Site speed and Core Web Vitals optimization

A customer on a phone waits three seconds, then goes back to the search results and picks a competitor. We measure how fast your site really is for the people using it, find what slows it down and fix it.

01

What Google measures

Core Web Vitals are three numbers Google collects through Chrome from the people who actually open your site:

  • LCP: how long until the main content appears, usually the large image or heading at the top. Good is under 2.5 seconds.
  • INP: how long the page takes to respond after a tap or keystroke. Good is under 200 milliseconds.
  • CLS: how much the page shifts while loading, for example when a button jumps away from your finger because a banner loaded above it.

Because it's measured on real visitors, the PageSpeed score and the Search Console report often don't match. The second one is the one that counts.

02

Why speed matters

Speed is part of technical SEO and Google does take it into account, but that isn't the main reason to work on it. The main reason is sales: on a slow page fewer people reach the form or the cart, especially on phones and weak mobile connections.

For online stores and ad landing pages the difference shows up in money. When you pay for a click and the visitor leaves before the page loads, that click is wasted.

03

What usually slows a site down

On the sites we check, the causes keep repeating:

  • a hero image of several megabytes in a format not meant for the web
  • fonts loaded from three different servers
  • a dozen analytics, chat and ad scripts that run before the content
  • a WordPress theme or page builder that loads everything it has on every page
  • a server with no caching that builds every page from the database for every visitor
  • banners and cookie notices that appear late and push the whole page down
04

What we do

We convert images to WebP or AVIF and serve a size that fits the screen. Critical CSS goes inline and the rest loads later. Scripts that aren't needed straight away are deferred, and the ones nobody uses are removed. We set up server caching and a CDN, and reserve space for elements that load late so the page doesn't shift.

For example, the website we built for the EC Pleasure aesthetic center scored 100 for performance in PageSpeed Insights on desktop (99 on mobile), with a Largest Contentful Paint of 0.5 s on desktop.

Most of this needs no redesign and no change in how the site looks. When the problem is the theme or the platform itself, we say so and estimate whether fixing it costs more than building a new website.

05

Speed degrades gradually

A site that's fast after optimization doesn't stay that way by itself. Marketing adds a new script, someone uploads a photo straight from the camera, a plugin gets updated. A few months later the score is back where it started.

So after the work is done we keep an eye on the Core Web Vitals report and let you know when something gets worse, before your customers notice.

FAQ

Questions about site speed

Less than content and relevance. A fast site won't outrank a better page just because it's faster. But when two pages are similar, speed can tip the balance, and it always affects sales.

Usually, yes. Images, caching, fonts and scripts can be fixed without changing the design. A redesign only makes sense when the theme or platform is so heavy that fixing it costs more than starting again.

For a brochure site, usually a few days. For a larger store or web application, two to three weeks, because each change has to be checked on several page types. The result shows up in Search Console only after 28 days, because Google reports an average for that period.

A PageSpeed test measures a single load on a simulated phone. Search Console shows what real visitors experienced over 28 days, on their own devices and connections. Google goes by the second number.

Is your site slow on mobile?

Send us the URL. We'll tell you what's slowing it down and how much work it takes to fix.

Request a free estimate →