--- title: Website Speed Statistics 2026 description: "In 2025, 48% of mobile sites and 56% of desktop sites passed Core Web Vitals. Sourced 2026 stats on load time, page weight, and sales impact." url: "https://hostingguider.com/blog/website-speed-statistics/" canonical_url: "https://hostingguider.com/blog/website-speed-statistics/" date: "2026-10-21T01:37:00+00:00" modified: "2026-10-10T07:09:06+00:00" author: Shaddam Hossain type: article site: HostingGuider language: en-US --- # Website Speed Statistics 2026 Less than half of mobile websites pass Core Web Vitals. The 2025 Web Almanac found 48% of mobile sites and 56% of desktop sites had good scores in July 2025. Speed also moves money. A Deloitte study of 37 brands linked a 0.1 second mobile speed gain to 8.4% more retail conversions. Every number below links to the original study, with its year and sample. They add hard data to the [website speed and server performance coverage](https://hostingguider.com/topics/website-performance/) on HostingGuider. Table of Contents ## Key website speed statistics The headline figures come from the HTTP Archive Web Almanac 2025, Deloitte’s original report, and Google’s web.dev case studies. - **48% of mobile websites** and **56% of desktop websites** had good Core Web Vitals in 2025 (CrUX, July 2025, per origin; [Web Almanac 2025 Performance](https://almanac.httparchive.org/en/2025/performance?utm_source=hostingguider.com)). - **44% of mobile websites** had a good Time to First Byte in 2025, and 17% a poor one (same chapter, CrUX July 2025). - **62% of mobile websites** and **74% of desktop websites** had a good Largest Contentful Paint in 2025 (same chapter). - **77% of mobile websites** had good Interaction to Next Paint in 2025, up from 74% in 2024 (same chapter). - **The median mobile home page weighed 2,559 KB**, against 2,862 KB on desktop, in the July 2025 crawl ([Web Almanac 2025 Page Weight](https://almanac.httparchive.org/en/2025/page-weight?utm_source=hostingguider.com)). - **A 0.1 second mobile speed gain went with 8.4% more retail conversions** and 10.1% more travel conversions. That’s 37 brands, about 30 million sessions, 2019 data ([Deloitte report, 2020](https://www.deloitte.com/content/dam/assets-zone2/ie/en/docs/services/consulting/2023/Milliseconds_Make_Millions_report.pdf?utm_source=hostingguider.com)). - **Vodafone saw 8% more sales** after a 31% LCP improvement in an A/B test, a single-company result ([web.dev, 2021](https://web.dev/case-studies/vodafone?utm_source=hostingguider.com)). **How to cite these statistics:** cite the original source linked on each line, with its year. If you used this summary, add “compiled by HostingGuider, updated October 4, 2026. ## Core Web Vitals pass rates About half of mobile websites and just over half of desktop websites pass all three Core Web Vitals, and mobile has improved faster since 2023. Core Web Vitals are Google’s three field metrics for loading, responsiveness, and visual stability. A site passes when it meets all three “goodthresholds. They are defined in the guide to [what LCP, INP, and CLS measure](https://hostingguider.com/blog/core-web-vitals/). The Performance chapter uses Chrome UX Report (CrUX) data. CrUX is Google’s public dataset of real Chrome user experiences, measured per website (origin). | Year | Mobile sites with good CWV | Desktop sites with good CWV | | --- | --- | --- | | 2021 | 32% | 41% | | 2022 | 31% | 44% | | 2023 | 36% | 48% | | 2024 | 44% | 55% | | 2025 | 48% | 56% | *Source: Web Almanac 2025, Performance chapter, CrUX data. The 2025 row uses July 2025 data.* Mobile gained 12 points between 2023 and 2025, while desktop barely moved, from 55% to 56% in the last year. **Two more findings from the same chapter:** - **Popularity matters, but not in a straight line.** On mobile, 51% of the top 1,000 sites pass, dropping to 37% for the top 100,000, then rising to 48% overall. - **Inner pages do better than home pages.