Website Load Speed — How Core Web Vitals Affect SEO and Sales [2026]
Website load speed and SEO vs. conversion. Data: -4.42% conversion/second, 38% WordPress vs 67% Astro in CWV. How to improve results and stop losing customers.
Website Load Speed — Why It Determines SEO and Sales
You have 3 seconds. That’s how long the average user gives you before closing the tab and moving to a competitor. Website load speed is not a technical curiosity — it’s a direct factor affecting your company’s revenue, Google position and conversion rate. In 2026, Google no longer simply asks whether your site is fast — it measures it quantitatively through a set of metrics called Core Web Vitals, and includes those scores in search rankings.
This article will explain exactly what Core Web Vitals measure, why WordPress achieves only a 38% pass rate, how modern technologies like Astro and Next.js reach 67%, and — most importantly — what you can do to make your site faster and attract more customers.
Hard Data — How Speed Affects Conversion and Google Position
Before we get into the technical details, look at the numbers. They’re unambiguous enough to persuade any CEO or COO:
- Every additional second of page load time reduces conversion by 4.42% — this is the Portent 2019 study result, confirmed many times by subsequent analyses. If your site loads in 4 seconds instead of 1 second, you lose about 13% of potential clients purely due to waiting time.
- 53% of mobile users abandon a site if it takes longer than 3 seconds to load — Google data from the “The Need for Mobile Speed” study. More than half of your mobile visitors will simply give up before the page even displays.
- Improving LCP by 0.1 seconds increases conversion by 8.4% — this is the Deloitte study conducted for Google, published in the “Milliseconds Make Millions” report. For a company generating 500,000 PLN annually from its website, this improvement means 42,000 PLN in additional revenue.
- Pages in the top position on Google load in an average of 1.65 seconds — Backlinko analysis of a million search results. The correlation between speed and position is strong and statistically significant.
These data points share one common denominator: website load speed is not a matter of aesthetics or technical preferences. It’s a measurable business factor that affects revenue as directly as product price or customer service quality.
Speed and the Google Algorithm
Google officially confirmed in 2021 that Core Web Vitals had become a ranking factor — the so-called “Page Experience Update.” This means two pages of identical content quality will be ranked differently if they differ in speed and technical parameters. The faster page wins.
In 2026 this effect is even more visible. Google Search Console clearly shows which pages have “Poor,” “Needs Improvement” or “Good” status in Core Web Vitals — and differentiates their organic visibility.
Core Web Vitals — What They Are and Why Google Measures Them
Core Web Vitals is a set of three metrics that Google has identified as key indicators of user experience. Each measures a different aspect of interaction with a page.
LCP (Largest Contentful Paint)
LCP measures the load time of the largest visible element on the page — typically the main image, a text block or a hero banner. It answers the question: when does the user see the actual page content?
| Rating | LCP time |
|---|---|
| Good | up to 2.5 s |
| Needs improvement | 2.5–4.0 s |
| Poor | above 4.0 s |
Most common causes of poor LCP: oversized images without compression and the right format (WebP/AVIF), slow server (high TTFB — Time to First Byte), JavaScript scripts blocking rendering when loaded in the page header, and lack of fetch priority for the hero image (fetchpriority="high").
CLS (Cumulative Layout Shift)
CLS measures page visual instability — how much elements “jump” on screen during loading. If you’ve ever clicked a button that moved down at the last moment and you ended up clicking an ad instead — that was a poor CLS experience.
| Rating | CLS value |
|---|---|
| Good | up to 0.1 |
| Needs improvement | 0.1–0.25 |
| Poor | above 0.25 |
Most common causes of poor CLS: images without defined width and height dimensions, web fonts loaded asynchronously (FOUT/FOIT), dynamically injected advertising and cookie banners, icon fonts replaced by images.
INP (Interaction to Next Paint)
INP is the newest of the metrics — it replaced FID (First Input Delay) in March 2024. It measures page responsiveness to user interactions: clicks, taps, form inputs. It answers the question: how quickly does the page respond when the user does something?
| Rating | INP time |
|---|---|
| Good | up to 200 ms |
| Needs improvement | 200–500 ms |
| Poor | above 500 ms |
The main cause of poor INP is large and poorly written JavaScript blocking the browser’s main thread. Heavy UI libraries, non-optimal React state management, too many CSS animations without will-change — these are typical sources of the problem.
Why WordPress Sites Load Slowly
WordPress powers approximately 43% of all websites in the world — an impressive statistic. But in the context of Core Web Vitals it’s less flattering: only 38% of WordPress sites pass the Core Web Vitals test according to HTTP Archive data (Web Almanac 2024).
