A real technical audit is not a Screaming Frog export with a cover page. Here are the 120 checks we run on every site, grouped into 8 categories, with priority and why each matters.
Most “technical audits“ we inherit from previous agencies are glorified Lighthouse screenshots. A real audit checks 120+ things across 8 categories, classifies them by business impact, and produces developer-ready tickets with acceptance criteria. Here is the exhaustive checklist we run.
1. Crawlability & indexation (22 checks)
- robots.txt syntax and sensibility
- XML sitemap presence, freshness and format
- Canonical tags on every URL
- Duplicate content via URL parameters
- HTTP vs HTTPS canonicalisation
- Trailing-slash consistency
- www vs non-www canonicalisation
- noindex tags on paginated pages
- Soft-404 detection
- Orphan pages
- Crawl budget waste (faceted filters, session IDs, search result pages)
- X-Robots-Tag header consistency
- Hreflang implementation if multi-regional
- Pagination: rel=next/prev legacy vs modern patterns
- JavaScript-rendered content crawlability
- Mobile vs desktop parity
- IP-based cloaking detection
- Server response codes (200 vs 301 vs 302)
- Redirect chains over 2 hops
- Redirect loops
- Broken internal links (4xx/5xx)
- Google Search Console coverage issues
2. Core Web Vitals (14 checks)
- LCP: largest contentful paint under 2.5s on field data
- INP: interaction to next paint under 200ms
- CLS: cumulative layout shift under 0.1
- FCP: first contentful paint
- TTFB: time to first byte
- Hero image preloading strategy
- Font loading strategy (swap, optional, block)
- Third-party script audit
- Render-blocking JavaScript
- Render-blocking CSS
- Main-thread blocking tasks
- Image lazy loading implementation
- Above-the-fold image format (AVIF, WebP)
- Responsive image srcset usage
3. Site architecture (16 checks)
- URL structure and readability
- Click depth from homepage
- Internal linking silo models
- Breadcrumb navigation + structured data
- Footer linking hygiene
- Anchor text distribution internal
- Orphan pages in internal graph
- Hub and spoke architecture
- Parent category structure
- Filter and facet URL rules
- Pagination architecture
- Tag page handling
- Author page canonicalisation
- Search result page noindex
- Faceted navigation canonical rules
- Mobile menu structure
4. Schema markup (12 checks)
- Organization schema on homepage
- WebSite schema with sitelinks search
- BreadcrumbList on every page
- Article schema on blog posts
- Product schema on e-commerce
- LocalBusiness schema if applicable
- FAQPage schema where relevant
- Review / AggregateRating schema
- Event, Recipe, HowTo as appropriate
- Valid JSON-LD syntax
- Rich Results Test validation
- Schema consistency across pages
5. Mobile & responsive (10 checks)
- Mobile-first indexing parity
- Viewport meta tag
- Tap target sizing
- Font size readability
- Horizontal scrolling detection
- Mobile menu usability
- Responsive image delivery
- Touch interaction delays
- Accelerated Mobile Pages (legacy audit)
- PWA manifest and service worker if applicable
6. Security (12 checks)
- HTTPS enforcement
- HSTS header
- Mixed content detection
- Content Security Policy
- X-Frame-Options
- X-Content-Type-Options
- Referrer-Policy
- Permissions-Policy
- TLS version (1.2+)
- Certificate validity and chain
- Vulnerable JavaScript library versions
- Directory listing exposure
7. International & accessibility (12 checks)
- Hreflang tag reciprocity
- Language-region code validity
- x-default implementation
- Language detection by URL, not IP
- Alt text on all images
- Semantic HTML5 landmarks
- Heading hierarchy (one H1 per page)
- Form label association
- Color contrast WCAG AA
- Keyboard navigation
- ARIA attributes correctness
- Screen reader testing
8. Analytics & tracking (12 checks)
- Google Analytics 4 implementation
- Google Search Console verification
- Bing Webmaster Tools verification
- Event tracking completeness
- Conversion goal setup
- Tag Manager container review
- Consent Mode v2 compliance
- Server-side tracking if applicable
- Sample rate and retention settings
- Cross-domain tracking
- Referral exclusions
- Bot and spam filtering
Prioritisation
Each finding is classified in three dimensions: severity (critical / high / medium / low), effort (S/M/L), and business impact (revenue / rankings / risk). The deliverable is not the audit — it is the prioritised ticket queue. A 120-check audit without prioritisation is a 400-page document nobody reads.
The output for each issue includes: what is wrong, why it matters, the fix (with code snippet if applicable), acceptance criteria, and expected impact. That is how you turn an audit into shipped improvements.