Short answer (TL;DR)
Core Web Vitals are Google's three user experience metrics: LCP measures loading speed, INP measures responsiveness, CLS measures visual stability. A page passes when 75 percent of visitors get an LCP under 2.5 seconds, an INP under 200 milliseconds and a CLS under 0.1. According to the HTTP Archive Web Almanac 2025, only 48 percent of pages meet all three on mobile. Slowness costs money: Akamai research found every second of delay cuts conversions by 7 percent, and Google's data shows 53 percent of mobile visitors abandon a page that takes more than 3 seconds to load. You can check your own site for free in PageSpeed Insights.
What Core Web Vitals are
Core Web Vitals are Google's standard for measuring the user experience of a web page. Google introduced them in 2020 and has used them as one ranking signal ever since. The data comes from real visits collected through the Chrome browser, in a dataset called CrUX (Chrome User Experience Report). Google is not measuring your site in a lab, it is measuring it on your actual customers' phones and their connections.
The score is calculated at the 75th percentile over a rolling 28-day window. In plain terms: at least three quarters of visits have to hit the "good" threshold on all three metrics at the same time. One clean lab test on a fast office computer means nothing if half your customers have a poor experience on their phones.
The three metrics: LCP, INP and CLS
| Metric | What it measures | Good | Poor |
|---|---|---|---|
| LCP (Largest Contentful Paint) | Time until the largest element on the page appears | under 2.5 s | over 4.0 s |
| INP (Interaction to Next Paint) | Responsiveness to a click, tap or keypress | under 200 ms | over 500 ms |
| CLS (Cumulative Layout Shift) | Content moving around while the page loads | under 0.1 | over 0.25 |
LCP: how fast the visitor sees the main content
LCP measures the time from clicking a link to the moment the largest element on the page appears. Usually that is the hero image or the main headline. According to the Web Almanac 2025, only 62 percent of pages hit a good LCP on mobile. That is the worst result of the three metrics and the main bottleneck on most websites.
INP: does the page respond to a click
INP replaced the older FID metric in March 2024 and measures the whole path from a click to a visible response on screen. If a visitor hits "Add to cart" and nothing happens for half a second, they hit it three more times or they leave. The main culprit behind a bad INP is too much JavaScript: tracking scripts, chat widgets and plugins that fire on every click.
CLS: does the content jump around
CLS measures how much content moves while the page loads. Classic example: you go to click a link, a banner loads right before your click and pushes the content down, and you click the wrong thing. The cause is almost always images, ads or fonts with no space reserved up front.
How many customers a slow site loses
The research numbers are brutal:
- Akamai research found every second of load delay cuts conversions by 7 percent.
- Google's data shows 53 percent of mobile visitors abandon a page that takes more than 3 seconds to load.
- A 2025 Yottaa analysis covering more than 500 million visits across 1,300 online stores measured that 63 percent of visitors leave pages that take more than 4 seconds to load.
- Google research shows bounce probability rises 32 percent when load time goes from 1 to 3 seconds.
It works in the other direction too. Vodafone improved LCP by 31 percent and increased sales by 8 percent (case study on web.dev). Rakuten 24 optimized Core Web Vitals and lifted conversions by 33 percent and revenue per visitor by 53 percent (web.dev). These are not marketing estimates, they are A/B test results on live traffic.
Does speed affect Google rankings
Yes, but less than the people selling miracle speed fixes claim. Core Web Vitals are one of several hundred ranking signals. Content and links still matter more. Google itself says great content with poor vitals still beats poor content with perfect vitals.
The realistic picture: speed decides between evenly matched competitors. If you and a local competitor are fighting over the same keyword with comparable content, the faster page wins. The indirect effect is even bigger. A slow site has a higher bounce rate, fewer pages viewed per visit and fewer conversions, and over time those signals show up in rankings too.
Where the web stands in 2026
According to the HTTP Archive Web Almanac 2025 (CrUX data from July 2025), 48 percent of mobile pages and 56 percent of desktop pages pass all three metrics at once. The trend is improving: on mobile, 36 percent of pages passed in 2023 and 44 percent in 2024. If your site passes, you are ahead of most of the web. If it does not, your competition probably does not either, which is an opportunity.
By individual metric on mobile: 62 percent of pages have a good LCP, 77 percent a good INP, 81 percent a good CLS. So the biggest room for improvement is loading speed, not responsiveness.
How to check your own site
Three free steps, ten minutes of work:
1. PageSpeed Insights (pagespeed.web.dev). Enter your URL and the tool shows real user data from CrUX plus a lab test. Look at the real user data section at the top, not just the 0 to 100 score. 2. Google Search Console, the Core Web Vitals report. It shows which groups of pages fail and on which metric. You need your site connected to Search Console, which any serious business should have anyway. 3. Test on your own phone. Open the site on cellular data, not on office wifi. If it feels slow to you, it feels slow to your customers.
Smaller small business sites often have no CrUX data in PageSpeed Insights because they do not get enough traffic. In that case go by the Lighthouse lab test and the thresholds in the table above.
The most common causes of a slow site
From practice, ordered by how often they show up:
1. Oversized images. A photo straight off a phone weighs 4 MB when it should be 150 KB on the page. By far the most common mistake on small business sites. 2. Cheap hosting. A server that takes a full second just to start responding (high TTFB) buries your LCP before the first image even loads. 3. Too many plugins and scripts. Every WordPress plugin, tracking script and chat widget adds JavaScript, which blocks responsiveness and wrecks INP. 4. Fonts and banners with no reserved space. Content jumps around while loading and CLS climbs. 5. Everything loads at once. A video in the footer does not need priority over the hero image, but on many sites it gets it.
How to fix speed, step by step
Ordered by impact per hour invested:
1. Compress your images and convert them to WebP or AVIF. Free with Squoosh or TinyPNG. This alone often cuts LCP in half. 2. Turn on caching and a CDN. Cloudflare has a free plan that shortens response times, including for visitors far from your server. 3. Remove plugins and scripts you do not use. Less code means a better INP. 4. Add fetchpriority="high" to your hero image and loading="lazy" to images below the fold. 5. Reserve space for images, banners and embeds with width and height attributes. CLS drops toward zero. 6. If TTFB stays above 600 ms, change hosts. No amount of optimization fixes a slow server.
On WordPress, a caching plugin (WP Rocket, LiteSpeed Cache) plus decent hosting covers most of this list. On custom-built sites, speed has to be part of development from day one, not an add-on at the end.
What speeding up costs and when it pays off
Optimizing an existing site costs less than the customers you are losing. Typical scope of work: measure, compress images, set up caching, clean up scripts, measure again. For a small brochure site that is a few hours, for a store with thousands of products a few days.
The math is simple. A store with 10,000 visits a month, a 2 percent conversion rate and a $60 average order does $12,000 in revenue. Using Akamai's 7 percent per second ratio, going from 5 seconds to 3 seconds delivers roughly 14 percent more conversions, or about $1,700 in extra revenue every month. The one-time cost of optimization pays for itself in the first month.
When is optimization not enough? If the site runs on a ten year old template, has no responsive design and is loaded with outdated plugins, patching it is throwing money away. At that point the right answer is a rebuild, with speed written into the project requirements.
Want a speed audit and a fix plan for your site? [Get in touch with NovusAgency](https://novusagency.co/en/contact).