Core Web Vitals Benchmark
Benchmark your website against up to four competitors using field data from real Chrome users. See who passes the Core Web Vitals, how far each site is behind the leader and how the numbers moved over the last 40 weeks. Below the tool you find pass rates by CMS, platform and industry. Free, no sign-up.
Benchmark against competitors with field data
The tool uses only the Chrome UX Report (CrUX): what real Chrome users experienced over the last 28 days. It is the same data Google uses for page experience and the same data you see in Search Console. A lab test measures one page load under fixed conditions, which makes it a poor way to compare website speed against competitors, because server location, cache state and plain chance move the result. Every value is the 75th percentile, meaning three out of four page loads were at least this fast.
Gap to the leader and gap to "good"
For each metric the comparison marks the best value and shows how far every other site is behind it, in milliseconds and in percent. Where a value is above the good threshold, it also tells you how much it has to drop. That number works as a target in a ticket or a proposal, because it is measurable and tells you when the job is done.
Next to the three Core Web Vitals the table shows First Contentful Paint and Time to First Byte. They do not count toward the assessment, but they often explain the difference: if a competitor's TTFB is half a second ahead, the LCP gap is usually about servers, CDN and caching rather than images.
LCP subparts show where the time goes
For pages whose largest element is an image, CrUX splits LCP into four parts: server response, the delay before the image starts loading, the image download and the render delay. The comparison puts these parts side by side for every site. If your image downloads as fast as your competitor's but the browser discovers it a second later, a preload or fetchpriority will do more than a smaller file format.
40 weeks of history and the trend
A single number does not tell you whether a site is getting faster or slower. The comparison loads the CrUX history with up to 40 weekly periods and draws LCP, INP and CLS for all sites in one chart. The trend compares the latest value with the value twelve weeks earlier, so a relaunch, a new consent banner or an extra tracking script shows up as a bend in the line.
Share, export and embed the result
Every comparison has a fixed address. Send the link to your team or a client and they see the current numbers without signing up. The CSV export contains every metric for your own analysis. For your own site there is a status badge you can embed on your website or in a slide deck.
When a site has no field data
The Chrome UX Report only includes sites with enough Chrome traffic. If a URL has no data, the comparison falls back to the site's origin. If the origin has none either, it tries the www variant, since CrUX treats both as separate sites. If that also fails, the site is too small for field data, and a lab test such as PerfDelta's free Core Web Vitals test is the next step.
This benchmark shows how often websites on the major content management systems and site builders pass the Core Web Vitals, based on field data from real Chrome users. It answers a practical question: what counts as a normal, a strong and a weak result, and where does a given platform stand? Every figure comes from a public source with its date, not from a lab test, so you can check it and place your own site against it.
Where the numbers come from
All values are based on the Chrome UX Report (CrUX), Google's public dataset of real-user experience, as processed by the HTTP Archive and its Core Web Vitals Technology Report. A site passes when all three metrics are good at the 75th percentile of its real page loads: Largest Contentful Paint up to 2.5 seconds, Interaction to Next Paint up to 200 milliseconds and Cumulative Layout Shift up to 0.1. All three have to pass at once. A fast INP does not make up for a slow LCP.
Pass rates are origin-weighted: every domain counts once, regardless of its traffic. This is the convention Google uses for CrUX aggregates. Origins without enough traffic for field data are not included. The platform figures below are as of November 2025, the global figures as of the 2025 Web Almanac, and the latest CrUX snapshot as of May 2026. The figures and dates change with each update.
Global benchmark: about half of all sites pass
Across the web, around 48 percent of mobile origins and 56 percent of desktop origins pass all three Core Web Vitals (CrUX via HTTP Archive Web Almanac, 2025). In the May 2026 CrUX release, 55.9 percent of roughly 18.4 million origins pass all three, with 68.6 percent good LCP, 81.3 percent good CLS and 86.6 percent good INP (CrUX, May 2026).
On mobile, loading performance is the bottleneck. About 62 percent of mobile origins pass LCP, against 77 percent for INP and 81 percent for CLS. If you want to raise your pass rate, Largest Contentful Paint is usually the place to start. The mobile pass rate has risen steadily, from about 32 percent in 2021 to 44 percent in 2024 and 48 percent in 2025.
Mobile is about eight points harder than desktop
The gap between mobile and desktop has stayed at roughly eight percentage points. Mobile combines weaker processors with slower connections, and it is also the view Google relies on for ranking. A site that passes on mobile almost always passes on desktop as well; the reverse is not true.
