Core Web Vitals: What They Measure and How to Fix Them
Core Web Vitals sound like a technical abstraction, but they measure something very ordinary: whether your website feels fast, responsive and stable to a real person on a real phone. Google uses them as a ranking signal, and they correlate closely with whether visitors stay or leave. This guide explains what each metric measures, what counts as good, and exactly what to change to improve each one.
Key takeaways
- Three metrics: LCP (loading), INP (responsiveness) and CLS (visual stability).
- Targets: LCP under 2.5 seconds, INP under 200 milliseconds, CLS under 0.1.
- Google judges the mobile experience, so test on mobile, not desktop.
- Images are the most common cause of poor LCP. Compression alone often fixes it.
- CLS is usually the cheapest to fix and the most irritating to visitors when ignored.
Why Google Measures This at All
Google's commercial interest is in sending people to pages they are satisfied with. A page that takes eight seconds to appear, freezes when tapped, or shifts under a finger produces an unhappy searcher. Rather than relying on subjective judgement, Google measures three specific aspects of that experience and uses them as one ranking signal among many.
It is worth being clear about weight. Core Web Vitals will not push an irrelevant page above a relevant one. Where two pages are otherwise comparable, they act as a tiebreaker, and increasingly a meaningful one. Their bigger value is commercial: faster, more stable pages convert better regardless of what Google does.
LCP: Largest Contentful Paint
What it measures
How long it takes for the largest visible element in the initial view to finish rendering. Usually that is your hero image, a banner, or a large block of heading text. In practical terms, LCP answers the question: how long before the visitor sees the main thing on the page?
| Rating | LCP time |
|---|---|
| Good | 2.5 seconds or less |
| Needs improvement | 2.5 to 4 seconds |
| Poor | Over 4 seconds |
What usually causes a poor LCP
- Oversized images. A hero photograph uploaded straight from a camera at several megabytes is the single most common cause.
- Slow server response. Cheap shared hosting, or hosting located far from your audience, delays everything that follows.
- Render blocking CSS and JavaScript. Files that must load before anything can display.
- Web fonts loading late. Text that cannot paint until a font file arrives.
How to fix it
- Compress and resize every image. Serve images at the size they actually display, in a modern format such as WebP. This alone frequently halves LCP.
- Do not lazy load the hero image. Lazy loading is excellent below the fold and harmful for the first visible image.
- Improve hosting. Quality hosting served close to your audience reduces server response time, which sets the floor for everything else.
- Add caching and a CDN. Cached pages skip most of the work; a CDN delivers assets from a nearer location.
- Preload the critical image and font. Tells the browser to fetch them immediately rather than discovering them late.
INP: Interaction to Next Paint
What it measures
How quickly the page visibly responds after someone interacts with it: tapping a menu, opening an accordion, clicking add to cart. INP replaced the older First Input Delay metric because it measures responsiveness across the whole visit rather than only the first interaction.
| Rating | INP time |
|---|---|
| Good | 200 milliseconds or less |
| Needs improvement | 200 to 500 milliseconds |
| Poor | Over 500 milliseconds |
What causes a poor INP
Almost always JavaScript. When the browser is busy executing scripts, it cannot respond to taps. Heavy third party code is a frequent offender: chat widgets, analytics stacks, advertising pixels, review widgets and popup tools all compete for the same thread. Page builders and plugin heavy WordPress sites accumulate this quietly over time.
How to fix it
- Audit third party scripts. Remove anything not earning its place. Most sites carry at least one tracker nobody uses any more.
- Defer non critical JavaScript so it loads after the page becomes interactive.
- Reduce plugin count where plugins overlap or are dormant.
- Delay chat and popup widgets until after the page is usable, or until the visitor scrolls.
CLS: Cumulative Layout Shift
What it measures
How much the page moves around while loading. If you have ever tried to tap a link and had an advert push it aside at the last moment, you have experienced a high CLS. It is measured as a score rather than a time.
| Rating | CLS score |
|---|---|
| Good | 0.1 or less |
| Needs improvement | 0.1 to 0.25 |
| Poor | Over 0.25 |
What causes it and how to fix it
- Images without dimensions. Always set width and height attributes so the browser reserves the space before the image arrives.
- Ads, embeds and iframes. Reserve a fixed container size rather than letting them expand on arrival.
- Late loading fonts. Use font-display swap and preload your main font so text does not reflow when the custom font loads.
- Content injected above existing content. Cookie banners and promotional bars should overlay rather than push the page down.
How to Test Your Own Site Properly
There are two kinds of measurement and confusing them causes a lot of unnecessary panic.
- Lab data is a simulated test run on demand, as in the main PageSpeed Insights score. It is useful for diagnosis and repeatable, but it is not what Google ranks on.
- Field data is collected from real Chrome users visiting your site. This is what Google actually uses. It appears at the top of PageSpeed Insights and in the Core Web Vitals report inside Search Console.
