SEO Audit Checklist 2026: The Complete 50-Point Framework

An SEO audit is a systematic evaluation of your website's search performance — covering technical infrastructure, on-page signals, content quality, mobile experience, structured data, and AI search readiness. This 50-point checklist gives you the complete framework, organized by category and priority, so nothing critical is missed.

Google's ranking system has evolved significantly. Core Web Vitals are now well-established ranking signals. AI Overviews appear in roughly 30% of search results. Answer engines like ChatGPT, Perplexity, and Claude are pulling content from the open web — and they cite structured, authoritative sources far more often than unstructured ones. A 2026 SEO audit must cover more ground than it did two years ago.

The average website scores just 62–65% on traditional technical and on-page SEO, and a dismal 28–35% on AI search readiness, according to data from seoscore.tools. That gap is the opportunity: websites scoring above 70% in AEO and GEO readiness receive roughly 4× more AI citations than those below 40%. Most of your competitors are below 35%.

Use this checklist as your complete audit framework. Work through it section by section, fix issues in priority order, and re-audit every three to six months. Pair it with a crawl tool like AuditDepot's free scanner to automate the heavy lifting, or run it manually alongside Google Search Console and PageSpeed Insights.

Quick Reference: The 7 Audit Categories

1. Crawlability & Indexation (8 Checks)

If search engines can't find or access your pages, nothing else matters. These checks verify that Google can crawl, render, and index your site correctly — and that you're not wasting crawl budget on dead ends.

• robots.txt is not blocking important pages. Fetch yourdomain.com/robots.txt directly. Confirm that key sections (homepage, blog, product pages) aren't disallowed. A single Disallow: / line de-indexes your entire site. Also ensure CSS and JS files are not blocked — Google needs them for rendering.
• XML sitemap is submitted and error-free. Check Google Search Console → Sitemaps. Confirm the sitemap is discoverable, all listed URLs return 200, and lastmod dates are accurate. Remove any URLs that redirect or return 4xx/5xx.
• Index coverage shows no unexpected exclusions. In GSC → Pages, review the "Not indexed" tab. Look for pages excluded by noindex tags you didn't set, soft 404s, and "Crawled — currently not indexed." Each signals a problem worth investigating.
• No important pages are orphaned. An orphan page has zero internal links pointing to it — Google may never discover it. Crawl your site with Screaming Frog and filter for pages with zero incoming internal links. Every page meant to rank needs at least one internal link from another indexable page.
• Canonical tags are self-referencing and consistent. Every page should have a <link rel="canonical"> pointing to itself — or to the canonical version if multiple URLs serve the same content. Pages without canonicals risk duplicate content issues; conflicting canonicals across URL variations confuse Google.
• No broken internal links (4xx) or redirect chains. Crawl the site and flag every internal link returning a 404 or 410. Also check for redirect chains — more than one hop between the clicked URL and the final destination wastes crawl budget and slows page load.
• Important pages are within 3 clicks of the homepage. If a key product or blog post requires 5+ clicks from the homepage, Google treats it as low-priority. Flatten your site architecture so high-value pages are reachable within 3 clicks.
• HTTPS is enforced everywhere. Every HTTP URL should 301-redirect to its HTTPS equivalent. Check for mixed content warnings (HTTP resources loaded on HTTPS pages) using Chrome DevTools or a crawler.

2. On-Page SEO (8 Checks)

On-page signals tell Google what each page is about and why it should rank. These checks ensure every page has a clear, unique topic signal — from title tags down to alt text.

