
Short answer: To improve website loading speed for SEO, measure real-user Core Web Vitals first, then fix the biggest bottleneck: usually a slow server response, an oversized hero image, render-blocking CSS and JavaScript, or heavy third-party scripts. Compress and resize images, enable caching and a CDN, defer non-critical scripts and remove unused plugins. Aim for LCP within 2.5 seconds, INP of 200 ms or less and CLS of 0.1 or less.
This guide is for site owners, marketers and developers who have seen a poor PageSpeed score or a failing Core Web Vitals report and want to know what to fix first. You will be able to read the right data, find the bottleneck on your own pages, work through fixes in priority order and judge when speed is no longer what is holding your rankings back.
Does website speed affect SEO?
Yes, but it helps to be precise about how. Google’s page experience documentation confirms that Core Web Vitals are used by its ranking systems. The same page is clear that Google always seeks to show the most relevant content, even if the page experience is poor, and that a great page experience helps most when many helpful pages are competing for the same query.
In practice that means speed is a tiebreaker and a conversion lever, not a substitute for good content. When I audit sites, I see three ways a slow site actually loses search traffic and revenue:
- Ranking: on competitive queries where several pages are equally relevant, a page that fails Core Web Vitals has a disadvantage.
- Behaviour: people abandon slow pages, especially on mobile data. The Deloitte study above found a 0.1 second improvement lifted retail conversions by 8.4% and travel conversions by 10.1%.
- Crawling: on large sites, a slow server means Googlebot fetches fewer pages in the time it allows, which delays indexing of new and updated content. My guide to crawl budget explains when this matters.
The metrics that matter
“Loading speed” is not one number. Google’s Core Web Vitals measure three parts of the experience, and a few supporting metrics help you diagnose them.
| Metric | What it measures | Good | Usual causes when it fails |
|---|---|---|---|
| LCP (Largest Contentful Paint) | When the main content, usually a hero image or headline, appears | 2.5 s or less | Slow server, large hero image, render-blocking CSS, lazy-loaded hero |
| INP (Interaction to Next Paint) | How quickly the page responds to taps, clicks and typing | 200 ms or less | Heavy JavaScript, page builders, chat widgets, tag managers |
| CLS (Cumulative Layout Shift) | How much the layout jumps while loading | 0.1 or less | Images without dimensions, late ads and banners, web fonts swapping |
| TTFB (Time to First Byte) | How long the server takes to start responding | Supporting metric | Cheap shared hosting, no page cache, slow database queries, distant server |
| Total page weight | Bytes downloaded for the page | Supporting metric | Uncompressed images, video backgrounds, unused CSS and JS libraries |
INP replaced First Input Delay as a Core Web Vital in March 2024. If a guide still tells you to optimise FID, it is out of date. For the details of each metric, see my guides on fixing INP and fixing layout shift.
Not sure if this is hurting your site?
Get the fixes that matter most, in priority order.
Step 1: Measure the right data
There are two kinds of speed data, and mixing them up is the most common mistake I see.
- Field data comes from real Chrome users over the last 28 days (the Chrome UX Report). This is what Google uses for page experience. You see it in Search Console’s Core Web Vitals report and at the top of PageSpeed Insights, where enough traffic exists.
- Lab data is a single simulated test, such as the Lighthouse score in PageSpeed Insights. It is useful for diagnosing causes and testing fixes, but the 0 to 100 score itself is not a ranking factor.
My routine: open Search Console’s Core Web Vitals report to see which groups of URLs fail and on which metric, then run a representative URL from each group through PageSpeed Insights or my free PageSpeed checker to see why. Test on mobile first, because Google indexes the mobile version and that is where most sites struggle.
Step 2: Estimate how heavy your pages are
Before tuning code, check the basics: how many megabytes does a typical page download, and how long does that take on the connections your visitors actually use? Many of my clients’ visitors in Pakistan browse on mobile data where real-world speeds vary widely through the day, so a 5 MB page that feels instant on office fibre can crawl on a phone.
Page weight download time calculator
How to read it: this is a simplified estimate (bytes divided by bandwidth, plus server response). Real loads also depend on latency, the number of requests, caching and rendering, so treat it as a sense check, not a prediction of your LCP. Get your actual page weight from my page size checker or the Network tab in Chrome DevTools.
Step 3: Fix speed problems in priority order
Most guides give you a list of 20 tactics with no order. This is the order I work in, because each layer limits the ones after it.
1. Server response and caching
If the server takes a second to respond, nothing else can start. On WordPress, a full-page cache (from your host, LiteSpeed Cache, WP Rocket or similar) is usually the single biggest win. Check that your hosting plan is not overloaded, that PHP is on a current version, and that the server is reasonably close to your audience; a site targeting Karachi served from a US data centre adds latency on every request. My guide to hosting, caching and CDNs goes deeper.
