Core Web Vitals are Google's attempt to reduce page experience to three numbers a team can actually act on: how fast the main content appears, how quickly the page responds when someone interacts with it, and how much the layout moves around while it loads.
They are a confirmed ranking signal, folded into page experience since June 2021. They are also, on their own, a weak one. Content relevance still dominates. The honest framing is that Core Web Vitals act as a tiebreaker: between two pages of comparable quality and relevance, the faster one has a measurable advantage, and in competitive niches that margin is worth having.
The Three Metrics and Their Thresholds
Largest Contentful Paint (LCP): under 2.5 seconds
LCP measures when the largest visible element in the viewport finishes rendering. Usually that is a hero image, a video poster or a large block of heading text. It is the closest available proxy for the moment a visitor feels the page has arrived.
Interaction to Next Paint (INP): under 200 milliseconds
INP measures how long the page takes to visibly respond after a user interacts with it, across the whole visit rather than just the first interaction. It replaced First Input Delay in March 2024, and it is a far harder test. FID measured only the delay before processing started. INP measures the full round trip through to the next paint.
Cumulative Layout Shift (CLS): under 0.1
CLS measures unexpected movement of visible content during loading. The familiar failure is reaching for a button just as an image finishes loading above it and pushes everything down.
The 75th Percentile Rule
This is the part teams most often get wrong, and it explains a lot of confusion about why a "fast" site fails.
Google evaluates each metric at the 75th percentile of real visitor data, gathered from actual Chrome users, not from a lab test. Three quarters of your visits must come in under the threshold. Your average does not matter. Your own experience on a fast connection on modern hardware definitely does not matter.
All three metrics must pass at the 75th percentile for a page to be assessed as good overall. Passing two out of three is a fail.
This also means lab tools and field data disagree for legitimate reasons. Lighthouse runs a simulated load on a synthetic device and cannot measure INP meaningfully at all, because there is no real user interacting. Treat lab tools as diagnostics and field data as the score.
INP Is Where Most Sites Fail
Around 43 percent of sites still miss the 200 millisecond INP threshold, making it comfortably the most commonly failed Core Web Vital in 2026. It is also the one that demands the deepest technical changes, because the usual cause is too much JavaScript competing for the main thread rather than an asset that can be compressed.
Because it is both the most-failed metric and the hardest to fix, INP deserves its own treatment. Our guide to diagnosing and fixing INP covers long tasks, input delay and the presentation delay that most teams forget to measure.
Fixing LCP
LCP problems are usually delivery problems, and they resolve in a predictable order.
- Find the actual LCP element first. Chrome DevTools names it. Optimizing the wrong image is the most common wasted afternoon in performance work.
- Serve modern formats. AVIF and WebP cut image weight substantially against JPEG at equivalent quality.
- Preload it. A
rel="preload"hint for the LCP image stops the browser discovering it late. - Never lazy-load it. Lazy loading the hero image is self-defeating and remains extremely common.
- Cut render-blocking resources. Inline the critical CSS, defer the rest.
- Fix the server. If time to first byte is already 800ms, no front-end work will save you.
Fixing CLS
CLS is usually the cheapest of the three to fix, because the causes are few and well understood.
- Set explicit
widthandheighton every image and video so the browser reserves the space before the file arrives. - Reserve space for ads, embeds and iframes with a min-height rather than letting them expand into the layout.
- Use
font-display: optionalor match fallback font metrics so a web font swap does not reflow the text. - Insert banners, notices and consent bars above or below existing content, never in a way that pushes it down.
Where to Start
Pull your field data from the Core Web Vitals report in Search Console. It groups URLs by the issue affecting them, which tells you whether you have one broken template or a sitewide problem. Fixing a template fixes every page built from it, and that is almost always the highest-return move available.
Then confirm improvements in field data over the following weeks. The report uses a 28-day rolling window, so changes take time to appear. Teams routinely conclude a fix did not work when the data has simply not caught up yet.