• Every page has a unique, keyword-optimized title tag (50–60 characters). The title tag is the single most important on-page ranking signal. Each page needs a unique title, with the primary keyword near the start. No duplicates across pages. No truncation in search results.
• Every page has a compelling meta description (140–160 characters). While meta descriptions aren't a direct ranking factor, they heavily influence click-through rate — which is. Include the primary keyword naturally, state a clear value proposition, and end with a soft call to action. No duplicate descriptions across pages.
• Single H1 per page containing the primary keyword. Every indexable page needs exactly one H1 tag, and it should include the page's primary keyword in a natural, readable way. Missing H1s and multiple H1s are both common — and both easy to fix.
• Heading hierarchy is logical (H1 → H2 → H3, no skips). Headings should form a nested outline: H1 as the page title, H2s for major sections, H3s for sub-sections within H2s. Never jump from H1 to H3 without an H2 in between. A logical hierarchy helps both Google and AI answer engines understand your content structure.
• Image alt text is descriptive and includes keywords where natural. Every content image needs alt text. For decorative images, use alt="" (empty, not missing). For meaningful images, describe what's shown and include a relevant keyword if it fits naturally — not forced.
• URL slugs are short, keyword-rich, and hyphenated. URLs should use lowercase, hyphens between words, and the primary keyword. Avoid dates, IDs, session parameters, and unnecessary folder depth. A good slug: /blog/seo-audit-checklist/. A bad one: /blog/2026/07/08/cat=seo&id=4429.
• Open Graph and Twitter Card tags are present on every page. og:title, og:description, og:image (1200×630px), and og:url should be set on every page. Twitter Cards should use summary_large_image. These control how your content appears when shared on social platforms — missing tags mean ugly, auto-generated previews.
• Breadcrumb navigation is on every page below the homepage. Breadcrumbs improve UX for users and provide Google with additional context about your site structure. They also appear in search results snippets, increasing click-through rate.

For a deeper walkthrough of on-page signals with specific optimization guidance, see the dedicated on-page SEO checklist for 2026.

3. Content Quality & E-E-A-T (8 Checks)

Google's Quality Rater Guidelines emphasize Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T). Content that fails these checks won't rank — regardless of technical perfection — and AI answer engines are even more selective about which sources they cite.

• Every article has a visible publish date and author byline. Google uses dates to assess freshness. AI engines use author attribution to evaluate expertise. No date and no author = low trust signal for both.
• Content demonstrates first-hand experience or original insight. Does the article include case studies, original examples, data, or frameworks that go beyond what competitors cover? Content that rehashes the same generic advice as the top 10 results won't outrank them. Every article needs at least one piece of original value — a unique framework, real data, or specific actionable steps competitors missed.
• External claims are cited with links to authoritative sources. If your article says "90% of web pages get zero organic traffic," link to the study. Unsourced statistical claims erode trust with both Google's algorithms and savvy readers. Aim for 2–3 external citations per article to high-authority domains.
• No thin content pages (fewer than 300 words of substantive body text). Pages with minimal content signal low value to Google. For informational articles, target 1,200–2,000 words of substantive, well-structured content. For product and category pages, ensure unique descriptions that go beyond manufacturer specs.
• No duplicate or near-duplicate content across pages. Two pages targeting the same keyword with substantially similar content cannibalize each other — splitting ranking potential and confusing Google about which page to serve. Use Screaming Frog to find near-duplicate title tags and content patterns.
• Content freshness: key pages updated within the last 12 months. Google favors fresh content, especially for topics where information changes rapidly (SEO, technology, finance, health). Articles older than 12–18 months should be reviewed for outdated statistics, deprecated tools, and changed best practices. Add a "Last updated" date when you refresh.
• Internal links are contextual and use descriptive anchor text. "Click here" and "learn more" tell Google nothing about the linked page. Anchor text should describe what the target page is about — ideally including its target keyword. Aim for 3–5 contextual internal links per article pointing to related content.
• Readability: paragraphs are short, sentences are scannable, key points use formatting. Walls of text lose readers. Use short paragraphs (2–4 sentences), bullet points for lists, bold text for key takeaways, and tables for comparisons. Readability signals dwell time — and dwell time correlates with rankings.

4. Mobile & Performance (8 Checks)

Google has used mobile-first indexing for all sites since late 2023. Core Web Vitals are confirmed ranking signals. If your site is slow or broken on mobile, you're losing rankings and users simultaneously.

