Skip to content
Khepri Digital
Book a strategy call
SEO

Technical SEO Checklist for 2026: What Google Actually Checks

By Gordon Hwa, Founder, Khepri Digital
Updated · First published 27 March 2023 · 13 min read

What is technical SEO?

Technical SEO is the part of search engine optimization (SEO) that deals with how your website is built rather than what it says: whether search engines can reach each page, read its content, understand which URL is the main one, and load it fast enough for users. Content and links decide whether a page deserves to rank in search results. Technical SEO decides whether it gets the chance. It's the part of an SEO strategy that's easiest to check and most often skipped, because nothing looks wrong on your website until search traffic drops.

Google describes the same four-step process for every page:

1. CrawlGooglebot fetches the URL 2. Renderruns the JavaScript 3. Indexstores the canonical URL 4. Rankshows it for a search WHAT BREAKS IT robots.txt blocks5xx server errorsno internal linksredirect chains content only afterJavaScript that failsblocked CSS or JSlinks without href noindex left onduplicate URLswrong canonicalsoft 404s relevance, qualitylinksCore Web Vitalsmobile version
The four steps from Google Search Central's documentation, and the most common technical faults we find at each.

A fault at an early step can't be fixed at a later one. A page Google can't crawl will never be indexed, however good the content. That's why a technical SEO checklist starts with crawlability.

Can search engines crawl your site?

Crawlability is whether search engines can reach the pages on your site. Most problems come from robots.txt directives, server errors or settings in your CMS that nobody remembers changing. Check four things:

  1. Your robots.txt file. It lives at yoursite.com/robots.txt and tells crawlers which paths they may fetch. One stray Disallow: / blocks the whole site. Google is clear about its limit: robots.txt "is not a mechanism for keeping a web page out of Google". A blocked page can still be indexed, without its content, if other pages link to it.
  2. Server responses. Pages should return 200. Google treats every 4xx error the same way, except 429 (too many requests), and slows its crawling when it sees many 5xx server errors.
  3. Internal links. Google finds most pages by following links. Use real <a href> links; Google says it can't reliably follow links made from JavaScript click handlers without an href.
  4. Crawl budget. Most sites don't need to worry about it. Google's crawl budget guide is written for sites with over a million pages, or over 10,000 pages that change daily. If you have a few hundred pages, fix errors and redirects instead.

Can Google render and index your pages?

Indexability is whether a crawled page can be stored in Google's index. Check:

  • No accidental noindex. A noindex meta tag or X-Robots-Tag header keeps a page out. It only works if Google can crawl the page, so don't block a noindexed page in robots.txt as well: Google will never see the instruction.
  • Content in the HTML. Google renders JavaScript, but rendering can be delayed and can fail. Important content and links are safest in the HTML the server sends. Check with URL Inspection in Google Search Console, which shows the rendered page Google sees.
  • One canonical URL per page. If the same content is available at several URLs (with and without www, with tracking parameters, http and https), tell Google which one is main. Google ranks the signals: redirects and rel=canonical are strong, and sitemap inclusion is weak.
  • Mobile content matches desktop. Mobile-first indexing has applied to all sites since 31 October 2023, so Google indexes the mobile version. Anything missing on mobile is missing from Google.

The Search Console Page indexing report lists every URL Google knows about and why each one isn't indexed. Read it monthly.

Do your sitemap and redirects work?

XML sitemaps list the URLs you want indexed. Google's rules:

  • Up to 50 MB or 50,000 URLs per file. Use a sitemap index for more.
  • Only canonical URLs that return 200. No redirects, no noindexed pages.
  • Google ignores the priority and changefreq values. It uses lastmod only if it's consistently accurate.
  • Submit it in Search Console and reference it in robots.txt.

Redirects move users and search engines from an old URL to a new one. A 301 and a 308 are both permanent and treated the same; 302 and 307 are temporary. Google follows up to 10 redirect hops, but every hop wastes time, so point each old URL straight at its final page.

When we relaunched khepridigital.com in October 2026, we mapped every URL on the old Webflow site: 91 old addresses now redirect in a single hop to the page that replaced them, the removed pages return 410, and the launch check found 0 broken internal links. That's the boring work that keeps a redesign from losing search traffic.