2. Use a CDN
A content delivery network serves static files, and often cached HTML, from locations near the visitor. Cloudflare’s free plan is a common starting point. It also handles Brotli or Gzip compression for text files, which a surprising number of sites still lack.
3. Images, especially the LCP image
- Resize images to the largest size they actually display at, and compress them. My free image compressor works for quick jobs.
- Serve WebP or AVIF where your CMS supports it.
- Add
widthandheightattributes so the browser reserves space and the layout does not shift. - Lazy-load images below the fold, but never the hero image. Give the hero
fetchpriority="high"instead.
4. Render-blocking CSS and JavaScript
Defer scripts that are not needed for the first view, remove unused CSS where your theme allows, and inline only a small amount of critical CSS. Load web fonts with font-display: swap and limit yourself to two families and a few weights.
5. Third-party scripts
Chat widgets, heatmaps, several analytics tags, social embeds and ad scripts often cost more than everything else combined, and they mainly hurt INP. List every third party in DevTools, ask who uses each one, and remove or delay the ones nobody needs.
6. Plugins, page builders and themes
Heavy page builders and plugin stacks add CSS and JavaScript to every page. Remove plugins you no longer use, avoid installing one plugin per small feature, and consider a lighter theme when you redesign. Check plugin settings after every major update.
7. Redirects
Each redirect hop adds a round trip. Link internally to final URLs and avoid chains like http to https to www to a trailing slash.
Worked example: a slow WordPress service page
A typical pattern I find on small business sites: a service page built with a page builder, a 2.4 MB PNG hero image, eleven plugins loading scripts on every page, a WhatsApp chat widget, and no page cache on shared hosting. Search Console flags LCP and INP as poor on mobile. The fix order would be: turn on full-page caching and Cloudflare; convert the hero to a properly sized WebP and mark it high priority; deactivate unused plugins and stop the slider and form plugins loading on pages that do not use them; replace the chat widget script with a plain WhatsApp link; then retest in PageSpeed Insights and wait for field data in Search Console to update over the following 28 days.
Common speed mistakes I see
- Chasing a 100 lab score. The Lighthouse score is a diagnostic. A page can score 70 and pass all Core Web Vitals in the field, and that is what matters for search.
- Testing only the homepage. Templates fail differently. Test a service page, a blog post, a category and a product page.
- Stacking optimisation plugins. Two caching or minification plugins fighting each other can break layouts and make things slower.
- Lazy-loading everything. Lazy-loading the hero image delays LCP.
- Ignoring mobile. Desktop scores are usually fine; the problems show on mid-range Android phones. See optimising Core Web Vitals for mobile.
- Expecting instant results. Field data is a rolling 28-day window, so Search Console takes weeks to reflect a fix.
- Treating speed as the reason for every traffic drop. If relevance or content quality is the problem, a faster page will not fix it.
When to stop optimising
Once your important templates pass all three Core Web Vitals in the field for mobile, further speed work rarely moves rankings. Keep monitoring so new plugins, scripts and images do not undo the work, and move your effort to content, internal links and authority. If you are not technical, my non-technical guide to website speed explains what to ask your developer for.
Frequently asked questions
How fast should a website load for SEO?
Google does not set a single load time. Its targets are Core Web Vitals: LCP within 2.5 seconds, INP of 200 milliseconds or less and CLS of 0.1 or less, for at least 75% of visits on mobile and desktop.
Is the PageSpeed Insights score a ranking factor?
No. The 0 to 100 score is a lab diagnostic. Google uses field data from real users, which you can see in Search Console’s Core Web Vitals report and the top section of PageSpeed Insights.
How long until speed improvements show in Search Console?
Field data covers the previous 28 days, so expect the report to update gradually over about a month after a fix. Any ranking effect, if there is one, comes after that.
Will a faster site rank higher on its own?
Only when speed was holding it back against equally relevant competitors. Google says it shows the most relevant content even when page experience is poor, so content and links still matter more.
Does a CDN help SEO?
Indirectly. It reduces the time for files to reach visitors far from your server and can lower server load, which helps LCP and crawling. It does not change rankings by itself.
What slows down a WordPress site the most?
In my audits: no page cache, cheap overloaded hosting, oversized images, heavy page builders and too many plugins and third-party scripts loading on every page.
Sources
- Google Search Central, “Understanding page experience in Google Search results” (last updated September 2026).
- web.dev, “Web Vitals” (last updated October 2024).
- web.dev, “Milliseconds make millions”, Deloitte study of 37 brand sites (2020).
- HTTP Archive, “Web Almanac 2024: Performance” (2024).
If your Core Web Vitals report has been red for months and you are not sure what is causing it, my website speed optimisation service starts with a template-by-template diagnosis, or you can send me the URL and I will tell you where I would start.
Request received
Thank you! Saad will contact you within the next 24 hours.
Keep an eye on your phone and inbox.
