Core Web Vitals case studies: what faster sites really gained
Short answer The documented wins are mostly sales and engagement, not rankings. In Google-published case studies, Vodafone got 8% more sales after cutting load time by 31%, and Rakuten 24 saw 33% more conversions. Google says Core Web Vitals count for ranking but never outweigh relevance. Speed up your site for customers; treat any ranking lift as a bonus.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callWhat the case studies actually show
Every result below is published on web.dev, Google's site for web developers, and written with the company involved. Each one measured money or engagement. None reports a ranking gain.
| Company | What they fixed | What changed | How it was measured |
|---|---|---|---|
| Vodafone (telecom) | Main content load (LCP) cut 31% on a landing page | 8% more sales, 15% better lead-to-visit rate | A/B test, half the paid traffic each way |
| Rakuten 24 (Japan, retail) | Core Web Vitals and related speed measures | 33.13% higher conversion rate, 53.37% more revenue per visitor | A/B test for a month, 50/50 split |
| redBus (bus tickets) | Tap response (INP) on search results | About 7% more sales | Before and after |
| Swappie (used phones) | LCP down 55%, layout shift (CLS) down 91% | Mobile revenue up 42% in three months | Before and after, mobile compared with desktop |
| Yahoo! JAPAN News | Layout shift, from about 0.2 to 0 | 15% more page views per session, 13% longer sessions | Before and after |
| The Economic Times (India, news) | LCP from 4.5 to 2.5 seconds, CLS from 0.25 to 0.09 | 43% lower bounce rate | Before and after |
Sources: Vodafone, Rakuten 24, redBus, Swappie, Yahoo! JAPAN News, The Economic Times.
The three Core Web Vitals, in plain words:
- LCP (Largest Contentful Paint): how long until the biggest image or block of text shows.
- INP (Interaction to Next Paint): how fast the page reacts when someone taps or clicks.
- CLS (Cumulative Layout Shift): how much the page jumps around while it loads.
INP replaced the older FID (First Input Delay) on March 12, 2024 (web.dev). A case study that still scores FID predates the change. Pass marks and how to read your own report are in Core Web Vitals: does website speed affect Google ranking?.
Did the faster sites rank higher?
We found no published case study that credibly ties a speed fix to a ranking jump. Articles that claim big ranking gains for famous brands rarely link to any data. Google's own words explain why the ranking effect is small:
"Google Search always seeks to show the most relevant content, even if the page experience is sub-par." (Google Search Central)
The same page says Core Web Vitals "are used by our ranking systems," that good scores don't "guarantee that your pages will rank at the top," and that chasing "a perfect score just for SEO reasons may not be the best use of your time."
So speed can tip a close race between two equally useful pages. It won't lift a weak page over a better answer. The money in the table came from visitors who stayed and bought, not from new visitors sent by Google.
How much to trust these numbers
Weigh each result by how it was measured.
- A/B tests are the strongest. Vodafone and Rakuten 24 split traffic between a fast and a slow version of the same page at the same time, so season and ad spend hit both equally (Vodafone, Rakuten 24).
- Before-and-after results can mix causes. NDTV cut LCP by 55% and saw bounce rate fall 50%, but the case study itself says other product changes happened at the same time, so speed can't be proven as the only cause (web.dev).
- These are success stories. Google publishes the wins. Projects that changed nothing don't get a case study.
- These are big sites. Millions of visits, large dev teams. A local service site with 1,500 visits a month can't measure a 7% change reliably.
The widest study is still useful as a direction, not a forecast. Across 37 large brand sites in late 2019, commissioned by Google and run by Deloitte and 55, a 0.1-second improvement on four speed measures came with up to 21.6% more lead-generation users reaching the form page (web.dev). Every page in the journey had to get faster to see it (same page).
What a small business can take from them
The fixes behind these results are ordinary. Match the fix to the metric that fails on your own pages.
| If this fails | What the case studies did | Small business version |
|---|---|---|
| LCP | Main content rendered on the server, smaller hero images (Swappie, Vodafone) | Shrink the top photo, drop slider carousels, use decent hosting |
| CLS | Reserved space for images before they load (Yahoo! JAPAN News) | Set image and embed sizes; stop banners from pushing the page down |
| INP | Lighter work on each tap (redBus) | Remove chat, tracking or booking scripts you don't use |
Sources: Swappie, Vodafone, Yahoo! JAPAN News, redBus.
Start with the page people book or call from, tested on a phone over mobile data. That is where Vodafone found its gain: a landing page fed by paid ads.
Example
Illustrative cases, not measured results.
A roofer paying for Google Ads
His ad landing page takes 7 seconds to show the quote form on a phone. He compresses a 3 MB header photo and removes an unused video embed. He judges the fix by cost per lead over the next two months, not by rankings.
A dental office chasing a score
Its PageSpeed score is 60, but real-visitor data passes all three metrics. Moving to 95 would cost $1,500 in developer time with little expected ranking effect. The office spends it on treatment pages instead.
An online boutique with jumping buttons
Shoppers tap "add to cart" just as a promo bar loads and pushes it down. Reserving space for the bar fixes the layout shift. The owner compares add-to-cart rate for the four weeks before and after, and notes any sales running in either period.
FAQ
Do Core Web Vitals affect rankings at all?
Yes, a little. Google says they are used by its ranking systems, but relevance comes first (Google Search Central).
Are there Core Web Vitals case studies for local businesses?
We found none with published data. The sourced studies are large retail, travel, telecom and news sites: a direction, not a forecast.
Which metric pays off most?
The one your key page fails. Vodafone gained from LCP, redBus from INP, Yahoo! JAPAN News from CLS.
How do I measure my own result?
Track one number (calls, form fills, bookings or sales) for four weeks before and after. Note other changes, like ad spend or season.
Different for you?
- Want the thresholds and why your score and Search Console disagree: Core Web Vitals and Google ranking.
- Fast site, still no calls: website not getting leads.
- Fixing the page your ads send people to: landing page best practices.
- Finding your slow pages in Search Console: Google Search Console for a small business.
- Want a fast site built and kept fast: Website Agent.