Do your pages pass Core Web Vitals?

Core Web Vitals are three measures of user experience on a web page: how fast your content appears, how quickly the page responds, and how stable it stays. They're the only part of page experience Google says is used in ranking: "Core Web Vitals are used by our ranking systems."

Metric What it measures Good Poor
Largest Contentful Paint (LCP) How fast the main content loads 2.5 s or less Over 4 s
Interaction to Next Paint (INP) How fast the page responds to taps and clicks Under 200 ms Over 500 ms
Cumulative Layout Shift (CLS) How much the layout jumps while loading Under 0.1 Over 0.25

A page passes when it's "good" on all three, measured at the 75th percentile of real visits on mobile and desktop. INP replaced First Input Delay (FID) as a Core Web Vital on 12 March 2024; any checklist that still mentions FID is out of date.

Most websites don't pass. The 2025 Web Almanac, built from Chrome's real-user data, found 48% of mobile websites and 56% of desktop websites had good Core Web Vitals.

Our own site was one of them. Before the relaunch, Google's Lighthouse gave khepridigital.com's main pages a mobile performance score of 27 to 51, with the largest content taking 7 to 21 seconds to appear and a cumulative layout shift as high as 0.35. On the rebuilt site the same pages score 98 to 100, load the main content in about 1.5 to 2.2 seconds, with a layout shift of 0 to 0.002. Those are lab tests on 26 October 2026, not field data; field data in Search Console takes about 28 days to catch up.

The usual causes and fixes:

  • Slow LCP: an oversized hero image or video, slow hosting, render-blocking scripts. Serve modern image formats at the right size, preload the main image, and load tag managers and chat widgets after the page.
  • Poor INP: heavy JavaScript on the main thread, often from third-party tags. Remove what you don't use.
  • High cumulative layout shift: images and embeds without set dimensions, banners and fonts that load late. Reserve space for everything.

Is your site secure and mobile-friendly?

  • HTTPS everywhere, with http redirecting to https. The Web Almanac puts HTTPS adoption above 91% of sites in 2025; being in the other 9% looks careless to users and browsers.
  • A responsive site with a viewport meta tag, readable text without zooming, and tap targets that aren't crammed together.
  • No intrusive interstitials: pop-ups that cover the content on mobile, especially on arrival from search.

Does your structured data still earn anything?

Structured data (schema.org markup, in JSON-LD) helps Google understand what a page is about, and makes some pages eligible for rich results. Half of websites now use it: 50% of home pages in the 2025 Web Almanac, 43% in JSON-LD.

What's changed:

  • FAQ rich results are gone. Google stopped showing them in May 2026 and removed the documentation in June 2026. HowTo rich results went in 2023, and Google phased out seven more types in June 2025.
  • Mark up what's still supported and matches what's visible on the page: Organization, LocalBusiness or a more specific type (such as Hotel or MedicalClinic), Article, Product, BreadcrumbList, Event, Review snippets on eligible types.
  • Never mark up content users can't see, or ratings you gave yourself. Both break Google's structured data policies and can lead to a manual action.

Our 2026 audit of 12 hotels found 9 had no Hotel schema and 4 had no structured data at all (the hotel AI search audit). It's one of the cheapest technical fixes there is.

What about AI search and AI crawlers?

Google says there are "no additional requirements" to appear in AI Overviews or AI Mode: a page needs to be indexed and eligible for a snippet, and normal SEO best practice applies. No special files, no special markup. Google added in June 2026 that llms.txt files "aren't needed for Google Search" and won't help or hurt rankings. Only about 2% of sites have one anyway.

Other AI companies run their own crawlers, which you control in your robots.txt file:

Company Crawler for search answers Crawler for model training
OpenAI OAI-SearchBot GPTBot
Anthropic Claude-SearchBot ClaudeBot
Perplexity PerplexityBot None (Perplexity says PerplexityBot isn't used for training)
Google Googlebot (AI Overviews use the normal index) Google-Extended (Gemini training only)

Blocking a search crawler keeps you out of that company's answers. Blocking a training crawler doesn't. Decide on purpose: in our hotel audit, one of the 12 hotels was blocking an AI crawler. GPTBot is named in about 4.5% of robots.txt files, so most sites have made no choice at all.

