Three numbers that decide how your page feels

Three numbers decide how your page feels to the people actually using it, and each one has a clear pass mark. Google calls them Core Web Vitals: LCP, INP and CLS. The names sound like a tax form. The ideas behind them are things you've felt on every slow website you've ever closed in a huff.

SEO Fix Friday is my weekly series: one technical idea, about a minute of video. Last Friday's episode was the map, the four stages every search runs through before a result appears. That one was about whether Google can find your page at all. This one starts after someone has landed on it. The page is found, it's ranked, a real person has tapped through, and now it has to behave.

LCP: the blank menu board

Largest Contentful Paint is the time until the biggest image or block of text in view appears on screen. Google's pass mark is 2.5 seconds or less.

Picture a takeaway shop where the door is open and the lights are on, but the menu board is still blank. Technically the shop is open. You still can't order anything. LCP measures how long you stand there before the big thing you came for shows up, usually the main photo or the headline.

INP: the crossing button

Interaction to Next Paint covers the moment someone clicks, taps or presses a key. It measures how fast the page shows a response on screen. Good is 200 milliseconds or less.

Think of the pedestrian crossing button that doesn't light up when you press it, so you press it four more times. That's poor INP. The button may well have worked the first time. Nobody could tell, so it felt broken.

It's also the newest of the three. INP replaced First Input Delay in 2024, which means any audit template that still talks about FID is out of date.

CLS: the page that won't sit still

Cumulative Layout Shift is the formal name for content jumping around when you don't expect it. Images without set sizes, web fonts swapping in and ads that resize are the usual causes. Good is 0.1 or less.

Mr. Mummu, who wanders into these posts now and then, was reading a dosa batter recipe on his phone with wet hands. Every time he went to tap "next step", a banner loaded above it and the whole page dropped. He kept tapping an advert for garden gnomes instead. So he did what he always does and built a system. He stuck a tiny sticker on his screen exactly where the button lived, to save himself hunting for it. Next load, the page jumped again, and the sticker was now pointing with total confidence at the gnomes. He opened his notebook marked "Urgent", wrote "sticker not the problem" just under "curry leaves", and announced to the kitchen: "The button should stay where they first put it."

He's right, and that's the whole metric. People aim at where things are. If the page moves under their thumb, they hit the wrong thing and blame you, fairly.

The number that catches people out

Here's the part I'd underline. Google judges these on real visits, at the 75th percentile, with mobile and desktop counted separately. Roughly speaking, three in four real visits need to clear the bar. So one quick test on a fast laptop with good office internet proves very little about someone loading your page on an older phone in a car park.

There's a lab catch too. Lighthouse can't measure INP, because nobody clicks anything in a lab test. It shows Total Blocking Time instead, which is only a stand-in. The real number comes from field data, the kind Search Console and PageSpeed Insights show.

Where I land on this

My view: treat Core Web Vitals as a bar to clear, then get back to the content.

Google's documentation is unusually blunt about it. Good scores help, but they don't guarantee top rankings. It says chasing a perfect score just for SEO reasons may not be the best use of your time, and that Google still tries to show the most relevant content even when the page experience is sub-par. Where the scores earn their keep is when lots of pages answer the same query about equally well. Then feel matters.

So a fast, steady page with nothing useful on it is still a page with nothing useful on it. And a genuinely helpful page that makes people wait, then jumps under their thumb, is leaving goodwill on the table.

One honest limitation. Everything above is a simplified version of what Google publishes in its Web Vitals guides and its Search Central documentation. Real pages have more moving parts, and two sites can fail the same metric for completely different reasons. Use the three numbers as better questions about your own pages, not as a full diagnosis.

And a soft disclaimer, because every site is different: this is general education, not advice for your specific site.

The video for this one runs about a minute and walks through all three numbers with their pass marks. It's episode two of SEO Fix Friday on my YouTube channel, and I'll drop it in the comments. It finishes where Google's docs point everyone first, the Core Web Vitals report in Search Console.

Learn what each number feels like from the visitor's side. The scores make a lot more sense after that.