so unrelated URLs are not changed.
Title tags and meta descriptions
Review titles and meta descriptions for clarity, distinctiveness, and alignment with page intent. Important pages should communicate their topic and value without keyword stuffing or unsupported promises. Duplicate titles may indicate template problems or pages that are too similar, but some duplication is a symptom rather than the root cause. Meta descriptions can influence how a result is presented, although search engines may choose different snippets. Prioritise high-value pages and page groups, and use a scalable template strategy where manually writing every description is not practical.
Headings and content structure
Check whether each page has a clear topic and a useful heading hierarchy. Headings should organise the content for readers rather than serve as containers for repeated keywords. In a WordPress site where the post title already supplies the H1, avoid adding an unnecessary second H1 in the body. Review headings for missing context, duplicate sections, abrupt jumps, and vague labels. A good structure helps readers scan the page and helps maintainers see whether the content actually answers the intended question.
Content usefulness and topical overlap
Audit content for accuracy, completeness, originality, and alignment with the intended audience. Identify pages that are outdated, thin, repetitive, or competing for the same intent. Consolidation can help when several pages answer essentially the same question, but do not merge content solely because tools show similar words. Preserve unique information, useful backlinks, and distinct audience needs. Refresh facts and examples, remove obsolete claims, and provide clear next steps. Content changes should make the page more useful, not simply longer.
Search intent and keyword mapping
Map important queries and topics to the pages that best satisfy them. Look for mismatches such as a broad service page targeting a very specific troubleshooting question, or several articles competing for the same main intent. Use Search Console query and landing-page data where available, and combine it with customer questions and competitive research. Avoid assigning a unique page to every minor keyword variation. A coherent content architecture should cover meaningful needs with clear page roles and natural internal links.
E-E-A-T and trust signals
For topics where experience, expertise, authoritativeness, and trust matter, make it easy for readers to understand who created the content, what expertise supports it, and how claims can be checked. Provide accurate author information, transparent business details, relevant references, contact paths, and clear policies. Avoid fabricating credentials, testimonials, statistics, or first-hand experience. Trust signals should be genuine and consistent with the business. They support reader confidence but do not guarantee rankings, and they cannot compensate for inaccurate or unhelpful content.
Image SEO and media accessibility
Check that important images load reliably, have suitable dimensions, and do not unnecessarily slow the page. Use descriptive alt text when an image conveys information; decorative images should not receive keyword-stuffed descriptions. Review responsive image delivery, compression, file naming, and image sitemap needs where relevant. Confirm that important images are not blocked and that product or service images have appropriate context. Video and embedded media should have a useful surrounding explanation and accessible controls or alternatives where needed.
Core Web Vitals and user experience
Review field data and lab diagnostics for loading performance, responsiveness, and layout stability. LCP, INP, and CLS can help identify friction, but they are not a complete measure of site quality. Inspect actual page templates and user journeys, including navigation, filters, forms, and mobile menus. Prioritise changes that address measured bottlenecks while preserving accessibility, privacy, and functionality. Use the official PageSpeed Insights and Search Console reports where available, and compare results after deployment instead of treating a single test score as conclusive.
Mobile usability and accessibility
Test key pages at common mobile widths and on representative devices. Check navigation, tap targets, form controls, readable text, sticky elements, cookie banners, and content that overflows horizontally. Review keyboard navigation, focus visibility, labels, contrast, and semantic structure. Accessibility is not just an SEO checkbox; it helps people use the website successfully. Avoid fixes that improve a metric while breaking assistive technology or hiding important content. Test important tasks end to end rather than relying solely on responsive screenshots.
Structured data and rich-result eligibility
Validate structured data against the actual visible page content and current search-engine guidelines. Check for syntax errors, invalid property values, misleading markup, and structured data that describes content users cannot see. Not every schema type leads to a rich result, and valid markup does not guarantee enhanced display. Choose types that accurately describe the page, keep business details consistent, and update markup when content changes. Use the Rich Results Test and Search Console enhancement reports where relevant, and avoid adding schema solely to chase a visual feature.