For Google, the control is in Search Console: the Search generative AI control manages your site's inclusion in AI Overviews and AI Mode, and Google added it to its documentation in October 2026 alongside the robots meta tag, data-nosnippet and X-Robots-Tag. Opting out keeps you out of those answers, so use it only if you mean to.

Being crawlable by each search engine is the start. Being cited by AI answers depends on consistent facts and clear pages, which is what our AI search visibility work covers.

The technical SEO checklist

Use this in order. Each line is something you can check today.

Crawling

  • robots.txt returns 200, blocks nothing you want found, and lists your sitemap
  • Key pages return 200; no spikes in 5xx errors in Search Console's Crawl stats
  • Every page you want found has at least one normal internal link pointing to it
  • AI crawlers allowed or blocked on purpose

Rendering and indexing

  • URL Inspection shows the main content and links in the rendered HTML
  • No noindex on pages you want indexed; noindexed pages aren't blocked in robots.txt
  • One canonical URL per page, self-referencing, on https and your preferred host
  • Mobile and desktop show the same content
  • Page indexing report reviewed; every "not indexed" reason understood

Sitemaps and redirects

  • XML sitemap lists only canonical 200 URLs, with accurate lastmod
  • Old URLs redirect in one hop (301 or 308) to the closest equivalent page
  • No internal links pointing at redirects or 404s

Core Web Vitals and site structure

  • LCP 2.5 s or less, INP under 200 ms, CLS under 0.1 in Search Console's Core Web Vitals report
  • HTTPS on every URL; http redirects to https
  • Site structure no deeper than three clicks from the home page for important pages
  • Descriptive URLs with hyphens, not IDs or underscores

Structured data

  • Organization (and LocalBusiness or a specific type) on the home page
  • Article markup on posts, with author and dates
  • No markup added only for retired rich results, no self-awarded ratings
  • Rich Results Test passes on each template

How often should you run a technical SEO audit?

Run the full checklist after any redesign, platform move or domain change, and before launch, not after: a migration is when search rankings are most often lost. Otherwise once a quarter, plus a monthly look at Search Console's Page indexing and Core Web Vitals reports. Most of the problems we find aren't hard to fix. They've just gone unnoticed for months.

Technical SEO is the foundation of our SEO service, and the first thing we check for every new client. If you're weighing outside help, here's what SEO costs in Singapore.

Frequently asked questions

What is a technical SEO checklist?

A list of checks that make sure search engines can crawl, render and index your website and that pages load well for users: robots.txt, status codes, canonical URLs, sitemaps, redirects, Core Web Vitals, HTTPS, mobile and structured data.

Does robots.txt stop a page being indexed?

No. robots.txt stops crawling. A blocked URL can still be indexed if other pages link to it. Use a noindex tag, and allow crawling so Google can see it.

What are good Core Web Vitals scores in 2026?

Largest Contentful Paint of 2.5 seconds or less, Interaction to Next Paint under 200 milliseconds and Cumulative Layout Shift under 0.1, at the 75th percentile of real visits.

Do I need an llms.txt file?

Not for Google. Google says llms.txt files aren't needed for Google Search and don't affect rankings. Other services may read them, but we haven't seen one decide whether a site gets cited.

Is FAQ schema still worth adding?

Not for rich results. Google stopped showing FAQ rich results in May 2026. Write the questions and answers on the page for readers. The markup does no harm, and other search engines and AI tools may read it, but it earns nothing visible in Google.

How long does it take to see results from technical fixes?

Google needs to recrawl the pages first, which can take days to weeks. Core Web Vitals field data in Search Console covers the previous 28 days, so improvements show there about a month after the fix.

Sources

Rewritten in October 2026. The first version of this article was published in March 2023.

This guide supports our SEO service.

More on SEO

See what your customers see when they search for you.

A 30-minute call with the people who would run your account. We look at your ads, your pages and what AI says about your brand, live.

Book a strategy call
Fixed monthly fee. Every account stays in your name.
Book a strategy call