• Core Web Vitals pass for both mobile and desktop. LCP (Largest Contentful Paint) under 2.5s, INP (Interaction to Next Paint) under 200ms, CLS (Cumulative Layout Shift) under 0.1. Check with Google PageSpeed Insights using field data (CrUX) rather than lab data alone. Lab data simulates; field data reflects real users.
• Site passes mobile-friendly test. Open Chrome DevTools → Lighthouse → Mobile audit. Confirm no horizontal scroll, touch targets are at least 48×48px, font sizes are readable without zooming, and no intrusive interstitials block content on mobile.
• Page load time under 3 seconds on 4G mobile. Every additional second of load time costs conversions. Compress images to WebP format, enable Brotli or Gzip compression, minify CSS and JavaScript, and use a CDN for static assets.
• Images are properly sized and lazy-loaded. Don't serve a 4000px-wide image in a 780px container. Use responsive srcset attributes or serve appropriately sized images. Add loading="lazy" to images below the fold to defer loading until they're about to enter the viewport.
• No render-blocking resources. CSS and JavaScript that block the initial page render delay the moment a user can interact with your content. Use Lighthouse to identify render-blocking resources, then defer or async-load non-critical scripts and inline critical CSS.
• Server response time (TTFB) under 800ms. Time to First Byte measures server responsiveness. Slow TTFB often indicates underpowered hosting, no caching layer, or slow database queries. Use a CDN with edge caching to reduce TTFB for global audiences.
• No layout shifts from dynamically loaded content. If an ad banner, image, or embedded widget loads after the page renders and pushes content down, users tap the wrong thing or lose their place. Reserve space with fixed dimensions for all dynamic elements to prevent CLS.
• Fonts are optimized for performance. Use font-display: swap so text renders immediately in a fallback font while the custom font loads. Preload critical font files. Subset fonts to include only the characters your site uses.

5. Structured Data & Schema (6 Checks)

Structured data helps Google understand your content and unlocks rich results — review stars, breadcrumbs, article carousels, and AI Overview citations. Missing or broken schema is a missed opportunity, and some schema types are restricted or deprecated.

• Article or BlogPosting schema on every blog post. Every article needs JSON-LD Article schema with: headline, datePublished, author (Person type with name), publisher (Organization type with name and logo URL). Use absolute URLs throughout.
• Organization schema on the homepage. Include name, url, logo, and sameAs links to social profiles. This helps Google build a knowledge graph entry for your brand.
• BreadcrumbList schema on all pages below the homepage. Breadcrumb schema powers the breadcrumb trail in search result snippets, improving click-through rate and providing Google with structural context.
• No restricted schema types: FAQPage is for government/healthcare only. Since August 2023, Google restricts FAQ rich results to well-known, authoritative government and health websites. If your site uses FAQ schema and you're not in those categories, remove it — it will not appear in search results and marks your structured data as non-compliant.
• No deprecated schema types: HowTo was retired September 2023. HowTo rich results no longer appear in Google Search. Any HowTo schema on your site is dead weight — remove it.
• All schema validates without errors. Test every schema type on your most important pages using the Google Rich Results Test and the Schema.org Validator. Warnings are acceptable; errors are not.

6. AI Search Readiness — AEO & GEO (7 Checks)

AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization) are the newest layers of an SEO audit. AI answer engines like ChatGPT, Perplexity, and Google's AI Overviews pull content from the open web — and they cite sources that are structured, authoritative, and quotable. Most websites score below 35% here, creating a significant opportunity for sites that optimize early.

