SEO Technical
When to Use
- Running a technical SEO audit on an existing site.
- Diagnosing why pages are not being indexed or ranked.
- Optimizing Core Web Vitals or page speed.
- Setting up site architecture (sitemaps, robots.txt, canonicalization).
- Implementing structured data / schema markup.
- Managing AI crawler access via robots.txt.
- Preparing a new site or migration for SEO readiness.
- Reviewing JavaScript rendering and its impact on indexation.
Context Required
- From startup-context: product description, target audience, primary keywords, current domain and site structure.
- From the user: site URL, access to Google Search Console data (if available), specific pages of concern, known issues, tech stack (static site, SPA, WordPress, etc.), whether the site has been previously audited.
Workflow
Audit across 9 technical dimensions. Score on a 100-point scale with category breakdowns.
- Crawlability audit — Foundation; nothing else matters if search engines cannot crawl the site.
- Check robots.txt for unintended disallow rules. - Verify XML sitemap exists, is submitted to Search Console, and is up to date. - Identify crawl errors (4xx, 5xx status codes). - Check for orphan pages (no internal links pointing to them). - Audit redirect chains — flatten to single hops. Chains of 3+ waste crawl budget and dilute link equity. - For sites with 10,000+ pages, analyze crawl budget efficiency and remove low-value pages from the index.
- Indexation audit — Ensure crawled pages are actually indexed.
- Check site:domain.com results vs. expected page count. - Look for noindex tags applied unintentionally. - Identify duplicate content issues and missing canonical tags. - Check for thin content pages that may be filtered out. - Review index coverage report in Search Console for excluded pages.
- Security and HTTPS — HTTPS is a ranking signal and a trust requirement.
- Verify HTTPS everywhere with proper redirects from HTTP. - Check for mixed content issues. - Validate SSL certificate and HSTS headers.
- URL structure — Short, descriptive, hyphenated URLs without parameters.
- Consistent URL patterns across the site. - No duplicate content from URL variations (trailing slashes, www vs. non-www). - Self-referencing canonical tags on every page.
- Mobile optimization — As of July 2024, Google crawls and indexes ALL websites exclusively with the mobile Googlebot user-agent. Mobile-first is not optional.
- Mobile and desktop content parity is mandatory. - Test with mobile-friendly tools. - Responsive design, no horizontal scroll, touch targets properly sized.
- Core Web Vitals — Google uses these as ranking signals. Current targets:
- LCP (Largest Contentful Paint): Target under 2.5 seconds. Fixes: optimize hero images, implement lazy loading, use CDN, reduce server response time. - INP (Interaction to Next Paint): Target under 200ms. INP replaced FID in March 2024. Fixes: reduce JavaScript execution time, break up long tasks, optimize event handlers. - CLS (Cumulative Layout Shift): Target under 0.1. Fixes: set explicit dimensions on images/embeds, avoid injecting content above existing content, use font-display: swap. - Check for render-blocking CSS/JS, unoptimized images, excessive third-party scripts.
- Structured data — Implement schema markup for rich results.
- Organization schema on the homepage. - Article/BlogPosting schema on content pages. - Product schema on product pages. - FAQ schema where applicable (boosts SERP real estate). - BreadcrumbList schema for navigation clarity. - Validate with Google's Rich Results Test.
- JavaScript rendering — Critical for SPAs and JS-heavy sites.
- Serve canonical tags, robots directives, and structured data in initial server-rendered HTML. Google may not reliably process these if injected via JavaScript. - Verify server-side rendering or pre-rendering is in place for critical content. - Test what Googlebot actually sees vs. what users see.
- AI crawler management — Managing AI crawlers via robots.txt is a critical 2025-2026 consideration.
- GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot, Bytespider (ByteDance), Google-Extended - Blocking Google-Extended does NOT affect Google Search indexing or AI Overviews — only Gemini training data. - Decide per-crawler based on business needs: allow, block, or rate-limit.
Output Format
A prioritized audit report scored on a 100-point scale:
- Critical issues (blocking indexation or causing major ranking loss): fix immediately
- High-impact improvements (will meaningfully improve rankings): fix within 2 weeks
- Medium improvements (incremental gains): fix within 30 days
- Low priority (minor optimizations): address when convenient
Each item includes: the issue, why it matters for rankings, how to fix it, and estimated effort (low/medium/high).
Frameworks & Best Practices
- Canonical tags are directives, not suggestions: While Google treats them as hints, properly implemented canonicals are followed in the vast majority of cases. Always self-canonical pages and canonical duplicates to the preferred version.
- JavaScript SEO: Critical metadata (canonical, robots, structured data) must be in initial server-rendered HTML. Google can render JS but with delays and reliability gaps.
- Mobile-first is complete: Since July 2024, there is no "desktop index." The mobile version IS the version Google indexes. Full content parity is non-negotiable.
- INP replaced FID: FID was fully removed from Chrome tools by September 2024. All performance optimization should target INP (under 200ms), not FID.
- Crawl budget: Rarely an issue for sites under 10,000 pages. For larger sites, prioritize by removing low-value pages and fixing redirect chains.
- Hreflang for international sites: Implement correctly or not at all. Errors cause worse indexation than no hreflang tags.
- Log file analysis: Server logs reveal what Googlebot actually crawls vs. what you want crawled. This is the ground truth of crawlability.
- Speed is a tiebreaker: Rarely causes dramatic ranking changes alone, but affects user behavior metrics (bounce rate, time on site) which indirectly impact rankings.
- E-E-A-T signals matter: Author bios with credentials, about page, contact info, links to authoritative sources, HTTPS, privacy policy, terms of service.
Related Skills
content-strategy— when the audit reveals content gaps or thin content that needs a broader content planlanding-page— when key landing pages need both CRO and on-page SEO optimization
Examples
Example 1: Full audit
"We launched our site 6 months ago and we're barely showing up in Google. Can you audit our SEO?"
Good output: A structured audit across all 9 dimensions with a 100-point score. Identifies that robots.txt is blocking the /blog/ directory, sitemap is missing, several pages have duplicate title tags, Core Web Vitals are failing on mobile (LCP at 4.2s due to unoptimized hero image), canonical tags and structured data are injected via JavaScript instead of server-rendered HTML, and no AI crawler policy is defined. Prioritized fix list with critical, high, medium, and low items with effort estimates.
Example 2: Specific issue
"Our blog posts were ranking well but traffic dropped 40% last month."
Good output: Diagnostic checklist starting with Search Console data review (manual actions, crawl errors, index coverage changes), then checking for algorithm update timing, content cannibalization between similar posts, accidental noindex or broken canonicals, Core Web Vitals regression (especially INP after a JS-heavy deploy), lost backlinks, and mobile rendering issues. Recommends specific data to gather and steps to isolate the cause.