Technical SEO is the invisible layer of search rankings. When it works well, nobody notices. When it fails, even the most carefully crafted content won’t rank because Google cannot crawl, index, or understand the website correctly. This guide covers all the technical SEO elements you need to review, how to spot the problems, and how to fix them.
Core Web Vitals: the performance factor Google measures
Google uses three user experience metrics as a direct ranking factor since 2021. Understanding what they measure and how to improve them is the starting point for technical SEO.
Largest Contentful Paint (LCP): measures how long it takes for the largest visible element in the viewport to load. The threshold is 2.5 seconds. The most common issues that degrade it are unoptimised images, server response time, and render-blocking JavaScript.
To improve LCP: compress images and serve them in WebP format, implement lazy loading for images outside the viewport, use a CDN to reduce latency, and remove or defer non-critical JavaScript.
Cumulative Layout Shift (CLS): measures visual stability — how much elements move while the page loads. The threshold is 0.1. The most common culprits are ads without reserved dimensions, images without size attributes, and fonts that shift the layout on load.
To improve CLS: always define width and height attributes on images, reserve space for adverts before they load, and use font-display: swap for custom fonts.
Interaction to Next Paint (INP): replaced First Input Delay in 2024.Measures the response to all user interactions during the visit, not just the first. The threshold is 200 milliseconds. The main issues are heavy JavaScript that blocks the main thread.
Crawlability and indexation
Google needs to be able to crawl your website in order to index it. There are several elements that can block that process.
robots.txt: the file that tells Google what it can and cannot crawl. A common mistake is accidentally blocking important directories. Check in Search Console that no pages blocked by robots.txt should actually be indexed.
XML Sitemap: it must include all the pages you want Google to index, with the last modified date kept up to date. It must be referenced in the robots.txt and submitted in Google Search Console.
Crawl errors: 404 errors on URLs that had organic traffic are direct losses of ranking. You need to implement 301 redirects to the most relevant equivalent content. 5xx server errors completely prevent those pages from being indexed.
Pagination: paginated results pages need a clear strategy — rel=next/prev is deprecated. The current solution is the canonical pointing to the first page, or the use of infinite scroll with the correct signals.
URL structure
URLs have an impact on SEO and user experience. The basic principles: short and descriptive URLs, without unnecessary parameters, with the main keyword included naturally, in lower case and with hyphens instead of underscores.
Common problems include dynamic URLs with multiple parameters that generate duplicate versions of the same content, uppercase URLs that create duplicate versions, and flat structure without hierarchy that fails to convey the site’s architecture.
Canonical tags
The canonical tag tells Google which is the main version of a page when several URLs display the same content. It is especially critical in e-commerce (where a product with different filters generates dozens of URLs) and on websites with HTTP/HTTPS or www/non-www versions.
The most common error is the canonical that points to itself when it should point to another URL, or the complete absence of canonicals on websites with dynamic content.
Schema markup
Schema markup is structured code that tells Google exactly what type of content each page contains. It is not a direct ranking factor but it does influence rich results — the enhanced results in the SERP that achieve a higher CTR.
The most relevant schema types by business type: Organization and LocalBusiness for the home page and contact pages, Service for service pages, Product for e-commerce, Article for the blog, FAQ for pages with frequently asked questions, Review for pages with testimonials.
Hreflang for multilingual websites
Hreflang tells Google which version of a page to show to users in each language and country. A poorly implemented hreflang can cause Google to index the wrong version for each market.
The most common errors: hreflang tags that are not reciprocal (the EN version must point to the ES and vice versa), hreflang tags with URLs that do not exist or that redirect, and hreflang tags with incorrect language codes.
The auditing tool that detects all of this
The manual audit process is slow and error-prone. Tools that automate most of it: Screaming Frog for full site crawling, Google Search Console for real indexation data and errors, PageSpeed Insights for Core Web Vitals, and Ahrefs or Semrush for backlink analysis and position tracking.
Our free SEO Auditor covers the most critical elements in an automated way and generates a report with prioritised recommendations.