• Content contains clear, quotable statements. AI engines cite specific sentences or paragraphs — not entire pages. Every article should include 3–5 sentences formatted as standalone, factual claims that an AI can extract and attribute. Example: "Google uses Core Web Vitals — LCP, INP, and CLS — as confirmed ranking signals in its page experience update."
• Semantic HTML is used correctly. Use <article> for the main content, <section> for major divisions, <header> for title areas, and <nav> for navigation. Semantic HTML helps AI parsers understand content structure — and it's a prerequisite for reliable AI citation.
• Entity-rich content that references recognized concepts. Use specific, unambiguous names for tools, frameworks, companies, and standards. "Google Search Console" is an entity Google recognizes; "the console" is not. When introducing a concept, define it in clear, structured language that an AI can extract as a standalone definition.
• Content covers question-answer patterns explicitly. Format key sections around specific questions your audience asks, then answer them directly in the first sentence of the section. AI engines prioritize content that mirrors the question-answer format their users expect.
• Page is crawlable by AI user agents. Check your robots.txt and server logs. Some sites block GPTBot, Claude-Web, or PerplexityBot — intentionally or accidentally. Decide whether you want AI visibility, and configure accordingly. If you're publishing content to be found, allow AI crawlers.
• Data and statistics are presented in structured formats. Tables, bulleted lists, and clearly labeled data points are easier for AI engines to extract than narrative text. If your article includes a comparison, use a table. If it includes a list of steps, use an ordered list.
• Author expertise signals are clear and linked. AI engines evaluate source credibility by checking author pages, bios, and external profiles. Each article's author should link to a detailed author page that includes credentials, experience, and links to verified profiles (LinkedIn, Google Scholar, professional site).

7. Analytics & Monitoring (5 Checks)

An audit is only as good as the data that validates it. These checks ensure you can measure the impact of your fixes and catch new issues early.

• Google Search Console is connected and verified. Without GSC, you're blind to how Google sees your site. Verify both the domain property and all URL-prefix variants. Check that data is flowing — the Performance report should show impressions and clicks for recent dates.
• Analytics tracking fires correctly on every page. Use Google Tag Assistant or GA4's real-time report to confirm pageviews are registering. Check that key events (form submissions, CTA clicks, purchases) are tracked and reporting correctly. Missing tracking means you can't measure the impact of your SEO fixes.
• Automated monitoring is set up for critical pages. Use a tool or script to check that your top 10–20 pages return 200 status codes daily. Pages going offline silently is a common and painful problem — set up monitoring so you're the first to know, not your users.
• Search performance is reviewed at least monthly. In GSC, review clicks, impressions, average position, and CTR for your top queries and pages. Look for: (1) queries at positions 4–15 with >10 impressions — these are your quick wins, (2) pages with declining clicks despite stable impressions — these need content refresh, (3) new queries appearing — these signal topics to double down on.
• Audit schedule is defined and recurring. A full audit every 3–6 months catches issues before they compound. Monthly spot-checks on crawl status, Core Web Vitals, and index coverage catch urgent problems between full audits. Calendar it — uncalendared audits become unperformed audits.

How to Run This Audit Efficiently

Running every check manually across a site with hundreds of pages is impractical. Here's a streamlined approach:

1. Automated crawl first. Run a full-site crawl with AuditDepot's free SEO scanner or Screaming Frog. This surfaces technical issues (broken links, missing canonicals, redirect chains, missing H1s) across every page in minutes — no manual checking required. For a comprehensive automated scan covering 40+ checks, see our comparison of the best free SEO audit tools in 2026.
2. Manual spot-check on high-value pages. For your top 5–10 pages by traffic, run the full 50-point checklist manually. Automated tools catch patterns; manual review catches nuance — especially on content quality and E-E-A-T checks.
3. Prioritize by impact. Fix high-priority items first (marked HIGH in this checklist), then medium. A single robots.txt fix that unblocks 200 pages is worth more than fixing alt text on 200 individual images.
4. Re-scan after fixes. Run the automated crawl again after addressing issues to confirm fixes took effect and no new problems were introduced during remediation.

For the technical infrastructure layer specifically — crawlability, indexation, Core Web Vitals, and site architecture — the technical SEO audit checklist provides a deeper 47-point framework with per-check tool recommendations. For on-page signals in more detail, see the on-page SEO checklist. And if you're running audits on content-heavy sites like WordPress, the WordPress-specific SEO audit checklist covers platform-level configuration checks that general audits miss.