Core Web Vitals and Revenue: What the Data Actually Says

When Vodafone ran an A/B test on their Italian e-commerce site, they sent 100,000 clicks per day to an optimised landing page and 100,000 to the original. A 31% improvement in Largest Contentful Paint led to 8% more sales, a 15% increase in leads and an 11% improvement in cart visits. There was no redesign and there were no new features, they just made the page faster.

That kind of result raises an obvious question: is this a one-off, or is there a predictable, measurable link between how fast your website loads and how much money it makes?

Years of published data from Google, Deloitte and dozens of major companies point to faster sites making more money, and we now have specific numbers to show it.

What Core Web Vitals actually measure

Google uses three metrics to judge what it's really like to visit your website, and together they're called Core Web Vitals.

Largest Contentful Paint (LCP) measures how long it takes for the main content of a page to appear on screen. This is the moment a visitor sees something useful, like a headline, a product image or a hero section. Google counts anything under 2.5 seconds as "good" and anything over 4 seconds as "poor."

Interaction to Next Paint (INP) measures how quickly a page responds when someone taps a button, clicks a link or types into a form. If your page feels slow when a visitor tries to do something, INP picks that up. Under 200 milliseconds is "good" and over 500 milliseconds is "poor." INP replaced the older First Input Delay metric in March 2024.

Cumulative Layout Shift (CLS) measures visual stability, which means whether things on the page jump around as it loads. If you've ever tried to tap a button on a mobile site only for the page to shift and send you somewhere else, that's a CLS problem. Below 0.1 is "good" and above 0.25 is "poor."

Between them, these three ask simple questions about a visit: did the page load quickly, did it respond when I used it, and did it stay still while I was reading?

Google judges your site using real user data from Chrome browsers. They use the 75th percentile, which means at least 75% of your visitors need a "good" experience for your site to pass. As of September 2025, only 53% of websites pass all three thresholds.

The evidence: what published case studies actually show

Major companies across lots of sectors have published their results, often working with Google, and the numbers and methods are out in the open for anyone to check.

Vodafone: 8% more sales from faster loading

Vodafone Italy ran a controlled A/B test, splitting traffic equally between a performance-optimised landing page and the original. The optimised page got a 31% improvement in LCP, and with it came 8% more sales, 15% more leads and 11% more cart visits. As Vodafone's team put it: "The introduction of LCP as the main KPI for page performance made it possible to significantly improve the customer experience of our e-commerce."

Rakuten 24: 53% more revenue per visitor

Rakuten 24, one of Japan's largest online retailers, put money into Core Web Vitals improvements and measured the impact carefully. Revenue per visitor went up by 53.37% and conversion rate went up by 33.13%. They also found that getting a good LCP score could lead to up to a 61% increase in conversion rate and a 26% increase in revenue per visitor, which are big gains.

Swappie: 42% increase in mobile revenue

Swappie, a European refurbished phone retailer, had spent years adding features without paying much attention to performance. When they noticed mobile results falling behind desktop, they changed their priorities. After reducing LCP by 55%, CLS by 91% and FID by 90%, mobile revenue went up by 42% within three months. Their relative mobile conversion rate improved from 24% to 34%.

AliExpress: 15% reduction in bounce rate

AliExpress improved their CLS by 10 times and doubled their LCP performance. That gave them a 15% reduction in bounce rate, so fewer visitors left straight away because the page was slow or jumpy.

Yahoo! Japan News: 15% more page views per session

Yahoo! Japan focused on reducing layout shifts across their news platform. After a 98% reduction in pages with poor CLS scores, they measured a 15% increase in page views per session and a 13% increase in session duration. For a publisher paid for by advertising, more page views per session means more revenue.

NDTV: 50% reduction in bounce rate

Indian news publisher NDTV improved their LCP by 55%, bringing it from 3.0 seconds to 1.6 seconds. They saw a 50% reduction in bounce rate, which is a big change in how people used the site from work on one metric.

Redbus: 80-100% improvement in mobile conversions

Redbus, a transport booking platform that runs in multiple countries, fixed their CLS (reducing it from 1.65 to zero) and cut their Time to Interactive from roughly 8 seconds to 4 seconds. They saw an 80–100% improvement in mobile conversion rate across their global sites, along with good gains in domain rankings.

The Telegraph: measurable conversion uplift from CLS fixes

The Telegraph, one of the UK's largest publishers, improved their 75th percentile CLS by 250% (from 0.25 to 0.1). The share of their pages passing CLS thresholds grew from 57% to 72%. Analysis using Quantum Metric showed that removing poorly performing code increased page performance by 17%, which created a 1.35% uplift in conversions. When a publisher handles millions of sessions, even small percentage gains add up to a lot of revenue.

iCook: 10% more ad revenue

Taiwanese recipe platform iCook improved their CLS by 15% by sizing ad slots better and saving space for them in the layout ahead of time. That led to a 10% increase in ad revenue, which they put down to the more stable layout.

The Deloitte study: even 0.1 seconds matters

In 2020, Deloitte published "Milliseconds Make Millions," a study that looked at over 30 million user sessions across 37 leading European and American brand sites. They found that a 0.1-second improvement in load time improved progression rates across the whole purchase funnel.

