| name | technical-seo-checklist |
| description | Audits and implements the technical SEO basics any modern web app should ship with — metadata, structured data, crawlability, Core Web Vitals, and international targeting. Use when launching a new site, auditing an existing one, or reviewing a PR that touches routing, rendering, or head tags. |
Technical SEO Checklist
Practical technical SEO for modern web apps (Next.js, Angular, SvelteKit, plain HTML). Covers the pieces that actually move search ranking — not link-building, not content strategy.
When to use
- Launching a new site and making sure it's indexable.
- Auditing an existing site where organic traffic is below expectations.
- Reviewing a PR that touches
<head>, routing, or rendering.
- Migrating frameworks (e.g. React CSR → Next.js SSR) and wanting to preserve rankings.
- Rolling out internationalization.
Before you start
Know these:
- Rendering model — CSR, SSR, SSG, or ISR? Google indexes all of them now, but CSR still costs you. Anything commerce- or SEO-critical should be server-rendered.
- Target audience and language(s) — one language or many? Same content or localized?
- Existing ranking, if any. Migrations that break URLs lose rankings; you need a 301 map before launch, not after.
- What tooling you have — Google Search Console, Lighthouse, CrUX, Ahrefs/Semrush, PageSpeed Insights. At minimum, Search Console + Lighthouse + a crawler (Screaming Frog free tier).
- What "success" looks like — organic clicks, keyword rankings, indexed-page count, Core Web Vitals scores. Pick 2–3 metrics and track them monthly.
Audit workflow
- Crawl the site. Screaming Frog (or equivalent) — get a list of every URL, status code, title, description, canonical, indexability.
- Check indexability of every important page.
robots.txt, meta robots, canonical, blocked by auth, noindex accidentally set. Most "why isn't this ranking" bugs live here.
- Check the
<head> on representative pages — title, description, canonical, Open Graph, viewport, lang, charset.
- Check structured data. Rich Results Test for templates that should qualify (article, product, recipe, FAQ, breadcrumb).
- Check Core Web Vitals on real traffic via CrUX or Search Console → Core Web Vitals.
- Check mobile usability — viewport, tap targets, text legibility, responsive breakpoints.
- Check international setup if multi-language —
hreflang, language-specific URLs, consistent canonical strategy.
- Submit sitemap. Every discovered URL in
sitemap.xml; sitemap registered in Search Console.
Non-negotiable rules
- Every indexable page has a unique
<title> and <meta description>. Duplicates waste crawl budget and confuse ranking.
- Every page has a canonical URL. Self-referencing on originals, pointing to the canonical on duplicates. Don't set canonical to the homepage for every page — that's a common bug that tanks indexation.
- One
<h1> per page, describing the page content. Not the site name.
- Images have
alt text, width, and height. Alt helps accessibility and image search. Width/height prevents Cumulative Layout Shift (CLS).
- Mobile viewport meta tag is set:
<meta name="viewport" content="width=device-width, initial-scale=1">. Missing this fails mobile usability immediately.
- Don't rely on client-side rendering alone for content you want indexed. Googlebot renders JS, but slowly and inconsistently. SSR or SSG your important pages.
- Never redirect 302 for permanent moves. Use 301. 302s don't pass ranking.
- Never
noindex pages unintentionally. Check staging configs don't leak X-Robots-Tag: noindex into production. This is the #1 traffic-killer at launch.
References
- Metadata and head tags — title, description, canonical, Open Graph, Twitter cards, favicons, viewport.
- Structured data (Schema.org / JSON-LD) — which types actually show rich results, how to implement, how to validate.
- Crawlability —
robots.txt, sitemap.xml, meta robots, X-Robots-Tag, pagination, faceted navigation traps.
- Core Web Vitals — LCP, INP, CLS — what they are, how to measure, the specific patterns that fix each one.
- International SEO —
hreflang, URL patterns for i18n, language detection vs explicit selection.
