Technical SEO has never been more consequential. As AI-driven search evolves and Google continues tightening its quality signals, the gap between sites that are technically sound and those that are not is widening fast. This checklist covers every layer of technical SEO that matters heading into 2026, from performance thresholds to schema implementation and the emerging challenge of AI crawler readiness.
Google's Core Web Vitals remain a confirmed ranking signal, and in 2026 there are three metrics you need to nail. For a full breakdown, refer to Google's Core Web Vitals documentation.
Largest Contentful Paint (LCP) measures how quickly the main content of a page loads. The target is under 2.5 seconds. Anything above 4 seconds is flagged as poor. LCP is most commonly impacted by unoptimised images, slow server response times, and render-blocking resources.
Interaction to Next Paint (INP) replaced First Input Delay (FID) in March 2024. INP measures overall responsiveness across all interactions during a page visit, not just the first. A good INP score is under 200 milliseconds. JavaScript bloat is the primary culprit when INP scores suffer.
Cumulative Layout Shift (CLS) measures visual stability. A score below 0.1 is good. Common causes of poor CLS include images without explicit dimensions, dynamically injected content, and web fonts that load without fallbacks.
Use Google PageSpeed Insights and Google Search Console to monitor these metrics across your real user data, not just lab scores.
Structured data does two things: it helps search engines understand your content precisely, and it makes you eligible for rich results that increase click-through rates. In 2026, the following schema types from schema.org are the highest priority:
Article and BlogPosting: Signals editorial content clearly. Critical for news eligibility and AI answer engine citations.
FAQPage: Enables FAQ rich results and is heavily used by AI answer engines to extract direct responses.
HowTo: Structured step-by-step formats are preferred by both featured snippet algorithms and generative AI.
Product: Essential for e-commerce. Enables pricing, availability, and review stars in search results.
Organization: Establishes your brand identity with Google, including logo, contact details, and social profiles.
BreadcrumbList: Improves site architecture signals and displays navigational context in search results.
Validate all structured data using Google's Rich Results Test and deploy via JSON-LD in the page head where possible.
Search in 2026 is not just Google. AI assistants including ChatGPT, Claude, and Perplexity are actively crawling and indexing web content to power their responses. Each crawler behaves differently, and your robots.txt file needs to reflect deliberate choices about access.
Googlebot remains the primary crawler. Ensure your sitemap is submitted via Google Search Console and that Googlebot is not inadvertently blocked by overly restrictive robots.txt rules.
GPTBot is OpenAI's crawler. By default, it will crawl your site unless blocked. To allow or disallow: add User-agent: GPTBot directives to your robots.txt file. Allowing it means your content may appear in ChatGPT responses, which is a significant distribution opportunity.
ClaudeBot is Anthropic's crawler. The same principle applies. Decide whether visibility in Claude's responses is valuable to your audience, then configure accordingly.
PerplexityBot powers Perplexity.ai's citation-based answers. Given Perplexity's growth among research-oriented users, blocking it is generally not in your interest unless content sensitivity demands it.
Audit your sitemap for consistency with your robots.txt. Pages blocked from crawling should not appear in your sitemap.
Crawl budget matters most for large sites (thousands of pages), but the underlying principle applies universally: every internal link is a signal of importance. Orphaned pages, excessive redirect chains, and paginated archives with no noindex directives all waste crawl budget.
Conduct a quarterly internal link audit. Ensure your highest-value pages receive the most internal links. Tools like Viaduct Generation's Blueprint service can help map your site architecture and identify structural gaps.
Google indexes the mobile version of your site first. If your mobile experience renders differently from desktop, or if mobile pages are missing content that exists on desktop, your rankings will suffer.
Key checks: ensure your mobile viewport meta tag is configured, tap targets are at least 44x44 pixels, font sizes are readable without zooming, and content is not hidden behind interstitials that Google cannot process.
These are foundational and non-negotiable.
HTTPS: All pages must be served over HTTPS. Mixed content warnings, where secure pages load non-secure resources, damage trust signals and can suppress rankings.
Canonical tags: Use the rel="canonical" attribute to signal your preferred URL version. This is especially important for paginated content, filtered product pages, and any syndicated content.
Hreflang: If you serve multiple language or regional variants, hreflang attributes prevent duplicate content issues and ensure the correct version ranks in the correct market.
Technical SEO is not a one-time project. It requires ongoing monitoring, scheduled audits, and a structured approach to fixing what breaks. If you want a diagnostic view of your current technical health, Viaduct Generation's Intelligence service surfaces technical issues alongside commercial opportunity data, so you fix the things that matter most first.
A technically sound site is the foundation every other growth activity depends on. Get this right and the returns from content, links, and AEO compound significantly.
