Imagine walking into a physical shop where the door sticks, the shelves wobble on their own, and the assistant takes a full minute to react when you speak to them. You'd probably walk out. Google measures the digital equivalent of that experience with three metrics it calls Core Web Vitals, and uses them to decide whether your site deserves to rank near the top.
The three that matter
LCP (Largest Contentful Paint): how long the biggest element on the screen (usually an image or a headline) takes to appear. Under 2.5 seconds is good; above that, it starts to show.
INP (Interaction to Next Paint): how long the page takes to react when someone clicks or taps something. Past 200 milliseconds, the feeling of "sluggishness" is already there.
CLS (Cumulative Layout Shift): how much the content "jumps" while loading (that moment when a button shifts right before you click and you end up tapping something else by mistake. The closer to 0, the better.
Why would a search engine care about this?
Because a slow or unstable site is frustrating, and a frustrated person goes back and picks another result. Google doesn't want to send traffic to sites that are going to disappoint, so it steadily) not dramatically — penalises anyone who falls short of these minimums.
The usual suspects
Uncompressed images, too many plugins or third-party scripts loading at once (a classic on a neglected WordPress site), fonts that cause text to jump around while loading, and shared hosting that can't keep up during peak traffic hours.
Free tools like PageSpeed Insights give you a score and, more importantly, a concrete list of what's failing. That's the real starting point: not the score itself, but the exact resource that needs fixing.
Mobile first, always
Google evaluates these metrics mainly using the mobile version of your site, not desktop. A website that flies on an office computer can be quite tight on a mid-range phone with average mobile data, which is how most of your real visitors browse. Any performance review should simulate those conditions, not the ideal ones.
Is it worth fixing if "sales are already fine"?
This is one of those improvements you don't notice until you measure it: many websites lose a real share of visits to people leaving before the page finishes loading, without anyone realising because those visits simply leave no trace in the usual statistics. Fixing Core Web Vitals isn't just an exercise to keep Google happy: it improves the experience and, with it, conversion.
Frequently asked questions
How much does this weigh in rankings compared to content?
Less than content and authority, but it can be an important tiebreaker between sites competing for the same searches. And above all, it directly affects how many visits turn into customers, beyond ranking itself.
Can this be fixed without touching code?
Partly: compressing images, enabling caching and choosing good hosting covers a good part of it. But INP and CLS issues usually require real technical adjustments to the template or theme.
How often should these metrics be reviewed?
At least whenever you make a big change to the site (new template, new plugin, redesign), and periodically even if nothing changes, since content growth itself can gradually slow the site down.