What is a technical SEO checklist?

It is a list of checks on how your site works for search engines: crawling, indexing, rendering, redirects, canonicals, speed, mobile, security and structured data. It is different from on-page SEO (words, titles, headings) and off-page SEO (links from other sites).

How often should I run a technical SEO audit?

Run the High priority checks after every major release (new theme, plugin, redesign or CMS move) and a full audit every 3 to 6 months. Watch the Page indexing and Core Web Vitals reports in Search Console monthly.

What are the Core Web Vitals thresholds in 2026?

Largest Contentful Paint 2.5 seconds or less, Interaction to Next Paint 200 milliseconds or less, and Cumulative Layout Shift 0.1 or less, measured at the 75th percentile of real page loads on mobile and desktop (web.dev).

Do I need special technical SEO for AI Overviews?

No. Google says there are no extra requirements or special markup for AI Overviews or AI Mode: the page must be indexed and eligible to show a snippet (Google). For ChatGPT, Claude and Perplexity, allow their search crawlers and serve your content as HTML.

Should I block GPTBot or Google-Extended?

That is a business choice about model training, not about search visibility. Blocking Google-Extended does not affect Google Search, according to Google. Blocking GPTBot does not block OAI-SearchBot, which is the crawler for ChatGPT search.

Is crawl budget something I need to worry about?

Usually not. Google's crawl budget guide is for sites with over 1 million pages that change weekly, or over 10,000 pages that change daily. Smaller sites should fix errors, duplicates and internal links instead.

Technical SEO Checklist 2026: 99 Checks (Free Sheet)

The 14 areas of a technical SEO audit

#	Area	The question it answers	Main tool
1	Crawlability	Can Googlebot reach the page?	robots.txt report, Crawl stats
2	Indexability	Is Google allowed to index it?	URL Inspection, view source, curl -I
3	Redirects and status codes	Does each URL return the right code in one hop?	A crawler (Screaming Frog, Sitebulb)
4	JavaScript rendering	Is the content there after rendering, and before it?	URL Inspection live test, curl
5	Site architecture	Can Google find every page through links?	Crawl depth and orphan reports
6	Canonicals	Does every signal name the same preferred URL?	URL Inspection (user vs Google canonical)
7	Hreflang	Do language versions point to each other correctly?	Crawler hreflang report
8	Structured data	Is the markup valid and true to the page?	Rich Results Test
9	Core Web Vitals	Is the page fast and stable for real users?	Core Web Vitals report, PageSpeed Insights
10	Mobile	Does the mobile version have everything?	Smartphone URL Inspection, Lighthouse
11	HTTPS and security	Is every page secure and clean?	HTTPS report, Security & Manual Actions
12	Server logs	What did crawlers really request?	Access logs, log analyzer
13	AI crawlers	Can AI search tools read and cite you?	robots.txt, WAF bot settings
14	Monitoring	Will you notice when a release breaks something?	Scheduled crawls, alerts

Start here: the 10 checks that break sites most often

This is a real excerpt from the spreadsheet. If you only have one hour, do these.

Check	How to check	How to fix
robots.txt returns 200 (or 404), never 5xx	Open /robots.txt. Search Console > Settings > robots.txt report	Fix the server rule. On a 5xx, Google stops crawling for 12 hours
robots.txt does not block pages, CSS or JS you need	Read every Disallow line. URL Inspection > Page resources	Remove broad rules. Block only cart, internal search, admin
No accidental noindex on pages that should rank	View source for "noindex". URL Inspection	Remove the tag or the CMS setting that adds it
No noindex in the X-Robots-Tag header	`curl -I https://yoursite.com/page`	Remove the server or CDN rule
Only one host version (https, www or non-www)	Open all 4 versions	301 each to the chosen one in one hop
Removed pages return 404 or 410, not 200	Open a made-up URL, check status	Make the error template send a real 404
Canonicals point to live 200 URLs	Crawler canonical report	Point to the final live URL
Firewall and CDN let Googlebot through	Crawl stats > Host status. WAF logs	Allowlist verified crawlers (reverse DNS)
Main content is in the rendered HTML	URL Inspection > View tested page	Server-side render key templates
Sitemap lists only canonical 200 URLs	Crawl the sitemap URLs	Regenerate it from your CMS

What changed in technical SEO from 2024 to 2026

Many top-ranking checklists still repeat old advice. Here is what is true in October 2026.

