Ortica
Let’s talk ↗

· 8 min read

Core Web Vitals: what they measure and why your site is slow

Core Web Vitals are three Google metrics for perceived speed: LCP (loading), INP (responsiveness) and CLS (visual stability). You can read them in PageSpeed Insights and Search Console; on WordPress the problems almost always come from hosting, caching, images and plugins.

In short: Core Web Vitals are three metrics Google uses to judge website speed from the point of view of the people using it: LCP (how long the main content takes to appear), INP (how quickly the page responds to a tap or a click) and CLS (how much it shifts around while loading). You can read them for free in PageSpeed Insights and Search Console. On a slow WordPress site the causes are almost always the same: hosting, caching, images, too many plugins and third-party scripts.

What Core Web Vitals are

Core Web Vitals are the set of metrics Google considers essential to the experience of a page. They don’t measure “how fast the server is”, but what a person sees and feels: when the thing they were looking for appears, whether the button responds straight away, whether the page stays still while they read it.

Since 12 March 2024 the three metrics have been LCP, INP and CLS. INP replaced FID, which only measured the first click and which almost every site passed without effort.

LCP, INP and CLS: what they measure and the thresholds

Metric

What it measures

Good

Needs improvement

Poor

LCP

when the main content appears

up to 2.5 s

up to 4 s

over 4 s

INP

how quickly the page responds to taps and clicks

up to 200 ms

up to 500 ms

over 500 ms

CLS

how much the page shifts while loading

up to 0.1

up to 0.25

over 0.25

  • LCP (Largest Contentful Paint) is the moment the largest element in the first screen appears: usually the image at the top or the headline.
  • INP (Interaction to Next Paint) is the time between an action (opening the menu, picking a size, adding to the cart) and the response on screen. What counts is one of the slowest interactions of the visit, not the average.
  • CLS (Cumulative Layout Shift) adds up the sudden shifts: text that drops down because a banner arrives, the button that jumps away just as you tap it.

The thresholds apply at the 75th percentile: a page passes if at least three visits out of four are under the threshold, and mobile and desktop are counted separately. Mobile is almost always the worst case, and that’s where it’s worth starting.

Where to read them: PageSpeed Insights and Search Console

  • [PageSpeed Insights](https://pagespeed.web.dev/) measures one page at a time. At the top it shows real visitor data from the last 28 days, if the site has enough of it; below is the Lighthouse test with the 0 to 100 score and the list of problems.
  • Google Search Console, in the Core Web Vitals report, shows the whole site: it groups similar pages and tells you which are good, need improvement or poor, again with real data from the last 28 days.
  • Chrome’s developer tools (Lighthouse and the Performance panel) help you understand what’s slowing down a specific page while you work on it.

Lab data and field data: why the numbers don’t match

This is the most common confusion. The PageSpeed score comes from a lab test: Lighthouse loads the page once, simulating a mid-range phone on a slow connection. It’s useful for finding problems, but it’s a simulation, and it changes a little with every run.

Field data comes from the Chrome User Experience Report: anonymous measurements from real visitors using Chrome, collected over the last 28 days. This is what Google uses, and this is what counts.

Three practical consequences:

  • a site can score 60 and have good Core Web Vitals, or the other way round;
  • the lab test doesn’t measure INP, because nobody touches the page: instead it looks at how long the browser is blocked by JavaScript (Total Blocking Time);
  • a site with few visits may have no field data at all: in that case you work from the tests and from measurements you take yourself.

After a change, field data updates gradually, because the window is 28 days: the improvement in Search Console shows up after a few weeks.

Do Core Web Vitals affect rankings?

Yes, but less than you read online. Google uses them among its page experience signals and recommends good values, but relevant content comes first: a slow page that answers the search better can rank above a fast one that doesn’t. Speed matters most to visitors: a page that takes too long to load loses people before they even read it, and on an online shop it loses orders.

That’s why we aim for “good” values on field data, not for 100 on PageSpeed.

Why a WordPress site is slow: the usual causes

When we take on a slow WordPress site, the causes are almost always on this list, often more than one at a time:

  • Underpowered hosting or old PHP: the server takes too long to respond, and everything else starts late. It weighs on LCP.
  • No page caching: every visit makes WordPress rebuild the page from scratch, with dozens of database queries.
  • Heavy images: photos of several megabytes, no modern formats (WebP, AVIF), the main image held back by lazy loading instead of loading first.
  • Page builders and multipurpose themes that load the CSS and JavaScript for every feature on every page, even where they’re not needed.
  • Too many plugins and third-party scripts: chat widgets, ad pixels, maps, videos, cookie banners. They’re the main cause of a high INP, especially on phones.
  • Images, banners and fonts with no space reserved: when they arrive they push the rest of the page down. A high CLS is almost always this.
  • A bloated database: revisions, data left by uninstalled plugins, ecommerce sessions. On WooCommerce with lots of products, search and filters are often the bottleneck.

How to improve them

You don’t start with a plugin that “speeds everything up”: you start with measurements. This is our method:

  • Measure: field data where it exists, tests on the most visited pages and on the ones that sell (home, category, product, cart).
  • Find the cause of each metric in the red: for LCP it’s usually the server, caching and the main image; for INP, JavaScript; for CLS, missing reserved space.
  • Work in order of impact, testing changes on a copy of the site (staging) before putting them live.
  • Measure again, and keep an eye on the data over the following weeks, because every new plugin or script can undo what was gained.

You can often gain a lot without rebuilding the site: proper hosting, caching, images and fewer scripts. A rebuild makes sense when the problem is the theme or the page builder itself. On Shopify stores the server can’t be touched, but themes and apps weigh just as much: we talk about it in the comparison between Shopify and WooCommerce.

What to ask anyone who promises you a fast site

  • Will you show me the before and after data, on field data and not just the PageSpeed score?
  • Which pages do you measure: just the home page, or products, categories and the cart too?
  • Do you test changes on a copy of the site first?
  • What happens to speed when I update plugins or add a script?
  • Be wary of anyone who guarantees a specific score: the lab score changes with every test, and what matters is how the site works for the people using it.

How we work is on the page about website performance. If the problem is the server, we handle it under server and DevOps; to keep the site fast over time there’s WordPress maintenance.

Ortica is two people, Riccardo and Davide, working on the web since 2016: the person who answers is the person doing the work, with no middlemen.

FAQ

The questions
people ask us.

What are Core Web Vitals, in plain terms?

They’re three Google metrics for how a page feels to use: how long the main content takes to appear (LCP), how quickly it responds when you tap something (INP) and how much it shifts while loading (CLS). They’re measured on real visits, separately for mobile and desktop.

Why does PageSpeed give me a different score every time?

Because the score comes from a simulation that depends on the network and the server at that moment. The field data at the top of the PageSpeed page and in Search Console is more stable: it’s real visits from the last 28 days.

Search Console says there isn’t enough data: what does that mean?

That the site doesn’t yet have enough Chrome visits to calculate real values. It isn’t an error: in that case you work from lab tests and from measurements taken by hand on the main pages.

Is a caching plugin enough to fix a slow WordPress site?

Sometimes it helps a lot, especially with LCP. But it doesn’t fix heavy images, third-party scripts or underpowered hosting, and if it’s badly configured it can break the cart and forms. Measure first, then decide what to do.

How long until the improvement shows in Search Console?

Field data covers the last 28 days, so the change shows up gradually in the weeks after the work. In PageSpeed tests, on the other hand, you see it straight away.

Notes

Want to talk it over?
Write to us.

Tell us what you need: you’ll hear back from us, with a written quote if it makes sense.

Let’s talk