What Are Core Web Vitals, in Plain English?

Search Console has flagged some of your URLs as "Poor" under Core Web Vitals, or a developer has forwarded a PageSpeed Insights report full of LCP, INP and CLS. You need to know whether it matters and who should fix it, and many guides still describe a metric Google retired in 2024.
Core Web Vitals are Google's three measurements of how a page feels to a real visitor. Largest Contentful Paint (LCP) is how long the main content takes to appear. Interaction to Next Paint (INP) is how fast the page reacts when someone taps, clicks or types. Cumulative Layout Shift (CLS) is how much the layout jumps while loading.
What do LCP, INP and CLS actually measure?
Each metric answers one question a visitor asks without realizing it. Together they cover the three moments when a page can let someone down: before anything useful appears, when they try to use it, and while it is still settling.
LCP: has the main content arrived?
LCP times how long the largest element in the visible part of the screen takes to render, counted from when the page started loading. On a product page that is usually the hero image. On a blog post it is often the headline block or the first large image.
It is not the moment the page finishes loading. It is the moment the visitor can see what they came for. That is why it tracks perceived speed better than older "page load" numbers did.
INP: does the page respond when tapped?
INP watches every click, tap and key press during a visit and reports one of the slowest. It measures the gap between the action and the moment the screen visibly changes in response.
If someone taps "Add to cart," nothing happens for a beat, and they tap again, that is an INP problem. The page may look finished. It is simply too busy running code to answer.
CLS: does the page stay still?
CLS measures how much visible elements move unexpectedly while the page loads. The classic case: you go to tap a button, an image or banner loads above it, everything shifts down, and you tap the wrong thing.
Unlike the other two, CLS is not a time. It is a score for how much of the screen moved and how far. Lower is better, and zero means nothing jumped.
Why did FID disappear from the list?
Older guides list First Input Delay (FID) as the third metric. INP replaced it in March 2024, and the difference comes down to what each one watched.
FID only measured the wait before the first interaction began to be processed. That covered one moment, the first one, and it did not include the full response. A page could pass FID comfortably and still feel sluggish every time someone used a filter or opened a menu afterwards.
INP closes both gaps. It looks at every interaction across the visit, and it times through to the moment the screen actually updates. That makes it harder to pass, and a fairer reflection of what using the page feels like.
The practical consequence is simple. Any audit, guide or agency deck that still lists FID as a Core Web Vital was written before the switch, and its advice on responsiveness is incomplete. Google changes the core set rarely and announces changes well ahead. Its developer documentation on web.dev is the place to confirm the current list, rather than a third-party summary.
Where do the scores come from?
Most confusion starts here. There are two kinds of data, and they do not always agree.
Field data comes from real visits. Chrome collects it from users who have not opted out of sharing usage data, and it is published through the Chrome User Experience Report. This is what Search Console's Core Web Vitals report and the top section of PageSpeed Insights show you. It is also the version that counts.
Lab data comes from a simulated load in a tool like Lighthouse, run on a single device profile over a single connection. It is useful for diagnosing why a page is slow, because you can rerun it after every change. But it is a rehearsal, not the performance. A page can score well in the lab and still fail in the field, because real visitors use older phones and patchier networks.
Three more things shape the field numbers:
- Assessment is page by page, not for the whole domain.
- Mobile and desktop are judged separately, so a page can pass on one and fail on the other.
- Each metric is judged across most real visits, not your best visit or the average. A page passes only when all three metrics hold up.
Exactly where those pass marks sit is covered in what a good LCP, INP and CLS score is.
Do Core Web Vitals affect rankings?
Yes, but less dramatically than many SEO pitches suggest. They are one of the signals Google uses to judge page experience, and Google recommends that site owners get them into good shape. It has never published how much weight they carry, and it frames good scores as a recommendation rather than a guaranteed lift. Relevance and content quality still matter more. A fast page with the wrong answer will not beat a slower page with the right one.
The stronger reason to care is the visitor. The three metrics describe the exact moments people give up:
- staring at an empty hero area
- tapping a button that does nothing
- hitting the wrong link because the page moved
Those moments cost you engagement whether or not Google is watching. They hit paid traffic just as hard as organic, and you have already paid for the click before the page lets the visitor down.
What should you do with a poor score?
Read it as three separate problems. Each usually has a different cause, and often a different owner.
- A poor LCP usually points to heavy images, a slow server response, or code that blocks the page from rendering. Compressing and properly sizing the hero image is often the first fix.
- A poor INP usually points to too much JavaScript running at once. Long tasks keep the browser busy when a visitor tries to interact. Breaking those tasks up, and auditing third-party scripts, is developer work.
- A poor CLS usually points to images, ads or embeds with no reserved space, or fonts that swap in late. Setting image dimensions and preloading fonts fixes much of it.
Start with mobile, where real visitors are on slower hardware. It is worth working through the broader question of whether your site is too slow on phones before you chase individual metrics.
Then fix templates rather than single URLs. A slow product template drags down every product page built on it, and one fix lifts them all.
Conclusion
Core Web Vitals are less mysterious than the acronyms make them look:
- LCP asks whether the main content has arrived.
- INP asks whether the page answers when touched.
- CLS asks whether the page stays still.
Google scores them from real Chrome visits, page by page and device by device, and treats them as one input among many. Read a poor score as a precise description of where visitors are getting frustrated. Hand each metric to the person who can fix it, and judge progress by field data rather than a single lab run.
Frequently Asked Questions
Lighthouse loads a page in a simulated environment with no real person using it, so there are no taps, clicks or key presses to time. INP can only come from field data gathered from real visitors. Use lab runs to find long-running scripts that are likely to slow interactions. Then confirm your actual INP in Search Console or in the field section of PageSpeed Insights.
Field data comes from real Chrome visits. A new page, a new site, or a page with very little traffic may not have enough visits to report. Until it does, Search Console and PageSpeed Insights cannot show a Core Web Vitals assessment for it. Use lab tools such as Lighthouse to catch obvious problems in the meantime, and expect field data to appear as traffic builds.
No. PageSpeed Insights shows two things on one page: field data from the Chrome User Experience Report, where enough exists, and a Lighthouse lab test run on demand. The field section reflects how real visitors experienced the page, and it is what Google assesses. The Lighthouse section is a diagnostic simulation. When the two disagree, trust field data for the verdict and lab data for the cause.
Get a complimentary audit to uncover AI opportunities hidden in your data.
Put these strategies to work


