One new layer, AI crawler access, and nineteen old ones that still quietly decide rankings. Every item below is pulled from real client audits, not theory.
Technical SEO Checklist for 2026: 20 Fixes That Still Move Rankings
One new layer, AI crawler access, and nineteen old ones that still quietly decide rankings. Every item below is pulled from real client audits, not theory.
Fix in order. Indexation problems mask everything downstream, so each phase depends on the one before it.
Technical SEO in 2026 has one new layer and nineteen old ones that still decide rankings. The new layer is AI crawler access, because AI engines cannot cite what they cannot crawl. The old layers are the crawlability, indexation, and rendering fundamentals that quietly break on real websites every week.
This checklist comes from the technical audits we run for clients on WordPress, Shopify, and custom stacks. Every item below is something we have actually found and fixed on a live site, not a theoretical best practice. Work through it top to bottom; the ordering reflects how much each layer depends on the one before it.
01Crawlability and Indexation (Items 1 to 6)
- Check robots.txt against reality. Fetch yourdomain.com/robots.txt and read it line by line. We regularly find staging rules that shipped to production, blanket Disallow rules from security plugins, and blocked CSS or JS directories that break rendering. If Googlebot cannot render your page the way users see it, everything downstream suffers.
- Reconcile your sitemap with your index. Your XML sitemap should contain only canonical, indexable, 200-status URLs. Compare sitemap count to the indexed count in GSC's Pages report. A large gap in either direction is a diagnosis waiting to happen.
- Audit the "Crawled, currently not indexed" bucket. This GSC status is where thin, duplicative, or low-value pages accumulate. On one Shopify client audit, hundreds of near-identical filtered collection URLs sat here, diluting crawl attention from pages that mattered. Consolidate, noindex, or improve; do not ignore.
- Hunt orphan pages. Crawl your site with Screaming Frog, then compare against sitemap and GSC data. Pages with zero internal links pointing at them rarely rank, and important pages end up orphaned more often than you would expect after redesigns.
- Fix redirect chains and loops. Every hop wastes crawl budget and dilutes signals. Anything beyond a single 301 hop should be flattened, especially in template-level links like navigation and footers.
- Verify your important pages return 200 to bots. Firewalls and bot-protection services sometimes serve challenge pages to crawlers. Test with the URL Inspection tool, not just your browser.
02Canonicals and Duplication (Items 7 to 10)
- Self-referencing canonicals on every indexable page. Simple, and still commonly broken by theme updates and plugins that overwrite each other.
- Parameter URLs must canonicalize to the clean version. Sort orders, session IDs, tracking parameters, and Shopify's ?variant= URLs should all point back to the primary URL. When we audited a Shopify apparel client's GSC coverage, parameter duplication was the single largest source of wasted crawl and split signals.
- Check for protocol and host duplication. http vs https, www vs non-www, and trailing-slash variants should all 301 to one canonical form. Test all four combinations of your homepage manually.
- International sites: validate hreflang reciprocity. Every hreflang annotation must be confirmed by a return tag on the target page. When we rebuilt hreflang for Scriveiner London across 13 Shopify markets, broken reciprocity was causing the wrong country pages to surface in the wrong markets, which is a revenue problem disguised as a technical one.
03Speed and Rendering (Items 11 to 14)
- Measure field data, not just lab data. Use the CrUX data in PageSpeed Insights, because Google ranks on real-user experience. Lab scores are for debugging, not for judging.
- Fix the largest LCP offender first. On most content and e-commerce sites it is an unoptimized hero image. Serve modern formats, size images to their containers, and preload the LCP asset. One fix, most of the gain.
- Kill layout shift at the template level. Reserve space for images, embeds, and ads with explicit dimensions. CLS problems are almost always template problems, so one fix repairs thousands of pages.
- Defer what does not paint. Third-party scripts, chat widgets, and tag-manager bloat are the usual suspects behind poor interactivity. Audit your tag manager quarterly; tags outlive the campaigns they were installed for.
04Structured Data (Items 15 to 17)
Structured data is the entity foundation both Google and AI engines build on:
Organization with full NAP and sameAs profiles, BreadcrumbList sitewide, and WebSite, on every site regardless of vertical.
Product with offers on e-commerce, Service plus LocalBusiness for service businesses, Article or BlogPosting with real author entities on content. Validate with the Rich Results Test.
FAQ schema must mirror the on-page FAQ text, dates must reflect real publish dates, and ratings must exist on the page.
Mismatches risk manual actions and, just as damaging, erode the trust signals AI engines rely on when deciding what to cite.
05The AI Crawler Layer (Items 18 to 20)
AI engines cannot cite what they cannot crawl, which makes this the newest layer of the checklist and, for most brands, the most neglected:
- 18 — Decide your AI crawler policy deliberately. Check robots.txt and CDN settings for GPTBot, Google-Extended, PerplexityBot, and ClaudeBot. Blocking them is a legitimate business choice, but it should be a choice, not a plugin default.
- 19 — Serve full content without JavaScript where possible. Several AI crawlers render JavaScript poorly or not at all. If your main content only exists after client-side rendering, you are invisible to part of the AI layer.
- 20 — Add an llms.txt file for your key resources. An emerging convention, not a standard, but it costs little and gives AI systems a curated map of your most citable content.
How to Run This as an Audit
Run items 1 to 6 first and fix what you find before touching anything else; indexation problems mask everything downstream. Then duplication, then speed, then schema, then the AI layer. Re-crawl after each phase rather than at the end, because fixes interact.
For a typical small-to-mid site this is two to three focused days of work. For large e-commerce catalogs, faceted navigation and parameter handling alone can be a project. If you would rather hand this off, a technical audit against this exact checklist is part of our SEO service work, and pairing it with the AI visibility audit gives you both halves of the 2026 picture.
Get a Free 20-Point Technical SEO Audit
We'll run your site against this exact checklist, flag what's costing you crawl budget and rankings, and hand you a prioritized fix list.
Related Reading
-
GUIDE
What Is GEO? AI Search Visibility in 2026
The AI visibility audit that pairs with this checklist — how to measure and earn citations across AI engines once crawl access is sorted.
-
CASE STUDY
Novoresume Growth Case Study
The same schema and content architecture discipline behind this checklist, applied to an e-commerce client's crawl and indexation problems.
-
COMING SOON
The Full Guide to llms.txt
A deeper walkthrough of Item 20 — how to structure an llms.txt file so AI systems can find your most citable content.
FAQ
Crawlability, same as every year. If search engines and AI crawlers cannot reach and render your pages, no other optimization matters. The 2026 addition is confirming that AI crawlers like GPTBot and PerplexityBot are part of that access check.
A full audit twice a year, plus a monthly check of GSC's Pages report and Core Web Vitals. Also audit immediately after any migration, redesign, or platform change, which is when most technical damage happens.
Yes, as part of Google's page experience signals, though content relevance outweighs them. Their bigger commercial impact is on conversion: faster pages convert better regardless of rankings.
Directly. AI engines depend on crawl access, clean structured data, and server-rendered content to understand and cite pages. A technically broken site underperforms in AI answers for the same reasons it underperforms in classic search.
Raj — Founder, Alpha Geeks Technologies
Raj is the founder of Alpha Geeks Technologies, a digital marketing agency serving clients across the UK, US, Australia, Hong Kong, and UAE. Raj and the Alpha Geeks team specialize in SEO, AI/GEO visibility, and conversion-focused web development, with hands-on experience running technical audits and schema implementations for clients across e-commerce, healthcare, and local service industries.