** On mobile, 45% of home pages pass, against 56% of secondary pages (page-level CrUX data). **What this means for your money:** if your mobile site fails, you’re in the majority. That’s no comfort. A fast site still beats about half the web. ## Page weight The median home page in July 2025 weighed 2,559 KB on mobile and 2,862 KB on desktop, and pages keep getting heavier. The Page Weight chapter measured home and inner pages in the July 2025 HTTP Archive crawl. On growth, it says “the median mobile home page is 2.6 MB, growing 8.4% from 2.4 MB in 2024.Here is what makes up the median mobile home page, from the same chapter: | Resource | Median size, mobile home page | | --- | --- | | Images | 911 KB | | JavaScript | 632 KB | | Fonts | 122 KB | | CSS | 77 KB | | HTML | 22 KB | Images and JavaScript make up most of the weight. Neither comes from your host. A faster server can’t shrink a 900 KB hero image. What page weight means for a plan’s monthly transfer allowance is worked out in a separate guide on how much bandwidth and storage you need. ## Load timing metrics: LCP, INP, CLS and TTFB Most websites now pass the responsiveness and layout tests, while loading speed and server response remain the weakest areas, especially on mobile. All figures in this section come from the Web Almanac 2025 Performance chapter, using CrUX data for July 2025. Shares are of websites with a “goodrating. | Metric (one-line meaning) | Good threshold | Mobile good | Desktop good | | --- | --- | --- | --- | | LCP: when the main content appears | 2.5 s or less | 62% | 74% | | INP: how fast the page reacts to taps and clicks | 200 ms or less | 77% | 97% | | CLS: how much the layout jumps | 0.1 or less | 81% | 72% | | TTFB: when the first byte arrives from the server | 0.8 s or less | 44% | 55% | | FCP: when anything first appears | 1.8 s or less | 55% | 70% | **A few details stand out:** - **Poor LCP is nearly twice as common on mobile.** 13% of mobile sites have poor LCP, against 7% on desktop. - **TTFB barely moved.** Good TTFB rose from 42% to 44% on mobile and from 54% to 55% on desktop since 2024. JavaScript is getting heavier at the same time. The chapter’s lab data shows median mobile Total Blocking Time rose to 1,916 milliseconds in 2025, from 1,209 in 2024. Lab and field numbers can disagree, as the guide to [lab scores versus field data](https://hostingguider.com/blog/pagespeed-score-vs-real-user-experience/) explains. ## Speed and business outcomes The strongest public evidence links small speed gains to higher conversions, but most of it comes from single companies or one study, so treat it as direction, not a promise. **Multi-brand study.** Deloitte’s Milliseconds Make Millions report studied mobile data from 37 brands in Europe and the US. It covered about 30 million sessions over 4 weeks in late 2019. **It found that a 0.1 second improvement went with these changes:** - **Retail:** “An 8.4% increase in conversionsand a 9.2% higher average order value. - **Travel:** “A 10.1% increase in conversionsand a 1.9% higher average order value. - **Lead generation:** conversion rates on mobile “decreased by almost 2%, even though users moved further through forms. The report says the 0.1 second figure “was the cumulative impact of all four speed metricsit tracked. The study was observational. It shows correlation, not proof that speed caused the change. **Single-company case studies (Google web.dev).** Each of these is one company’s result. Your site may see more, less, or nothing. - **Vodafone (2021):** an A/B test with a 31% better LCP produced 8% more sales, per the case study linked above. - **Renault (2021):** across over 10 million visits in 33 countries, a 1-second LCP improvement “can lead to a 14 percentage points (ppt) decrease in bounce rate and 13% increase in conversions.Source: [Renault web.dev case study](https://web.dev/case-studies/renault?utm_source=hostingguider.com). - **The Economic Times (2023):** INP fell from over 1,000 ms to 257 ms. The team reports a 50% lower bounce rate and 43% more pageviews on topic pages. Source: [Economic Times web.dev case study](https://web.dev/case-studies/economic-times-inp?utm_source=hostingguider.com). ## Famous speed statistics, checked Three widely repeated speed claims are older and less clearly sourced than most listicles suggest, so here is what each original actually says. - **“53% of mobile users leave after 3 seconds.** A 2016 Google [DoubleClick mobile speed toolkit](https://www.thinkwithgoogle.com/_qs/documents/3702/9e6a3_DoubleClick_Mobile_Web_Speed_Toolkit_10_17_2016.pdf?utm_source=hostingguider.com) says “53% of mobile site visits are abandoned if a page takes longer than 3 seconds to load.It cites a September 2016 DoubleClick study we could not find published. - **“A 1-second delay costs 7% of conversions.** Akamai’s [April 2017 retail performance release](https://www.ir.akamai.com/news-releases/news-release-details/akamai-online-retail-performance-report-milliseconds-are?utm_source=hostingguider.com) says “A 100-millisecond delay in website load time can hurt conversion rates by 7 percent.That’s 100 ms, not 1 second. Its sample was about 10 billion visits to retail sites over one month, and the full report link points to a retired SOASTA address we could not open. - **“Amazon loses 1% of sales per 100 ms.