A client forwards us a PageSpeed report every few months with a red number circled and the question “is this why we’re not ranking?” Usually the answer is no, or not mostly. But sometimes the report is pointing at something that’s genuinely costing them customers, and Core Web Vitals is how you tell the difference.
Core Web Vitals are three measurements Google uses to judge how a page feels to use: how quickly it loads, how quickly it responds, and whether it jumps around. Each has a pass mark, and each maps to something you’d notice in a shop.
The three measurements
LCP: how long until the main thing appears
Largest Contentful Paint measures how long it takes for the biggest thing on screen to show up. On most small business sites that’s the hero image or the main heading.
Think of walking into a shop. LCP is how long you stand in the doorway before the lights come on and you can see what’s for sale. You can put up with a second or two. Five seconds and you’re wondering whether they’re open.
The usual causes on small sites:
- A huge hero image, often a full-size photo straight from a camera or phone.
- Slow hosting, so the server takes ages to start sending anything.
- The hero image being lazy-loaded, which tells the browser to wait before fetching the one image it needs first.
- Heavy sliders or background videos at the top of the page.
- Custom fonts that hold up the text.
INP: how long until a tap does something
Interaction to Next Paint measures the delay between someone tapping or clicking and the page visibly responding. It replaced an older measure called First Input Delay in 2024.
This is the shop assistant who looks up when you ask a question. A quick nod is fine. A long pause while they finish something else makes you wonder whether they heard you. On a website, that pause is the menu that doesn’t open straight away, or the form button that seems dead for a moment.
The usual causes:
- Too much JavaScript, often from a pile of plugins each loading their own scripts.
- Third-party extras like chat widgets, review carousels, booking embeds and tracking tags.
- Big page-builder layouts with many nested elements, which take longer for a phone to process.
CLS: does the page jump around
Cumulative Layout Shift measures how much the content moves while you’re trying to read or tap it. You go to tap “Call now”, a banner loads above it, everything drops down, and you tap “Directions” instead.
It’s the shop where someone keeps rearranging the shelves while you’re reaching for something.
The usual causes:
- Images without set dimensions, so the browser doesn’t reserve space for them.
- Cookie banners, promo bars or popups that push content down instead of sitting over it.
- Embeds (maps, videos, social feeds) that load late and expand.
- Web fonts that swap in and change the size of the text.
The pass marks
These are Google’s current thresholds, as published on web.dev at the time of writing (late 2026).
| Measure | Good | Needs improvement | Poor |
|---|---|---|---|
| LCP (loading) | 2.5 seconds or less | 2.5 to 4 seconds | Over 4 seconds |
| INP (responsiveness) | 200 milliseconds or less | 200 to 500 milliseconds | Over 500 milliseconds |
| CLS (stability) | 0.1 or less | 0.1 to 0.25 | Over 0.25 |
Two details matter. First, they’re measured at the 75th percentile of real visits, so to pass, three out of four visitors need a good experience, not just you on your office wifi. Second, mobile and desktop are measured separately, and mobile is nearly always worse.
Where to find your real numbers
Run a page through PageSpeed Insights. The top section, which describes what real users are experiencing, comes from Chrome users who visited your site. That’s the data that counts. The score out of 100 underneath comes from a one-off simulated test. It’s useful for finding problems, but it’s not what Google uses to assess your site, and it swings around from run to run.
Many small sites don’t have enough visitors for the real-user section to appear. That’s normal. In that case, use the simulated test as a guide, and don’t obsess over the score.
Search Console has a Core Web Vitals report too, under Experience, which groups your pages by status using the same real-user data.
How much this affects rankings
Less than the industry noise suggests. Google’s own page on Core Web Vitals says good scores align with what its ranking systems “seek to reward”, and it recommends getting them right, but it’s one signal among many. A fast page with thin content won’t outrank a slower page that answers the search properly.
Where Core Web Vitals really bite is with your customers. Someone on a patchy mobile connection in a regional town, waiting for a giant hero image, will leave before they see your phone number. That’s a lost job whether or not Google noticed.
So our position is this. Get out of the Poor band on all three, because that’s usually a sign of a real problem. Aim for Good on your homepage and key service pages. Don’t spend thousands chasing a perfect 100.
What usually fixes each one
On the WordPress and Elementor sites we build and look after, these are the fixes that make the biggest difference, roughly in the order we try them.
For LCP: resize and compress the hero image (WebP, usually no wider than about 2000 pixels), exclude it from lazy loading, and make sure the page is cached. If the server is slow to respond before anything loads, it’s a hosting problem, and our guide to hosting in plain English explains why cheap hosting is a false economy. Caching matters a lot here too, and our WordPress caching explainer covers what it does.
For INP: audit your plugins and third-party scripts. Remove what you’re not using, load chat and booking widgets only on the pages that need them, and delay non-essential scripts until after the page loads. Simplify heavy sections built from many nested containers.
For CLS: make sure every image has width and height set (WordPress and Elementor normally do this, but custom code and embeds often don’t), give embeds a fixed-height space to load into, and make cookie and promo banners overlay the page rather than push it down.
Before you pay anyone
Run your homepage and your most important service page through PageSpeed Insights on mobile. Note whether you have real-user data, and which of the three measures is worst. That one number tells you where to look first, and stops anyone selling you a generic “speed optimisation package” that fixes the wrong thing.
Want a second opinion on what that number means for your site? Our SEO team can tell you which fixes are worth paying for and which aren’t, and when a slow site is past patching, a rebuild on decent hosting is often the cheaper fix over a few years.