Back to guides

Seven in ten UK-visited websites pass Google's phone speed test

Seven in ten UK-visited websites pass Google's phone speed test

Data study

Most of the websites UK shoppers use on their phones already clear Google's Core Web Vitals — but slow loading remains the weak link, and the platform a shop runs on splits the field far more sharply than the size of its budget.

70%
UK-visited phone sites passing all three tests
Chrome UX Report, September 2026, 491,119 sites
80.8%
Sites with a good loading score on phones
Largest Contentful Paint; weakest of the three tests
89.4%
Sites with a good tap-response score on phones
Interaction to Next Paint; 380,716 sites assessed
90.2%
Sites with a good layout-stability score on phones
Cumulative Layout Shift; 491,119 sites assessed

A shopper on a bus taps a link to a winter coat. The screen stays blank, then a banner slides in from nowhere, and the photo of the coat arrives last. By the time she reaches for the button, it has moved.

Google has a name for each of those small irritations, and a pass mark for all three. It calls them Core Web Vitals: how fast the main content loads, whether the page jumps about, and how quickly it responds to a tap. With Cyber Monday, the online-only day of the retail peak, a few weeks away, the common assumption among shop owners is a gloomy one. Most of the web is said to fail, and the big names with big budgets are assumed to be the ones getting it right.

The data tells a different story on both counts. Of the 491,119 websites that people in the UK visited on their phones using Chrome in September 2026, 70% passed all three tests; the remainder were too slow to load, too jumpy or too slow to respond for at least a quarter of visits.

Most sites clear the bar, and loading is the hurdle that remains

That pass rate is a level for a single month. On phones that means mostly Android, since visits from Safari on iPhones are not included. The sites are the ones UK people use, wherever they happen to be owned, so this covers far more than British businesses. A site only passes if at least three in four of its visits meet Google's mark on every test.

That makes the gloomy assumption look overdone. A clear majority of the sites in UK shoppers' pockets are doing the basics well enough for Google. The question worth asking is which of the three tests the rest are failing.

Loading speed is the weakest of the three tests

Loading is the answer. About four in five UK-visited websites had a good loading score on phones, which makes it the weakest of the three tests by some distance. Responding to taps and keeping the layout steady both come in at roughly nine in ten.

Google's loading test times how long the largest image or block of text on the screen takes to appear. On a shop's home page or product page, that is the moment the shopper first sees what she came for.

020406080100Loading (Largest Contentful Paint) · Sites with a good score (%): 80.880.8Loading (Largest Contentf…Responding to taps (Interaction to Next Paint) · Sites with a good score (%): 89.489.4Responding to taps (Inter…Layout stability (Cumulative Layout Shift) · Sites with a good score (%): 90.290.2Layout stability (Cumulat…Sites with a good score (%)Test (phones, September 2026)
Test (phones, September 2026) Sites with a good score (%) Sites assessed
Loading (Largest Contentful Paint) 80.8 491,119
Responding to taps (Interaction to Next Paint) 89.4 380,716
Layout stability (Cumulative Layout Shift) 90.2 491,119

Fewer sites have enough data to be judged on taps, which is why that row has a smaller base. The order is still plain: jumping pages and laggy buttons are the rarer faults, and slow loading is the common one.

The biggest sites pass a little less often than the rest

Here the second half of the assumption breaks. Among UK-visited websites in the global top 10,000, 66.8% passed on phones, against 70.1% of the less popular sites below them. Popularity here is Google's worldwide ranking, so a site can be huge globally and modest in the UK, or the reverse.

The gap is a few points, and the data does not say why it exists. What it does show is that a famous name and a large engineering team are no guarantee of a quick, steady page on a mid-range Android phone. A small independent shop is competing on a level field more often than its owner might think.

The shop platform splits the field

The platform a shop runs on lines up with the result far more sharply than size does. Of the six online-shop platforms with at least 1,000 UK-visited websites, Wix eCommerce sites passed on phones most often, at more than nine in ten, and WooCommerce sites least often, at about six in ten.

020406080100Wix eCommerce · Sites passing (%): 91.891.8Wix eCommerceShopify · Sites passing (%): 87.987.9ShopifySquarespace Commerce · Sites passing (%): 87.287.2Squarespace CommerceBigCommerce · Sites passing (%): 72.172.1BigCommerceMagento · Sites passing (%): 60.560.5MagentoWooCommerce · Sites passing (%): 60.260.2WooCommerceSites passing (%)Shop platform (phones, September 2026)
Shop platform (phones, September 2026) Sites passing (%) UK-visited sites
Wix eCommerce 91.8 8,466
Shopify 87.9 37,733
Squarespace Commerce 87.2 7,885
BigCommerce 72.1 1,014
Magento 60.5 3,252
WooCommerce 60.2 23,421

Two cautions apply. The platform is identified from the site's home page in a separate public crawl of the same month, and sites missing from that crawl are left out. A platform's pass rate also reflects the kind of shops built on it: their size, their themes, the apps and add-ons bolted on. A low score describes the sites, and the platform may be only part of the reason. Magento and WooCommerce sit almost level at the bottom of the six, near-parity at this precision.

What this means for UK businesses

  • Passing is normal, so failing stands out. Most of the sites your customers visit clear Google's mark on phones. A slow shop is now the exception shoppers notice.
  • Start with loading. It is the test sites fail most often. The main image on your home page and product pages is usually the place to look first.
  • Size will not save you. The most popular sites do no better than the rest, so a small shop can match or beat much bigger rivals on speed.
  • Look hard at your set-up. Pass rates differ widely between shop platforms. If you are on a platform near the bottom of the table, your theme, apps and add-ons deserve an audit before the peak.
  • Check the phone version. These results are for phones. Test on a real, ordinary Android handset, not only on the office desktop.
Screenshot developersgooglecom

Google's own explanation of the three Core Web Vitals and how they are judged. Source: developers.google.com, captured 15 October 2026.

How we built it

All pass rates come from Google's Chrome UX Report for September 2026, covering websites visited in the UK with enough traffic to be included. The report is built from real page loads by Chrome users who share usage statistics, so it is a one-month snapshot of real visits from a self-selecting group of users, not a census of every visit. Phone data comes mainly from Chrome on Android; iPhone visits in Safari are not included. It covers the websites UK people use, whether or not they are UK-owned.

A site "passes Core Web Vitals" when, at the 75th percentile of its visits, Largest Contentful Paint is 2.5 seconds or less, Cumulative Layout Shift is 0.1 or less and Interaction to Next Paint is 200 milliseconds or less. Sites without enough tap data are judged on loading and layout alone, which is why the responsiveness test has a smaller base.

Popularity bands use the Chrome UX Report's own global rank buckets. The "less popular" group combines every site outside the global top 10,000. Platforms were identified by matching each site to the technology detected on its home page in the HTTP Archive mobile crawl of the same month; sites not in that crawl are excluded. The platform comparison is limited to online-shop platforms with at least 1,000 UK-visited websites, and a platform's pass rate reflects the mix of sites built on it as well as the platform. The difference between the top 10,000 and the rest, and the spread between the highest and lowest shop platforms, are statistically significant.

Year-on-year changes are not reported in the article because the set of sites with enough traffic to appear in the report changes from month to month, so part of any move reflects which sites are included.