JavaScript rendering and resource health
For JavaScript-powered sites, verify that important content and links can be discovered and rendered reliably. Check server responses for direct route requests, resource loading, console errors, and whether content depends on user interaction to appear. Compare rendered output with raw HTML and use URL inspection for representative pages. Make sure robots rules do not block essential resources. When possible, expose critical links in crawlable HTML and use stable routes with meaningful status codes. Rendering issues should be diagnosed with evidence rather than assumed from the presence of JavaScript.
Local SEO and business information
For businesses serving defined areas, audit the accuracy and consistency of business name, address, phone, opening hours, service areas, and contact information across the website and relevant profiles. Ensure location pages offer genuine local value instead of repeating near-identical city pages. Check map links, contact forms, and structured data for consistency. Avoid inventing addresses or service areas and do not create misleading listings. Local visibility also depends on relevance, prominence, and user trust, so technical checks should be paired with accurate business information and useful service content.
Analytics, conversions, and measurement
Verify that analytics records important user actions and that conversion events are not duplicated or missing after site changes. Review organic landing pages, engagement, form completion, calls, bookings, and other meaningful outcomes where measurement is available. Separate tracking problems from actual performance changes. Check consent settings and privacy requirements before adding or changing tags. SEO success should not be reported only as impressions or average position; connect search visibility to the outcomes the business actually values, while recognising that attribution has limitations.
Backlinks and external signals
Review important external links to understand which pages attract references and whether those pages still exist or redirect appropriately. Investigate sudden link losses or legacy links to missing pages, but avoid assuming that every low-quality link requires disavowal. Prioritise the quality and relevance of references rather than raw volume. If a linked resource has moved, a relevant redirect may preserve a useful journey; if no replacement exists, consider whether restoring or rebuilding the resource is warranted. Never invent backlink data or promise rankings from link acquisition.
Competitor comparison without copying
Competitor research can reveal topics, page types, and user questions your site has not addressed. Compare the purpose and usefulness of pages, information architecture, service explanations, evidence, and conversion paths. Do not copy competitor text or imitate claims you cannot support. Search results differ by location, device, intent, and time, so one snapshot is not a complete picture. Use competitor observations to generate hypotheses, then validate them against your audience, capabilities, and first-party data.
Prioritise fixes by impact and effort
A practical audit scores issues by severity, affected URLs, business importance, confidence in the diagnosis, implementation effort, and risk. Fix widespread causes such as a broken template or wrong canonical rule before manually editing hundreds of pages. Address critical access failures and broken conversion journeys before cosmetic metadata issues. Some improvements require coordination across content, engineering, analytics, and business owners. Write down the expected result for each action so the team can verify whether the change actually solved the problem.
Create a remediation roadmap
Group recommendations into immediate fixes, short-term improvements, and longer-term projects. For each item, include examples, evidence, the likely root cause, affected templates, an owner, dependencies, a target date, and a verification method. Make clear which items are confirmed defects and which are opportunities requiring a business decision. Avoid overwhelming teams with a flat list of hundreds of issues. A smaller set of high-impact changes, implemented and validated well, is often more valuable than a large report that no one can act on.
Verify changes after deployment
After implementation, rerun the relevant tests and compare them with the baseline. Confirm the live response code, canonical, indexability, internal links, sitemap state, rendered content, and user flow. Use Search Console and logs to monitor how search engines respond over time; those systems may not update immediately. Keep a record of what changed and when. If the result is worse, investigate the difference and roll back when necessary. Verification is part of the fix, not an optional final step.
Common SEO audit mistakes
Avoid treating every tool warning as equally urgent, changing several major signals at once, blindly blocking URL patterns, writing repetitive keyword-stuffed metadata, and measuring success only by a single score. Do not recommend a redirect without checking whether the destination is relevant. Do not prune content solely because it has low traffic, and do not add structured data that does not match visible content. A good audit explains context, preserves working functionality, and distinguishes technical correctness from business value.
