SEO Testing (Technical SEO QA)
Auditing technical SEO and search visibility — crawlability, indexation, robots.txt, sitemaps, canonical/hreflang, meta and Open Graph tags, structured data (JSON-LD), Core Web Vitals, JS-rendering, redirects, and automating SEO regression checks in CI.
Ship-blocking SEO bugs are cheap to catch pre-release (accidental noindex, broken canonicals, blocked JS) and brutal in production. Automate them.
Crawl & index basics
| Check | Pass condition | How | |
|---|---|---|---|
robots.txt | 200, doesn't block CSS/JS/important paths; lists sitemap; staging blocked by auth, not just robots | curl -s site/robots.txt | |
| XML sitemap | Valid, ≤50k URLs/50 MB, only 200-status canonical indexable URLs, lastmod accurate | xmllint --noout sitemap.xml | |
| Status codes | Indexable pages 200; moved = single-hop 301; removed = 410/404; no soft-404s; no redirect chains/loops | curl -sI -L -o /dev/null -w "%{http_code} %{num_redirects}\n" URL | |
| Meta robots | No stray noindex/nofollow (also X-Robots-Tag header) | `curl -sI URL \ | grep -i robots` |
| Canonical | Self-referencing absolute URL on indexable pages; consistent with sitemap & hreflang | view-source / crawler | |
| Rendering | Critical content & links present in rendered DOM (not only after user action); SSR/SSG/prerender for JS apps | Search Console URL Inspection, Playwright page.content() |
On-page
Unique <title> (≈50–60 chars) & meta description (≈120–155) · single <h1> · logical heading order · descriptive anchor text · image alt + width/height + modern formats + lazy-load below fold · internal links ≤3 clicks deep · pagination via crawlable links · faceted URLs controlled (canonical/noindex/robots).
International
hreflang reciprocal (each page lists all alternates + self), valid ISO codes, x-default; one language per URL; no auto-redirect by IP that blocks crawlers.
Social & rich results
Open Graph (og:title/description/image 1200×630/url/type), Twitter card, JSON-LD structured data valid: Organization, WebSite + SearchAction, BreadcrumbList, Article, Product, FAQPage, SoftwareSourceCode. Validate with Google Rich Results Test / Schema.org validator; keep markup matching visible content.
Performance & UX signals
Core Web Vitals (LCP ≤2.5s, INP ≤200ms, CLS ≤0.1) via Lighthouse/CrUX (web-performance-testing) · HTTPS + HSTS · mobile-friendly, no intrusive interstitials · accessibility overlaps (accessibility-testing).
Automation
npx lighthouse https://example.com --only-categories=seo,performance --output=json --output-path=lh.json --chrome-flags="--headless"
npx unlighthouse --site https://example.com # whole-site crawl report
screamingfrogseospider --crawl https://example.com --headless --export-tabs "Internal:All" # if licensed
# CI guard: no noindex on prod, canonical present
import requests, re
h = requests.get(URL).text
assert 'noindex' not in re.findall(r'<meta[^>]+robots[^>]*>', h, re.I).__str__()
assert re.search(r'<link[^>]+rel="canonical"', h)
Playwright/pytest checks per template: title length, one h1, canonical, JSON-LD parses (JSON.parse), OG image 200. Diff sitemap URL count release-over-release.
Release checklist
Staging noindex removed at launch · redirects map tested (old→new, 100% of top URLs) · analytics still firing · robots/sitemap updated · Search Console property verified & sitemap submitted · monitor coverage & CWV for 4 weeks.
Related
web-performance-testing, accessibility-testing, analytics-tracking-testing, cross-browser-testing, localization-testing