Technical SEO Checklist for 2026
This is the checklist our team works through at the start of every technical audit. It is ordered by dependency: there is little point fixing structured data on pages that search engines cannot index.
How to use this checklist
- Work top to bottom. Each section depends on the ones before it.
- Check templates, not individual pages. Most technical problems repeat across every page built from the same template.
- Use field data and real crawls. Assumptions about how a site behaves are often wrong.
1. Crawling
Search engines need to reach your pages before anything else can happen.
| Check | What good looks like | Common failure |
|---|---|---|
| robots.txt | Blocks only what genuinely should not be crawled | Staging rules copied to production, blocking whole sections |
| Server responses | Important URLs return 200; removed pages return 404 or 410 | Soft 404s: “not found” pages returning 200 |
| Crawl traps | Filters, calendars and session IDs do not create infinite URLs | Faceted navigation generating millions of combinations |
| Server log data | Crawlers spend time on valuable URLs | Most crawl activity going to parameters and old redirects |
| Response time | Consistently fast server responses | Slow responses that reduce how much gets crawled |
If you have access to server logs, use them. They are the only direct record of what search engine crawlers actually request, as opposed to what a crawling tool assumes they request.
2. Rendering
Many sites now rely on JavaScript to display content. Search engines can render JavaScript, but rendering is resource-intensive and can be delayed.
- Compare the raw HTML with the rendered page. Is the main content, including headings and internal links, present in the initial HTML?
- Check that navigation links are real
<a href>elements, not click handlers. - Confirm that lazy-loaded content appears without user interaction such as scrolling or clicking.
- Test important templates with a URL inspection tool to see the page as the search engine renders it.
Server-side rendering or static generation for important content removes most of this risk.
3. Indexation
Being crawled is not the same as being indexed. This section is where we most often find large, fixable problems.
Canonical tags
Every indexable page should have a canonical tag pointing to its preferred URL. Look for:
- Canonicals pointing to redirected, non-indexable or 404 URLs
- Paginated pages all canonicalising to page one, which hides deeper content
- Filtered URLs self-canonicalising when they should point to the parent category
- Conflicts between canonical tags, sitemaps and internal links
Noindex directives
Check that noindex is applied deliberately. A noindex left on a template after a site launch is one of the most damaging and easiest-to-miss errors in SEO.
Duplicate and near-duplicate content
Protocol variants (http and https), trailing slashes, uppercase URLs and tracking parameters can all create duplicates. Each should resolve to one version through redirects or canonicals.
Index coverage
Compare three numbers: pages you want indexed, pages in your XML sitemaps, and pages search engines report as indexed. Large gaps in either direction are worth investigating. We cover the most common causes in 15 technical SEO problems that quietly hurt organic visibility.
4. Site architecture and internal linking
- Important pages should be reachable within a few clicks of the homepage.
- Category and hub pages should link to their most important children, and children should link back.
- Orphan pages, those with no internal links, should be linked or removed.
- Anchor text should describe the destination page.
- Breadcrumbs should reflect the real hierarchy and be marked up with structured data.
Internal linking is one of the few ranking factors entirely under your control. On most sites we audit, it is also one of the most neglected.
5. Redirects
- Redirects should be single hops wherever possible. Chains of three or more slow crawlers and users.
- Use permanent (301 or 308) redirects for permanent moves.
- After migrations, update internal links to point directly at final URLs rather than relying on redirects.
- Keep a redirect map from every migration, and keep important legacy redirects in place for the long term.
6. XML sitemaps
- Include only canonical, indexable URLs that return 200.
- Split large sites into multiple sitemaps by section so coverage can be monitored per section.
- Keep
lastmoddates accurate. Updating every date on every build teaches search engines to ignore them. - Reference sitemaps in robots.txt and submit them in search console tools.
7. Performance and Core Web Vitals
Use field data from real users wherever available, and diagnose by template.
| Metric | Measures | Typical causes of poor scores |
|---|---|---|
| Largest Contentful Paint | How quickly the main content appears | Large unoptimised hero images, slow servers, render-blocking resources |
| Interaction to Next Paint | How quickly the page responds to input | Heavy JavaScript, long tasks, third-party scripts |
| Cumulative Layout Shift | How much the layout moves while loading | Images without dimensions, late-loading ads and banners, web fonts |
Performance is rarely a dramatic ranking lever on its own, but it affects every visitor and every conversion.
8. Structured data
- Use schema types that accurately describe visible content: Organization, Article, Product, BreadcrumbList, LocalBusiness and so on.
- Do not mark up content that users cannot see. FAQ markup should only appear where the questions and answers are visible on the page.
- Validate templates after each release; structured data often breaks silently.
9. International and multi-language sites
- Hreflang annotations should be reciprocal: each language version references all others, including itself.
- Hreflang URLs must be canonical and indexable.
- Avoid automatic redirects based on IP address, which can prevent crawlers from seeing other versions.
- Translate meaningfully. Machine-translated pages with no local adaptation rarely perform.
10. Monitoring after the audit
An audit describes a site on the day it was run. Releases, CMS changes and new content constantly change the picture. Set up:
- Scheduled crawls of key templates, with alerts on changes to status codes, canonicals, noindex and titles
- Weekly checks of index coverage
- Monitoring of Core Web Vitals field data by template
This continuous monitoring is exactly what Advino AI’s website intelligence is designed for. Whether you use our platform or your own tools, the principle is the same: technical SEO is maintenance, not a one-off project.
If you would like our technical SEO team to work through this list on your site, we are happy to talk.