Always check the mobile tab, because mobile is what Google indexes and ranks. And note that field data updates on a rolling basis, so improvements typically take a few weeks to appear even after the fix is live. Do not assume the change failed because the score did not move the next morning.
A Realistic Order of Work
| Step | Action | Typical effort |
|---|---|---|
| 1 | Compress and correctly size all images | Low effort, large gain |
| 2 | Set width and height on every image | Low effort, fixes most CLS |
| 3 | Enable caching | Low effort, large gain |
| 4 | Remove unused plugins and third party scripts | Medium effort, improves INP |
| 5 | Preload critical font and hero image | Medium effort, improves LCP |
| 6 | Upgrade hosting or add a CDN | Higher cost, raises the ceiling |
| 7 | Replace a bloated theme | Highest effort, last resort |
What Good Scores Are Actually Worth
Passing Core Web Vitals will not transform a site with thin content or no local presence. What it does is remove a handicap, and improve the commercial performance of every visit you already receive. Faster pages hold more visitors, convert more of them, and lose fewer at checkout.
Treat it as maintenance rather than a project. Scores degrade as content, plugins and tracking accumulate, so check quarterly. If you would like the work handled properly, our website speed optimisation service covers image and code optimisation, caching, hosting review and Core Web Vitals remediation with before and after measurements.
Why UAE Sites Fail These Metrics More Often Than You Would Expect
Two patterns show up repeatedly when we audit websites here. The first is image heavy design. Property listings, restaurant menus, salon galleries and product catalogues are all visual by nature, and photographs are frequently uploaded straight from a phone or camera at several megabytes each. A single page carrying eight uncompressed images can be twenty times heavier than it needs to be.
The second is hosting distance. A site hosted on a cheap plan with servers on another continent adds hundreds of milliseconds to every single request before any optimisation work has a chance to help. For a business whose customers are all in the Emirates, hosting served from or near the region is one of the highest return technical decisions available.
The mobile network factor
Field data is collected from real visitors, and a large share of UAE browsing happens on mobile data while people are moving. A page that loads instantly on office fibre can perform poorly in field measurements for exactly that reason. This is why testing on your own phone, on mobile data, away from your office wifi, is the most honest check you can run.
Turning Scores Into Business Decisions
It is easy to become preoccupied with reaching a perfect score, which is rarely the best use of money. The commercially sensible approach is to get comfortably into the good range on all three metrics and then stop, redirecting effort toward content and conversion instead.
- If you are failing badly, fix it. The gap between a four second and a two second load is customers.
- If you are marginally below the threshold, the cheap fixes are worth doing and the expensive ones usually are not.
- If you are comfortably passing, stop optimising and check again in three months.
- If a fix requires rebuilding the site, weigh it against everything else that rebuild would deliver rather than treating speed as the only justification.
A Note on Tools and Conflicting Scores
You will find plenty of speed testing tools online and they will disagree with each other, sometimes dramatically. This is normal and not worth worrying about. Each tool tests from a different location, on a different simulated connection, with different assumptions about device power. A site can score 92 on one and 61 on another within the same minute.
Pick one source of truth and ignore the rest. For Core Web Vitals specifically, that source should be the field data in Google Search Console, because it is measured from your actual visitors and it is the data Google uses. Treat everything else as a diagnostic aid for finding what to fix, not as a scoreboard.
Frequently Asked Questions
How much do Core Web Vitals affect rankings?
They are one signal among many and act mainly as a tiebreaker between pages of comparable relevance. Their bigger effect is commercial: faster, more stable pages keep more visitors and convert better, which benefits you regardless of ranking impact.
My PageSpeed score is 95 but Search Console says I am failing. Why?
The PageSpeed score is lab data from a simulated test, while Search Console reports field data from real visitors on real devices and connections. Real users on mid range phones and variable mobile networks have a slower experience than a simulated test. Trust the field data, because that is what Google uses.
How long before improvements show in Search Console?
Field data is calculated on a rolling 28 day basis, so expect several weeks before a fix is fully reflected. Use lab tools to confirm the change worked immediately, then wait for the field data to catch up rather than making further changes in the meantime.
Do Core Web Vitals matter for a small local business site?
Yes, though perhaps less for rankings than for conversions. Local searches are overwhelmingly mobile and often urgent, so a slow or unstable page loses customers who were ready to call. A simple local site should comfortably pass all three metrics, and failing usually indicates oversized images or a bloated theme.
Can a WordPress site pass Core Web Vitals?
Easily, provided it is built with care. Passing sites typically use a lightweight theme, compressed and correctly sized images, a caching plugin, a lean plugin stack and decent hosting. Failing WordPress sites usually combine a heavy page builder theme with a large number of plugins on cheap shared hosting.
Is it worth paying someone to fix this?
If your site is failing badly and you rely on search traffic, yes. Most of the gain comes from a handful of fixes that a professional completes in hours rather than days. If you are only slightly below the thresholds, compressing your images and enabling caching may be enough to do it yourself.
