Technical SEO playbook for 2026
Crawl budget, rendering paths, and the schema mix that survives the next core update.
Crawl budget is real on sites over 10k URLs
Below 10,000 URLs crawl budget is rarely a constraint. Above that, the architecture starts to matter: parameterised URLs balloon the crawl frontier, faceted navigation creates infinite spaces, and Googlebot starts deprioritising thin or duplicate clusters. Fix it with canonicals, robots.txt directives on parameterised paths, and a sitemap that emits only canonical URLs.
JS rendering is the silent killer
Client-side rendered apps still index, but with a delay (the two-wave indexing problem). Critical content, internal links, structured data, and canonicals MUST be in the initial HTML response. Use SSR or static generation for any page that needs to rank. The cost-of-edge-rendering argument lost three years ago.
Schema coverage is a ranking signal
Google's 2024 confirmed: structured data does not boost rankings directly, but it is required to compete for rich results, and rich results compound CTR. Cover the basics: Article, FAQPage, Product, LocalBusiness, BreadcrumbList. Use JSON-LD, not microdata.
Common questions
Is XML sitemap still required?
Yes for any site over a few hundred URLs. Submit via GSC, include only canonicals, and split into a sitemap index once you exceed 50,000 URLs per file.
Should I use lazy loading?
For images below the fold, yes. For LCP candidate images, no — eager-load and preconnect to the CDN.