Back in 2021 we wrote a short explainer on Core Web Vitals when Google first turned them into a ranking signal. Five years on, that post is due an update because Google quietly swapped one of the three metrics out, and a fair bit of the advice still floating around the internet is now out of date.

So here’s where things actually stand in 2026, minus the scaremongering.

A quick refresher: what are Core Web Vitals?

Core Web Vitals are Google’s attempt to put a number on how a page feels to a real person using it. Not how it looks in a design tool, and not how fast it loads on a developer’s fast laptop — how it behaves for the actual human on their phone, on real-world internet. There are three of them, and they cover three different frustrations: waiting for a page to show up, tapping something and having nothing happen, and the layout jumping around while you’re trying to read it.

The one thing that’s changed: goodbye FID, hello INP

If you read anything older than 2024 (including our own last post), it’ll mention a metric called FID – First Input Delay. That one’s gone. On 12 March 2024 Google replaced it with INP – Interaction to Next Paint, and INP is a tougher, fairer measure.

The old FID only looked at the very first tap or click on a page, and only measured how long the browser took to start reacting. INP watches every interaction across the whole visit – every tap, click and keypress – and measures the full round trip until the screen actually updates. In plain terms: it’s much harder to game, and it lines up far better with the “why is this page so sluggish?” feeling everyone recognises. So if your site felt fine under FID but suddenly looks worse, that’s usually why.

The three that matter in 2026

Here’s the current set, what each one is really measuring, and the score Google considers “good”:

  • LCP – Largest Contentful Paint (loading): how quickly the main thing on the page – usually your hero image or headline – actually shows up. Good is 2.5 seconds or less.
  • INP – Interaction to Next Paint (responsiveness): when someone taps or clicks, how quickly the page reacts. Good is 200 milliseconds or less.
  • CLS – Cumulative Layout Shift (visual stability): how much the page jumps around as it loads. Think of a button sliding down just as you go to tap it. Good is 0.1 or less.

One detail worth knowing, because it trips people up: these scores come from real visitors to your site over the previous 28 days, not from a one-off speed test. And Google grades you at the 75th percentile – meaning three out of four of your visitors need to hit “good”, not just the average one. Mobile and desktop are scored separately too, and mobile is almost always the harder of the two. So a page can look great in a lab test and still not pass, if enough of your real visitors are on slower phones.

Does this actually affect my Google ranking?

Google’s own position hasn’t really changed: Core Web Vitals are part of a broader “page experience” picture, and they’re a secondary signal. Relevant, genuinely useful content is still what wins rankings. A fast page won’t outrank a better answer, and a brilliant article on a slightly slow page will still do just fine. Where Vitals tend to matter is as a tie-breaker – when two pages are otherwise neck and neck, the better experience can edge ahead.

The stronger reason to care isn’t Google at all – it’s the people you actually want to keep on your site. A page that’s slow to load, sluggish to tap, or jumps around while someone’s reading is a page people leave. That shows up in your bounce rate and your enquiries long before it shows up in a ranking report.

And it’s not a niche problem. On the latest industry-wide numbers, only around half of mobile sites pass all three Vitals – so if yours does, you’re already ahead of a lot of your competitors. Loading speed (LCP) is the one most sites fall down on.

So how do we improve them?

Every site’s different, but the levers are fairly consistent. Here’s the short version of what we look at:

For loading (LCP)

Get the biggest image on the screen loading first, not last – and never lazy-load your hero image. Serve images in modern formats (like WebP or AVIF) at the right size rather than shrinking a huge file in the browser. And make sure your hosting and caching aren’t leaving people waiting before anything even starts. Cheap, oversold hosting is a common culprit here.

For responsiveness (INP)

This one is almost always a JavaScript problem, and usually it’s the pile of third-party scripts a site accumulates over time — chat widgets, tag managers, tracking pixels, that one plugin nobody remembers installing. Trimming what you don’t need, and making sure the essential stuff doesn’t hog the browser, does most of the work.

For stability (CLS)

Mostly this comes down to reserving space in advance – giving images, videos, ads and embeds a defined size so nothing shoves the rest of the page around when it pops in. Fonts are a sneaky offender too, if they swap in and reflow the text.

What’s coming next?

Nothing you need to act on today – but worth knowing so you’re not caught out. The three metrics and their targets are stable for 2026; despite what a few blogs claim, Google hasn’t made the thresholds stricter and hasn’t added a fourth metric. The main thing on the horizon is better measurement of app-style sites (single-page apps, where the page never fully reloads), which the current Vitals don’t capture well. It’s still experimental and doesn’t affect anyone’s ranking yet. We keep an eye on it so our clients don’t have to.

The short version

Three metrics — loading, responsiveness, stability. FID is out, INP is in. The targets are 2.5 seconds, 200 milliseconds, and 0.1, measured on your real visitors. It’s a real thing worth getting right, mostly for your visitors and a little for Google – but it’s not something to lose sleep over, and it’s definitely not worth panic-buying an “emergency SEO fix” for.

If you’re not sure how your site’s doing, we’re happy to take a look and give you a straight answer . Get in touch for a no-obligation chat.