Change	Date	What to do	Source
INP replaced FID as a Core Web Vital	March 12, 2024	Measure INP (good is 200 ms or less). Ignore FID advice	web.dev
Mobile Usability report and Mobile-Friendly Test retired	December 1, 2023	Use Lighthouse and real phones	Google
Breadcrumb rich results shown on desktop only	January 2025	Keep breadcrumbs for users and links. Do not expect them on mobile results	Google updates
Googlebot fetches the first 2 MB of HTML for Search	Documented February 2026	Keep key content early. Move large inline data out of the HTML	Googlebot docs
FAQ rich results stopped showing	May 7, 2026	Stop adding FAQ markup for rich results. Keep useful FAQs on the page	Google updates
llms.txt not needed for Google Search	June 2026	Optional. No ranking effect in Google	Google updates
No special markup needed for AI Overviews and AI Mode	Current	A page must be indexed and eligible for a snippet. That's it	Google AI features

How to run a technical SEO audit in about 2 hours

Open Search Console (a Domain property). Check Settings > Crawling > Crawl stats and the robots.txt report. Note any 5xx, host problems or a crawl drop.
Read your robots.txt. Every Disallow line should block only junk. Check the AI crawler lines too.
Run a crawl of up to a few thousand URLs with any crawler. Export status codes, redirects, canonicals, indexability, depth and orphan pages.
Inspect your 10 most important URLs with URL Inspection. Compare user-declared and Google-selected canonical. Look at the rendered HTML and screenshot.
Open the Page indexing report. Only care about reasons that hit pages you want in Google. Our Page indexing fix guide covers each reason.
Open the Core Web Vitals report. Fix the URL groups marked Poor first, mobile before desktop.
Test one URL per template (home, category, product or service, post) in the Rich Results Test and PageSpeed Insights.
Mark each row in the spreadsheet as Done, To do or Not needed. The Progress sheet shows what is left.

Worked example (fictional site). northwind-outdoors.com moved to a new theme. Clicks fell 30% in two weeks. The audit found two problems in 20 minutes: the new theme added a second canonical tag that pointed every product to the homepage, and the CDN's new bot rule was challenging Googlebot (Crawl stats showed a jump in 5xx). Removing the duplicate canonical and allowlisting verified Googlebot IPs fixed both. Neither would have shown up in an "on-page" checklist.

Technical SEO for AI search

AI answers are built from pages that crawlers can fetch. Two things matter.

1. Let the right crawlers in. Search crawlers decide if you can be cited. Training crawlers are a separate choice.

User agent	Company	Purpose	Our default
Googlebot	Google	Search, AI Overviews, AI Mode	Allow
Google-Extended	Google	Gemini training and grounding. No effect on Google Search	Your choice
OAI-SearchBot	OpenAI	ChatGPT search answers	Allow
GPTBot	OpenAI	Model training	Your choice
Claude-SearchBot	Anthropic	Claude search results	Allow
ClaudeBot	Anthropic	Model training	Your choice
PerplexityBot	Perplexity	Perplexity search results	Allow

OpenAI says sites that opt out of OAI-SearchBot are not shown in ChatGPT search answers (OpenAI docs). Check your firewall too: some CDN "block AI bots" settings catch search crawlers along with training ones.

2. Serve the main content as HTML. Many AI crawlers fetch the raw HTML and do not run JavaScript. If curl https://yoursite.com/page does not show your main text, they may see an empty page.

What this checklist does not cover

This is about how your site works. Other SEOmatic resources go deeper on the rest:

Each Page indexing reason (crawled or discovered not indexed, soft 404, duplicates): Page indexing fix guide
Search Console settings and reports, row by row: Search Console audit checklist
Internal links and orphan pages in depth: Internal linking playbook
Platform-specific settings: WordPress and Shopify SEO checklists
Titles and meta descriptions: Title and meta swipe file

Who this is for

Site owners and marketers who want a clear list they can hand to a developer, with the "where to check" spelled out.
Freelancers and agencies who need a repeatable technical audit template for client sites.
Developers who want the Google rules (status codes, canonicals, rendering, hreflang) in one place with sources.

How to use this with an AI assistant

You can run this checklist with ChatGPT, Claude, Cursor or any MCP client connected to SEOmatic. Connect SEOmatic at https://seomatic.ai/mcp, then paste one of the prompts below.

The free check needs no account. Reading your Search Console data is free after sign-in. Staging fixes needs the 14-day trial, and nothing goes live until you approve it.