Among the three metrics, Cumulative Layout Shift is the easiest to keep in the good range. Interaction to Next Paint, which replaced First Input Delay in March 2024, is often the hardest for highly interactive pages. Largest Contentful Paint is the most common reason a mobile page fails. Some platforms lead clearly on individual metrics: around 95 percent of Shopify sites have a good Time to First Byte, the base every other loading metric builds on (HostingStep, 2025).
Pass rates by CMS and platform
The platform has a strong influence on how likely a site is to pass. The share of origins passing all three metrics differs by almost 40 percentage points between the major systems (Core Web Vitals Technology Report, November 2025):
| Platform | Share of origins passing all three Core Web Vitals |
|---|---|
| Duda | 84.9 % |
| Wix | 74.9 % |
| Squarespace | 70.4 % |
| WordPress | 46.3 % |
Source: Core Web Vitals Technology Report based on CrUX field data, November 2025. Origin-weighted, all three metrics at once.
The report tracks many more systems, including Drupal, Joomla, TYPO3 and Shopify. WordPress, by far the most widely used CMS, has the lowest pass rate of the large platforms and pulls the global average down. The current figure for any system can be looked up directly in the Technology Report.
Managed platforms lead because they ship performance by default
The platform sets the floor, not the ceiling. Two sites on the same CMS can score very differently. A WordPress site with a lean theme, good hosting and disciplined handling of images and scripts passes comfortably, while a site on a fast platform can still fail with too many third-party scripts.
Duda, Wix and Squarespace come out on top because their customers inherit performance work for free: lean templates, automatic image optimization and delivery through a CDN are standard there. On WordPress, theme, plugins and hosting decide the result, which is why the spread is so wide.
Hosting explains much of the WordPress spread
Within one platform, hosting is often the largest single factor. Analyses across millions of WordPress sites show that a small set of decisions explains most of the variation: the quality of the host, especially whether edge caching is included, and a handful of plugins for caching, image optimization and handling render-blocking scripts. A capable host with built-in caching improves server response time, and with it LCP, more than most individual front-end changes. A site on a platform with a low average can therefore still end up well above that platform's median.
E-commerce: heavy pages and many scripts
Online shops are a case of their own. Large catalogues, many images, variant logic and a long list of marketing and tracking scripts weigh mainly on LCP and interactivity. Magento sits at the lower end of the large systems with about 40 percent desktop pass rate, driven by large DOMs and often self-hosted infrastructure (HostingStep, 2025). Shopify reaches very good TTFB values through its hosted infrastructure but often loses the overall assessment to themes and apps that load a lot of JavaScript. For shops in particular, the number of installed apps and scripts decides the result more than the platform does.
Pass rates by industry
The industry matters too, because it shapes what a typical page contains. Technology sites have the highest pass rate at around 65 percent, news and media sites the lowest at around 39 percent (HTTP Archive, 2024/2025). News sites carry many ads, trackers and late-loading elements that hurt layout stability and interactivity, while technology sites tend to be leaner. PerfDelta publishes its own figures from the sites tested with the tool in the Core Web Vitals benchmarks by industry, once an industry has enough measured domains.
Why pass rates keep rising
The global pass rate has climbed by roughly four percentage points a year since 2021. Browsers and frameworks ship better defaults such as native lazy loading and modern image formats, managed platforms roll out performance work to all their customers at once, and the role of Core Web Vitals in rankings has prompted investment. At the same time the spread between the best and weakest platforms is widening, because managed systems move faster than the large and varied WordPress base.
Country and network quality
Where your visitors are also shapes the numbers, because network coverage and typical devices differ between countries. South Korea consistently leads across studies, while regions with slower mobile networks show lower pass rates. For a fair comparison, look at your target region rather than the global average alone.
How to read the benchmark
A benchmark is easy to misread when the wrong numbers are compared:
- Field, not lab. The benchmark is based on field data at the 75th percentile. Comparing it with a single Lighthouse score mixes two different things.
- Every domain counts once. Because the values are origin-weighted, a large and a small site on the same platform weigh the same in the average.
- A distribution, not a promise. 46 percent for WordPress does not mean each WordPress site passes 46 percent of the time. It means 46 percent of WordPress origins pass all three metrics. Your own site can sit far above or below that.
Placing your own site
A benchmark becomes useful once you put your own numbers next to it. What counts is the field data of your site from the Chrome UX Report, not a single Lighthouse run: Lighthouse is a diagnostic tool, while Google uses field data for ranking. If you are above your platform's median, you are doing a lot right. If you are below it, there is most likely a concrete, fixable problem, usually with LCP.