Here's how that broke down by sector:

• Retail: conversions increased by 8.4% and average order value increased by 9.2%
• Travel: conversions increased by 10.1%
• Luxury brands: page views per session increased by 8.6%, with a 40.1% increase in users moving from product detail to add-to-basket
• Lead generation: bounce rate improved by 8.3%

Those gains came from taking just one tenth of a second off load time, without a full overhaul of the site.

The broader pattern: every second costs you money

Research from Portent, based on 20 websites and over 100 million page views, found that website conversion rates drop by an average of 4.42% with each extra second of load time between 0 and 5 seconds. A site that loads in 1 second has a conversion rate 3 times higher than a site that loads in 5 seconds, and for B2B sites that goes up to 5 times.

Google's own data shows that 53% of mobile users leave a page if it takes more than 3 seconds to load. As load time goes up, the number of people who give up rises faster and faster, so the first few seconds matter a lot and the cost of a slow page builds up quickly.

How Core Web Vitals affect your search rankings

Since 2021, Core Web Vitals have been an official ranking signal in Google's algorithm. They're part of the wider "page experience" signals, alongside mobile-friendliness, HTTPS and not having pop-ups that get in the way of the content.

Google has been careful to say that relevant content still matters most, so a slow page with excellent content will still rank above a fast page with thin content. But when several pages with similar content compete for the same search, performance can decide which one wins, and in busy markets that can decide who makes it onto page one.

Over time the effects build on each other. Better Core Web Vitals help your rankings, better rankings bring more traffic, and more traffic to a faster site brings more conversions. Those extra conversions then make it easier to justify more work on performance, so the businesses that start early tend to get the most out of it.

The March 2024 core update backed this up. Sites with poor metrics saw traffic drops, while sites that had put work into performance saw gains. Google keeps saying the same thing, that user experience is a ranking factor and Core Web Vitals are how they measure it.

Where to start: prioritising by business impact

Some performance work pays back more than others. So if you're deciding where to spend first, what does the data suggest?

1. Fix LCP first

LCP has the clearest and best documented link to revenue. The Vodafone, Rakuten, NDTV and Swappie case studies all show LCP improvements leading to measurable increases in sales and conversions. Common fixes include optimising images, improving server response time and removing render-blocking resources.

2. Address CLS on high-traffic pages

CLS problems often come from ads, images without set dimensions, or content that loads late and pushes other things around. The Yahoo! Japan, Telegraph and iCook studies show that CLS fixes improve engagement and revenue, especially on pages with lots of content and ads. The fixes are usually simple, like reserving space for ads, adding width and height attributes to images and avoiding late-loading DOM injections.

3. Measure INP on interactive pages

INP matters most on pages where people are actively doing things, like filling out forms, filtering products, using configurators or finding their way around more complicated screens. If your visitors are mostly reading, INP may already be fine. If your conversion flow needs people to click and type, though, poor INP will cost you completions.

4. Use real user data, not lab tests

Lab tools like Lighthouse are useful for finding issues, but they test performance under controlled conditions. The Chrome User Experience Report (CrUX) measures what your actual visitors experience, and Google uses CrUX data for rankings, while Lighthouse scores play no part in them. So it's worth making sure you're watching the same numbers Google is.

5. Test changes with A/B experiments

The most credible case studies, including Vodafone, Rakuten and Swappie, all used A/B testing to separate out the impact of performance changes, and it's the most reliable way to know. If you can't run formal A/B tests, at least track your Core Web Vitals in Google Search Console alongside your conversion data over time, and compare the trends.

The bottom line

The link between Core Web Vitals and revenue has been measured, published and repeated across dozens of companies, in many industries and countries. Vodafone saw 8% more sales, Rakuten saw 53% more revenue per visitor and Swappie saw 42% more mobile revenue. Deloitte showed that even 0.1 seconds makes a measurable difference to conversion rates.

If you run a UK business in a crowded market, whether that's e-commerce, professional services or lead generation, website performance affects your revenue. It deserves the same attention you give your pricing, positioning and marketing spend, so try not to hand it to a developer and forget about it.

Most performance improvements use known, well documented fixes, and they don't need a full rebuild. A good first step is to open Google Search Console, look at the Core Web Vitals report for your busiest pages, and start with whichever ones fail LCP.

References

1. Vodafone: A 31% improvement in LCP increased sales by 8%, web.dev
2. How Rakuten 24's investment in Core Web Vitals increased revenue per visitor by 53.37%, web.dev
3. How Swappie increased mobile revenue by 42% by focusing on Core Web Vitals, web.dev
4. The business impact of Core Web Vitals, web.dev (includes AliExpress, NDTV, Redbus, iCook, and Yahoo! Japan data)
5. Improving Cumulative Layout Shift at Telegraph Media Group, web.dev
6. How CLS optimizations increased Yahoo! JAPAN News's page views per session by 15%, web.dev
7. NDTV achieved a 55% improvement in LCP by optimizing for Core Web Vitals, web.dev
8. Milliseconds Make Millions, Deloitte / Google, web.dev
9. Site Speed is (Still) Impacting Your Conversion Rate, Portent
10. Web Vitals, web.dev (metric definitions and thresholds)