A repeatable 30-day audit plan
Week one: define goals, assemble URL sources, crawl the site, and identify critical access or indexation problems. Week two: investigate architecture, internal links, canonicals, sitemaps, and template-level issues. Week three: review content quality, search intent, metadata, structured data, mobile usability, and performance. Week four: prioritise fixes, assign owners, validate the first changes, and set up monitoring. Large websites may need a longer cycle, but a staged plan makes progress visible and prevents the audit from becoming an endless research exercise.
Frequently asked questions
How often should a site be audited? Run checks continuously for critical technical changes and schedule broader reviews based on site complexity and release frequency. Does every warning need fixing? No; confirm relevance and impact first. Can an audit guarantee higher rankings? No; it identifies risks and opportunities, but outcomes depend on competition, content, user needs, and many other factors. Should a small site use enterprise tools? Only if they solve a real need. What should be fixed first? Resolve access failures, broken key journeys, and systemic technical issues before lower-impact refinements.
Official resources and further reading
Use current primary documentation when validating findings: Google’s SEO Starter Guide, the robots.txt introduction, the sitemap overview, PageSpeed Insights, and Google Search Console. Tool interfaces and documentation change, so confirm the current guidance before applying a technical recommendation.
Related My Advisers technical SEO guides
For deeper implementation details, read the website migration SEO checklist, the advanced canonicalisation guide, the JavaScript SEO guide, and the log file analysis guide. Use them alongside a live crawl and the actual evidence available for your website.
How My Advisers can help
My Advisers provides technology consulting and digital growth support for startups, professionals, and small and medium-sized businesses. An SEO audit can combine crawling, indexation checks, content and keyword mapping, internal-link analysis, performance review, structured-data validation, and a prioritised remediation plan. The aim is to turn diagnostic findings into practical improvements that support real user journeys. Recommendations should be tailored to the website’s goals and validated after implementation rather than presented as a generic checklist.
Maintain an audit evidence log
Keep a central evidence log with the URL or template, finding, date, source, screenshot or response details where appropriate, severity, confidence, and next action. This prevents repeated investigations and makes it easier to explain why a recommendation exists. When an issue is fixed, record the deployment or change reference and the verification result. If a warning returns later, compare it with the earlier evidence to determine whether the same cause has returned or a different issue is producing a similar symptom.
Audit templates before editing individual pages
If dozens or hundreds of URLs share the same defect, inspect the template and content-management rules first. A missing canonical, repeated title pattern, oversized hero image, or broken breadcrumb can originate in one shared component. Fixing the template may resolve the issue consistently and prevent recurrence. Still sample representative pages after deployment, because shared changes can have exceptions for special landing pages, languages, or custom content types. Avoid making large-scale changes based only on one example URL.
Protect measurement and privacy
Audits often involve analytics exports, logs, form data, and third-party tags. Use only the data needed to answer the audit question, protect access, and follow retention and consent requirements. Avoid copying personal information into reports when aggregated evidence will suffice. Test tracking changes in a way that does not expose credentials or customer details. A technically successful SEO change should not introduce privacy risks or break consent behaviour.
Use a change calendar
Maintain a calendar of releases, migrations, theme updates, plugin changes, content launches, and tracking changes. Correlating a sudden rise in errors or a change in crawl patterns with a known release can speed up diagnosis. The calendar should include responsible owners and rollback notes for high-risk changes. It also helps distinguish a long-standing issue from a regression introduced recently. Pair it with alerts for critical status codes and conversion failures so that high-impact problems are not discovered only during the next scheduled audit.
Turn findings into ongoing governance
A one-time audit has limited value if the same issues return. Define standards for new pages, URL changes, redirects, metadata, image uploads, structured data, and release checks. Assign ownership so content editors know what they control and developers know which technical requirements must be preserved. Add lightweight automated tests for high-risk routes and periodically review the standards as the site evolves. Governance reduces the cost of future audits by preventing recurring mistakes at the source.