PerfDelta measures several URLs of your site on mobile and desktop, shows field and lab data side by side and compares your domain with competitors. How each value is collected is described in the PerfDelta methodology.
Improving your pass rate
The levers are the same on every platform; only the implementation differs. The Core Web Vitals guide covers all three metrics, and each one has its own guide: Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift. Because LCP is the most common bottleneck on mobile, start with a fast server response time, remove render-blocking resources, and optimize images with modern formats and correct sizing.
PerfDelta measurements by industry
Alongside the public figures above, PerfDelta breaks down its own measurements by industry. Each industry has a page with pass rate, typical values and a platform comparison:
- Online retail: 1110 measured domains
- Media and publishing: 570 measured domains
- Finance and insurance: 527 measured domains
- Education: 363 measured domains
- Public sector and non-profit: 361 measured domains
- Entertainment and leisure: 339 measured domains
- Software and SaaS: 317 measured domains
- Health and medical: 282 measured domains
- Food and restaurants: 273 measured domains
- Travel and hospitality: 247 measured domains
- Telecommunications: 194 measured domains
- Real estate: 192 measured domains
- Automotive and mobility: 190 measured domains
- Manufacturing and industry: 175 measured domains
- Professional services: 167 measured domains
- Energy and utilities: 165 measured domains
- Logistics and transport: 151 measured domains
- Trades and local services: 98 measured domains
All industries at a glance: Core Web Vitals by industry.
Frequently asked questions
Is the Core Web Vitals benchmark free?
Yes. It is free and needs no account and no email address. You can benchmark up to five websites or URLs at once.
Where does the data come from?
From Google's Chrome UX Report. It aggregates what real Chrome users experienced over the last 28 days and is updated daily. PerfDelta caches each response for twelve hours.
Why do the numbers differ from PageSpeed Insights?
The field data at the top of PageSpeed Insights comes from the same source and should match. The performance score below it is a lab test with a simulated device and network, which is why it differs. This comparison shows field data only.
Can I compare single pages instead of whole domains?
Yes. Enter the full URL including the path, for example a competitor's product page. If that URL has too few visitors, the comparison shows the origin's data and labels it as such.
How is the ranking calculated?
Passing the Core Web Vitals comes first, then the number of good metrics. Ties are broken by the average of LCP, INP and CLS, each divided by its good threshold.
Can I track competitors over time?
With a PerfDelta account on the Solo plan or higher, your site and your competitors are checked every day, and you get an alert when a metric gets worse.
Which CMS has the best Core Web Vitals?
Among the large platforms, managed systems lead. In the Core Web Vitals Technology Report of November 2025, 84.9 percent of Duda origins, 74.9 percent of Wix origins and 70.4 percent of Squarespace origins pass all three metrics, against 46.3 percent for WordPress.
Why does WordPress score lower?
WordPress itself is not slow. The wide range of themes, plugins and hosting pulls the average down. Managed platforms include performance work by default; on WordPress you have to set it up yourself. A well-built WordPress site passes comfortably.
What is a good Core Web Vitals result?
Passing all three metrics at the 75th percentile: LCP up to 2.5 seconds, INP up to 200 milliseconds and CLS up to 0.1. Since only about half of all sites pass all three, a passed assessment already puts you ahead of a large part of the web.
Does the benchmark use lab or field data?
Field data. Google assesses Core Web Vitals from the Chrome UX Report, meaning real users over a 28-day window at the 75th percentile. Lighthouse provides lab values for diagnosis that do not feed into rankings directly.
How do I see where my site stands against the benchmark?
Check your own URLs with a tool that shows CrUX field data for mobile and desktop, and compare the result with your platform's pass rate. That shows which of the three metrics causes the gap.
How much does hosting matter for Core Web Vitals?
Within one platform, hosting is often the largest single factor. A host with built-in edge caching lowers server response time and therefore improves LCP, often more than any single front-end change.
Does the benchmark apply to small websites?
Platform pass rates are origin-weighted, so small and large sites count the same. Sites with too little traffic for field data are not in CrUX, however, and stay outside the benchmark until they have enough real users.
More free tools
PerfDelta is built by Christian Schreiber, who has been working on SEO and web performance for more than 15 years.
Thresholds and sources
Sources:
All sources retrieved on October 9, 2026. Thresholds and documentation change; the linked page is authoritative.
A free account saves your tests and shows how your values change from one test to the next. Daily monitoring with field data and alerts starts with the Solo plan.