Core Web Vitals in Plain Words: What LCP, INP and CLS Measure and Which One to Fix First

A 3D yellow figure presses a button on a large screen, while a drawn man holds a stopwatch nearby. The scene includes plants, windows, and cityscape outlines in the background.

Three scary acronyms, one simple idea. Core Web Vitals are Google’s way of scoring how your page feels to a real person, not to a machine.

They measure three things a visitor notices without thinking: did the main thing load fast enough (that’s LCP), did the page respond when they tapped or clicked (INP), and did it hold still while it loaded instead of jumping around (CLS). Each one has a “good” score to aim for, and Google checks them using data from actual visitors, not a lab test.

Which one should you fix first? The one your real visitors are failing, and for most small business websites that turns out to be the loading score or the jumping score, not the one owners tend to worry about. The scary sounding one is usually fine.

What Core Web Vitals actually measure

Forget the acronyms for a second. When someone lands on your page, their body asks three quick questions:

  • Did it load? Did the big thing I came to see, the headline, the photo, the main text, show up quickly, or did I stare at a blank screen?
  • Did it respond? When I tapped a button or a menu, did something happen right away, or did the page freeze for a beat?
  • Did it hold still? Did the page settle down, or did stuff keep shifting so I tapped the wrong thing?

Core Web Vitals put a number on each of those three feelings. That’s the whole idea. Here are the three scores and what “good” looks like:

  • LCP (Largest Contentful Paint): loading. Good is under 2.5 seconds.
  • INP (Interaction to Next Paint): responding. Good is under 200 milliseconds, which is a fifth of a second.
  • CLS (Cumulative Layout Shift): holding still. Good is under 0.1, a small number where lower is steadier.

You don’t need to memorize the numbers. You need to know which of the three questions your site is failing, and why.


LCP: did the main thing load fast enough?

LCP times how long it takes for the largest piece of content on the screen to show up. On most pages that’s the hero image at the top or the big headline. Google wants that to happen in under 2.5 seconds. Over 4 seconds is rated poor.

For small business sites, the usual culprit is one thing: a huge image. Owners upload a photo straight off their phone, which can be 5 or 6 megabytes, and the browser has to drag that whole file down before the page looks ready. A slow or cheap web host makes it worse.

The plain fix is to shrink your images before they go up. Resize them to the width they actually display at, save them in a modern format like WebP, and let a caching plugin or your host do the rest. That one habit fixes more slow pages than any other single change.


INP: did the page respond when you tapped?

INP measures the lag between a visitor doing something, tapping a button, opening a menu, and the page actually reacting. Good is under 200 milliseconds. Anything a person would call “laggy” lives above that.

This is the newest of the three. In March 2024 it replaced an older metric called First Input Delay, or FID. The old one only timed the very first tap. INP watches every interaction during the visit and reports the slow ones, so it’s a much truer picture of whether a page feels snappy or stuck.

When INP is bad, the cause is almost always too much code running in the background, usually a pile of plugins or a bloated theme fighting for the browser’s attention. Most simple small business sites, a few pages and a contact form, pass INP without trying. If yours fails, that’s a sign you’ve stacked on more plugins and trackers than the site can carry, and thinning them out is the fix.


CLS: did the page hold still while it loaded?

CLS scores how much your page jumps around as it loads. You’ve felt this: you go to tap a link, an image finishes loading above it, everything shifts down, and you tap an ad instead. Good is under 0.1, where lower means steadier.

The jumping happens when the browser doesn’t know how much room something needs until it arrives. The main offenders are:

  • Images and videos with no width and height set, so the page reflows when they load.
  • Ads or embedded widgets that push content down when they appear.
  • Web fonts that swap in and shove the text around.

The fix is to reserve the space in advance. Give every image its dimensions so the browser holds the spot, and leave a fixed box for anything that loads late. A decent modern theme does most of this for you, which is one more reason a clean theme beats a heavily customized one.


Which one to fix first

Don’t try to fix all three at once. Fix the one your real visitors are actually failing, and only that one, until it passes.