Why such a discrepancy? It stems from several fundamental characteristics of the WordPress architecture:
1. PHP and dynamic server-side rendering. Every visit to a WordPress page triggers MySQL database queries, loads dozens of PHP files, processes plugins and only then sends HTML to the browser. This process takes time — typically 200–800 ms of TTFB alone, before the browser receives even a byte of content.
2. Plugin accumulation. The average WordPress site uses 20–30 plugins. Each adds its JS scripts and CSS stylesheets. Result: pages loading 15–30 JS files and a similar number of CSS files, often blocking rendering.
3. Page builders. Elementor, Divi, WPBakery — popular but devastating for performance. They generate redundant HTML code, add their own JS libraries (jQuery, custom animations), produce inline CSS for every element. A site built in Elementor can weigh 3–5 MB.
4. Shared hosting. Most WordPress sites run on cheap shared servers where CPU and RAM resources are shared between hundreds of other sites. A traffic peak at a neighbour’s site means your site slows down.
5. No default image optimisation. WordPress doesn’t automatically convert images to WebP format or lazy-load them optimally. Users upload photos from their phones (5–10 MB) and they land on the site at their original size.
These problems can be fixed — caching (WP Rocket, W3 Total Cache), CDN, better hosting (Kinsta, WP Engine) — but it requires continuous effort and money. And it still rarely matches the native performance of modern frameworks.
How Astro/Next.js Achieves a 67% CWV Pass Rate
Modern JavaScript frameworks like Astro and Next.js approach performance from fundamentally different foundations — and the results are measurable. HTTP Archive data shows that sites built on Astro achieve a 67% Core Web Vitals pass rate — nearly twice as good as WordPress.
Why?
Astro — “zero JavaScript by default” architecture. Astro generates static HTML at build time, not on every user request. The result is HTML files ready to be served immediately from a CDN without any server-side processing. JavaScript reaches the browser only where it’s absolutely necessary — in so-called “islands” (islands architecture). A typical Astro site sends 0–10 KB of JavaScript on first load.
Next.js — Server-Side Rendering and Static Generation with granular control. Next.js allows choosing for each page between SSG (static generation), SSR (server-side rendering) and ISR (incremental static regeneration). It automatically optimises images (<Image> component), prefetches links on hover, and splits JavaScript code into smaller bundles.
Shared advantages of both frameworks:
- Automatic image optimisation to WebP/AVIF with correct dimensions
- Code splitting — the user downloads only the JavaScript needed for the current page
- No jQuery or other heavy libraries in the default configuration
- Integration with CDN (Vercel Edge Network, Cloudflare) with minimal TTFB
- Tree-shaking — elimination of unused code from the final bundle
For a company choosing technology for a new website, this difference — 38% vs 67% CWV pass — is a concrete difference in Google visibility and sales conversions.
How to Measure Page Speed?
Before you start optimising, you need to know where you stand. Here are three essential tools:
Google PageSpeed Insights (pagespeed.web.dev) — a free Google tool that analyses a page against both lab data (simulation) and real user data (Chrome UX Report data). It gives a score of 0–100 for mobile and desktop and specific corrective recommendations. This is the first reference point for any analysis.
GTmetrix (gtmetrix.com) — a tool with a more detailed report, the ability to choose the testing server location and a results history. The free plan allows several tests per day, the paid plan enables recurring monitoring. A good choice for comparing a page version before and after optimisation.
Lighthouse in Chrome DevTools — a built-in Chrome browser tool (F12 → “Lighthouse” tab). It tests the page locally, without sending data to external servers. Ideal for testing pages in a development or staging environment before they go live.
Bonus — Google Search Console → the “Core Web Vitals” section shows real data from actual users of your site, grouped as “Good,” “Needs Improvement,” “Poor.” This is the only source that tells you how your site performs for your customers — not under lab conditions.
10 Ways to Speed Up Your Site
Regardless of technology, the following methods deliver measurable results:
-
Compress and convert images to WebP/AVIF — the single biggest gain. A 2 MB PNG image → 200 KB WebP at identical visual quality. Use tools like Squoosh, ShortPixel or the framework’s built-in optimisation.
-
Lazy load below-fold images — add
loading="lazy"to all images not visible on first load. Mark the main hero image withloading="eager"andfetchpriority="high". -
Minify and compress files — enable Gzip or Brotli on the server. Minify HTML, CSS and JavaScript (removes comments, whitespace, shortens variable names).
-
Implement a CDN — serve static assets from nodes geographically close to the user. Cloudflare (free plan is sufficient for most SMEs), Vercel Edge Network, AWS CloudFront.