No sign-in: check_site gives a free first look at one site.
After free sign-in: your assistant can read your Search Console data and audit your pages.
Staging fixes needs the 14-day trial. Nothing goes live until you approve it, and every applied change can be undone.

What the tools cannot do (so your assistant does not pretend): they do not read server logs, run Lighthouse or field Core Web Vitals data, or test hreflang sets across a whole site. site_audit covers up to 30 pages per run. For those rows of the checklist, the assistant should tell you which tool to use (listed in the spreadsheet) instead of guessing.

Copy-paste prompt

Prompt

Use SEOmatic to run the technical SEO checklist on mysite.com.

1. Run check_site on mysite.com. Report the homepage status, redirects, noindex, canonical, robots.txt rules for Googlebot and AI crawlers, JavaScript-only rendering, and the sitemap sample.
2. Sign me in with connect_seomatic, then call get_started for the overview.
3. Run site_audit. Group what it finds into the checklist sections: Indexability (noindex leaks), Canonicals (canonicals pointing elsewhere), Redirects and status codes (dead pages, broken internal links), Content (missing or duplicate titles and metas, near-empty pages), Speed (slow pages).
4. Check sitemaps with gsc_insights (action: sitemaps).
5. Take my 20 most important URLs (from gsc_performance, action: top_pages, last 28 days) and inspect them with gsc_indexing (action: batch_inspect). For each, compare the user-declared and Google-selected canonical and flag any page that is not indexed.
6. For any template that looks wrong, run site_pages (action: analyze) on one example URL and site_pages (action: crawl) to show its title, canonical, headings, links and schema.
7. Check for duplicate pages competing for the same searches with gsc_insights (action: cannibalization).
8. If AI crawler tracking is set up, call get_ai_crawler_activity and tell me which AI bots visit and whether robots.txt lets them in.
9. Give me one table: section, check, evidence (with the URL), fix, priority (High/Medium/Low). High = stops crawling or indexing, or affects every page. Say plainly which checklist rows you could not check with these tools.
10. For fixes SEOmatic can make, stage them with task_manage (action: create), up to 5 per call, then call show_approval_cards so I can approve or dismiss each one. Do not change anything live without my approval. If I want an applied change back, use task_manage (action: rollback).

Tools this prompt uses

MCP tool (action)	What it does
check_site	Free, no sign-in: homepage status, redirects, noindex, canonical, robots.txt for Googlebot and AI crawlers, JavaScript-only rendering, up to 10 sitemap pages, top fixes
connect_seomatic	Shows the free sign-in prompt
get_started	Overview of your site: verdict and top 3 actions with numbers
site_audit	Crawls up to 30 pages: duplicate titles and metas, missing metas, near-empty pages, noindex leaks (meta or X-Robots-Tag), canonicals pointing elsewhere, slow pages, dead pages, broken internal links
gsc_insights (sitemaps)	Health of every submitted sitemap: errors, warnings, pending, stale
gsc_insights (cannibalization)	Queries where two of your pages compete
gsc_performance (top_pages)	Your top pages by clicks from Search Console
gsc_indexing (batch_inspect)	Google's index status for up to 50 URLs per call. Google caps URL Inspection at 2,000 requests per day per property
site_pages (analyze)	One page: 0-100 score and issues ranked by impact
site_pages (crawl)	One page: title, meta, headings, links, images, schema, Open Graph
get_ai_crawler_activity	Which AI bots visit your site and how. Needs a one-time setup (WordPress plugin or Cloudflare Worker); otherwise it says visits are unmeasured and checks robots.txt access
task_manage (create, decide, rollback)	Stages fixes as proposals (never runs them), records your approve or dismiss, undoes an applied change
show_approval_cards	Shows Approve and Dismiss cards for staged fixes in the chat

Which checklist fixes SEOmatic can stage

Staged as tasks for your approval (applied in your CMS where the connection allows it, otherwise you get the exact change to paste): noindex removal, canonical fixes, broken internal link fixes, redirects (on Shopify, BigCommerce, Wix, HubSpot, Framer, Drupal, and WordPress with the SEOmatic Redirects plugin), robots.txt rules, structured data (schema and repairs), hreflang, image dimensions and LCP loading fixes (cwv_fix), alt text, headings, sitemap submission and index requests in Search Console.

Not done by SEOmatic: server, CDN or firewall settings, hosting speed, HTTPS certificates, JavaScript framework changes (SSR), and server log analysis.

Sources (24)