To find out which, look at your field data (more on where in a second). Then use this order:

  • If LCP is failing, start there. It’s the most common failure on small sites and usually the easiest win, because it’s almost always an image problem you can fix in an afternoon.
  • If CLS is failing, that’s next. Set image dimensions and pin down anything that loads late.
  • If INP is failing, look at your plugins. It’s the least common problem for a simple site, and when it shows up it’s usually a sign you’re running too much.

Notice the pattern. The two that trip up most small business sites are the loading score and the jumping score, both usually caused by images and layout, not by code. The responsiveness score that sounds the most technical is the one you’ll probably touch last. If you want the bigger picture of how these site numbers ripple into every channel, we cover that in our guide to Core Web Vitals and page speed.


Where to see your real numbers

There are two kinds of Core Web Vitals data, and mixing them up is where people get confused.

Lab data is a single test run in a controlled setting. You get it from Google’s free PageSpeed Insights tool: paste your URL and it grades the page. It’s useful for spotting problems, but it’s one run on one simulated connection, so it can look fine while real visitors struggle.

Field data is the real thing: actual scores from real people who visited your site over the last 28 days. This is what Google grades you on. It shows up in PageSpeed Insights (at the top, when your site has enough traffic) and in the Core Web Vitals report inside Google Search Console.

One detail matters here. Google grades you at the 75th percentile, which means 3 out of 4 of your visitors need a good experience for the page to pass.

So the score leans toward your slower visitors, not your best phone and not your average one. That’s why a page can feel fast on your own device and still fail: some of your visitors are on older phones and slower connections, and they count.


How much this actually matters for ranking

Core Web Vitals are a real ranking factor, but a moderate one. They work like a tiebreaker. When two pages are close on the stuff that matters most, useful content that answers the question, they help the faster, steadier one edge ahead. They will not push a thin page above a great one.

So keep it in proportion. Good content and a page that answers the search come first. But once those are in place, a site that loads fast, responds quickly, and holds still keeps more visitors, converts more of them, and gives Google one more reason to prefer you.

That’s true whether or not it ever moves your ranking, which is the real argument for doing it. For how this fits with everything else that gets you found, see our overview of how search marketing actually works for a small business.

If you only do one thing after reading this, open PageSpeed Insights, paste your homepage, and look at the field data at the top. It tells you in ten seconds which of the three to fix first, and now you know what each one means.


Frequently asked questions

What are Core Web Vitals in simple terms?

They’re three scores for how your page feels to a real visitor: how fast the main content loads (LCP), how quickly the page responds when someone taps (INP), and how much it jumps around while loading (CLS). Google measures them from real visits and uses them as one signal in ranking.

What is a good score for each one?

Aim for LCP under 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1. Google rates a page “poor” above 4 seconds for LCP, above 500 milliseconds for INP, or above 0.25 for CLS. All three need to be good for the page to pass.

What replaced FID?

INP, or Interaction to Next Paint, replaced First Input Delay in March 2024. FID only measured the delay on a visitor’s first interaction. INP looks at every interaction during the visit, so it’s a more honest measure of whether the page feels responsive.

Do Core Web Vitals affect SEO?

Yes, but modestly. They’re a genuine ranking signal that acts as a tiebreaker between pages of similar quality. They won’t rescue weak content or outrank a much better page, so treat them as polish on top of good content, not a substitute for it.

What’s the difference between lab data and field data?

Lab data is a single test run on a simulated device, which you get from PageSpeed Insights. Field data is the real scores from actual visitors over the last 28 days, and it’s what Google grades you on. A page can pass in the lab and fail in the field, so always check the field numbers.

Which Core Web Vital is hardest to fix?

INP tends to be the toughest because it usually means trimming or replacing the code and plugins running on your site, which takes more work than swapping an image. Most simple small business sites pass INP already, so you may never need to touch it.

Ready to take the first step?

A group of people in business attire collaborate in a modern office, standing by a large whiteboard covered with diagrams, notes, and sticky notes, while others sit at a table with laptops and papers.