-
Eliminate render-blocking JS/CSS — load scripts with the
deferorasyncattribute. Remove or conditionally load unused styles and scripts. -
Browser and server caching — set
Cache-Controlheaders for static assets (images, fonts, icons) to a minimum of 1 year. Implement server-side caching (Redis, Varnish) for dynamic pages. -
Preconnect to external resources — if you use Google Fonts, Typekit fonts or analytics scripts, add
<link rel="preconnect">in<head>. Eliminates DNS lookup and TLS negotiation delay. -
Optimise fonts — download fonts locally instead of loading from Google Fonts on every request. Use
font-display: swapto prevent invisible text during loading. -
Reduce third-party JavaScript — every social media widget, live chat, advertising pixel and tracker adds tens of kilobytes of JS. Load them conditionally (after user interaction or after
DOMContentLoaded). -
Upgrade hosting or change technology — if you’re doing everything above and still have poor results, the problem may lie deeper: in the site’s architecture or server quality. Migrating from WordPress to Astro or Next.js on good hosting (Vercel, Cloudflare Pages) often delivers a jump from 45 to 90 PageSpeed points without any additional optimisation.
How Much Does Speed Optimisation Cost?
Cost depends on the current state of the site and the chosen improvement method:
| Scope of optimisation | Lead time | Approximate cost |
|---|---|---|
| Performance audit + report | 1–2 days | 500–2,000 PLN |
| Image optimisation and caching | 1–3 days | 1,000–3,000 PLN |
| Comprehensive WordPress optimisation | 3–7 days | 3,000–8,000 PLN |
| WordPress → Astro/Next.js migration | 2–6 weeks | 8,000–30,000 PLN |
| New site in Astro/Next.js (from scratch) | 4–10 weeks | 8,000–40,000 PLN |
For existing WordPress sites with moderate issues (PageSpeed score 40–60), comprehensive optimisation at 3,000–8,000 PLN is a sensible step before considering migration. It can achieve a score of 70–85, which significantly improves CWV and conversions.
If the site is old, built on a heavy page builder or has a very low score (below 30–40 points) — optimisation is patching a leaking vessel. A new site in Astro from 8,000–15,000 PLN with 90+ PageSpeed often pays back within 6–12 months through higher conversions and better organic traffic.
FAQ — Website Load Speed
Does page speed really affect Google position?
Yes — officially since 2021 (Page Experience Update). Core Web Vitals are a ranking factor. Pages with a “Good” rating have an advantage over pages with a “Poor” rating for identical content quality. The effect is particularly visible in highly competitive niches.
What PageSpeed Insights score should I achieve?
For mobile, above 70 is the minimum, above 85 is a good result. 90+ (green) is the goal for sites where speed has critical business significance. Desktop is typically 10–20 points higher than mobile — optimise primarily for mobile.
My site loads in 2 seconds in Chrome on my laptop — is that a good result?
Not necessarily. Your laptop and connection are not a representative user. Google simulates a slower device (equivalent to a mid-range mobile phone) and a 4G network in its lab tests. Always test through PageSpeed Insights and look at the “field data” (real users) in Search Console.
Do cache plugins fix a slow WordPress site?
They help, but rarely solve the problem completely. WP Rocket, W3 Total Cache or LiteSpeed Cache can improve the score by 10–30 points. But if the problem lies in a poor architecture (heavy page builder, 30+ plugins, cheap hosting) — caching is a plaster on a broken leg.
How long does optimisation take and when will I see results in Google?
Technical changes are visible immediately after implementation in measurement tools. Google re-indexes the page typically within 1–4 weeks. Effects on organic positions may be visible after 4–12 weeks — SEO requires patience, but performance changes work faster than, for example, link building.
Summary — Website Load Speed
Website load speed in 2026 is one of the most important business factors for any company with an online presence. -4.42% conversion per second of delay is not a theory — it’s money your potential clients are leaving with your competitors.
Core Web Vitals — LCP, CLS and INP — are concrete, measurable indicators that Google uses to evaluate your page. Technologies like Astro and Next.js achieve a 67% pass rate, while WordPress sits at 38%. This is not coincidental — it’s the result of fundamental architectural differences.
The good news: it can be fixed. Image compression, CDN, caching, eliminating unnecessary JavaScript — each of these steps delivers measurable results. And if you want a lasting solution with a guaranteed high Core Web Vitals score — a new site in modern technology is an investment that pays back in concrete revenue.
Want to know how fast your site is and exactly what’s slowing it down?
Check our website services → /strony-www/ and see how we build sites with 90+ PageSpeed scores that rank high in Google and convert visitors into customers.