What Are Core Web Vitals?

Core Web Vitals are three measures of how solid a page feels in real use. One watches loading, one watches responsiveness, one watches visual stability. Google uses the trio as an experience signal. It is not a magic ranking key. An irrelevant page can score “good” on all three and still fail to answer the query. A bad trio is still an unnecessary obstacle in front of good content.

Names can change, and thresholds are updated in Google’s own documentation. This article speaks with the names in use today and the “good” limits Google has published. Memorising the limits is not the job. Finding which file on which page breaks the limit is the job. How speed relates to rankings is in website speed. Here, the measure itself is the subject.

LCP: when the main content appears

LCP is largest contentful paint. In practice it is the moment the main image or headline on the first screen becomes visible. The good limit Google publishes is 2.5 seconds or under. Above 4 seconds counts as poor. The usual culprit is the hero: a photo that looks small on screen and is huge as a file. A late font can hold the headline back too.

The fix is not a score trick. Crop the image, serve the right size, delay whatever sits below the first screen. “Show an empty box first and save the score” still makes a person wait. The measure is when useful content appears, not when a box does.

INP: the answer to a tap

INP is the delay after an interaction. You press a button, nothing happens, then it does. The good limit Google publishes is 200 milliseconds or under. Above 500 milliseconds counts as poor. It replaced an older metric, FID. It looks at interaction delay across the page, not only the first tap.

The usual culprit is heavy JavaScript. Three chat tools, a heatmap, a needless animation and a swollen plugin put the tap in a queue. The user thinks the form was sent, taps again, and you get a double submission. Speed here is a behaviour problem, not a picture problem. On WordPress or custom software, removing a script you do not use gains more INP than adding another library.

CLS: the page jumping

CLS is cumulative layout shift. The line you started reading gets pushed down by a late image or an ad, and you tap the wrong place. The good limit Google publishes is 0.1 or under. Above 0.25 counts as poor. The culprits are an image with no dimensions, a banner injected late, and text that shoves the line when the font swaps.

Giving images a width and height removes a large share of the shift. Pasting an ad into the page without reserving space breaks both CLS and trust. “Hide the ad to fix the score” is a shortcut that breaks the business. Reserve the space, measure the shift, then seat the ad in that space.

• LCP: the main content of the first screen visible at 2.5 seconds or under.
• INP: a tap answered at 200 milliseconds or under.
• CLS: shift kept at 0.1 or under, so the button does not move.

The lab and the field are not the same

Tools like Lighthouse simulate one controlled visit. The experience report in Search Console looks at the spread of real users. They can disagree. The lab can look bad while the field looks good, or the reverse. The verdict, when you can have it, comes from the field. If there is no field data — a new site, or thin traffic — the lab is only a hint. Do not read the hint as “this is our ranking”.

Templates differ too. The homepage can be clean while the product page is heavy. Rounding the report up to “the site” hides the template that is actually broken. Start with the template that gets the most traffic. On a technical SEO list this trio comes after redirects and indexation. The vitals of a page that is never crawled are a score nobody sees. The order is set together with the technical SEO checklist.

Do not break the page while you improve it

Yuz Medya does not “fix” vitals by breaking the form or the measurement tag. The image gets smaller, the page stops jumping, and one press of the button sends once. The before-and-after trace says which template it was taken on. An “all green” screen is not handed over until the contact form has been tried on a phone. The three measures describe a visitor’s patience. Patience is earned. Then the content and the offer do their work inside it. The measure does not replace the offer.

Frequently asked questions

If all three are not green, is the ranking over?

No. The trio is one part of experience. If the content is irrelevant, a green score does not make you relevant. If all three are poor, you can reasonably expect a minus against otherwise equal competitors. Fix them. Do not turn that into a ranking guarantee on its own.

Do the thresholds change?

Google can update the documentation. The 2.5 seconds, 200 milliseconds and 0.1 in this article are the published “good” limits at the time of writing. Check the current document before you decide. Memory does not replace measurement.

Is fixing the homepage enough?

No. Measure the template that makes the money: the service, the product, the booking. A clean homepage and a jumping form page tire the visitor at the exact moment of decision. Pick the template by traffic and by the business goal.