** We found no public Amazon study or data behind this line, so we don’t use it. Our advice: use the 2025 pass rates and the Deloitte report when you need a number. They have a date, a sample, and a document you can open. ## Hosting-related speed statistics Hosting shows up most clearly in Time to First Byte, the one metric in the timing table that measures the server and the network. The Web Almanac describes TTFB as “the network and server overhead. Since LCP includes TTFB, a slow server delays everything after it, and the server-side remedies are in the guide to [server-level TTFB fixes](https://hostingguider.com/blog/fix-ttfb-hosting-level/). **Our own data (4 accounts we bought, not a market sample).** Every account runs the same WordPress test site, hostingguider.online. These are single tests from our reviews, not averages over months. - **PageSpeed (lab):** mobile scores were 99 on Hostinger Business, Hostinger Cloud Startup and Namecheap Stellar Plus, and 100 on Kinsta WP 2 (first 3 days of that account, August 2026). - **Distance shows in TTFB.** On the [Namecheap Stellar Plus account we tested](https://hostingguider.com/reviews/namecheap-hosting/) in August 2026, with a USA data center, TTFB was 126 ms from San Francisco and 707 ms from Bangalore. Its monitoring history is still short. The full city-by-city figures for all four accounts sit in a separate set of TTFB benchmarks by city and hosting type. Market-wide figures, like provider share, belong to the [hosting market statistics](https://hostingguider.com/blog/web-hosting-statistics/). ## What the numbers mean for choosing hosting Hosting fixes the server part of speed, so pick a host near your visitors with fast server response, then fix images and scripts yourself. - **Server location first.** The Namecheap gap above came from distance alone. Pick a data center near most of your visitors, because [how data center distance adds delay](https://hostingguider.com/blog/how-server-location-affects-web-hosting-performance/) applies to every plan. - **Test TTFB, not the PageSpeed score.** Near-perfect lab scores on all four accounts said nothing about how far-away visitors would fare. - **Don’t expect a host to fix page weight.** Images and JavaScript make up most of a median page. That’s theme, plugin, and content work. - **WordPress sites carry extra weight.** Plugin and theme choices add weight that no plan removes. Platform numbers are in the [WordPress usage figures](https://hostingguider.com/blog/wordpress-statistics/). If your visitors are spread across regions and you want caching handled for you, look at [managed WordPress plans with server caching](https://hostingguider.com/best/best-managed-wordpress-hosting/). My verdict: most small sites don’t need that yet. Upgrade when shared hosting can’t keep TTFB in Google’s “goodrange for your main audience. The Website Speed Checklist sets the order to fix a slow site before you spend on a bigger plan. ## Frequently asked questions ### What percentage of websites pass Core Web Vitals? About half. The Web Almanac 2025 puts it at 48% on mobile and 56% on desktop, using Chrome UX Report data for July 2025. ### Which Core Web Vital do most sites fail? On mobile, LCP. Only 62% of mobile websites had a good LCP in 2025, the lowest of the three. On desktop, CLS was the weakest at 72%. ### What is the average website load time? There’s no reliable single average. Load time changes with device, network and page, so the Web Almanac reports the share of sites meeting each threshold instead. ### How much does a 1-second delay cost in conversions? No current, open study gives one number. The closest is Deloitte’s retail finding for a 0.1 second gain. Akamai’s widely quoted 2017 “7%figure was about a 100 ms delay. ### Is the “53% leave after 3 secondsstatistic real? It traces to a 2016 Google DoubleClick toolkit whose underlying study we could not find, so treat it as dated rather than current. ### How big is the average web page in 2025? A median mobile home page was about 2.6 MB in July 2025, and images were its largest part at 911 KB, per the Web Almanac. ## How we researched this We went to each original publisher between October 3 and 4, 2026. That meant three Web Almanac 2025 chapters, Deloitte’s report PDF, three web.dev case studies, and Akamai’s 2017 release. The Almanac used the July 2025 HTTP Archive crawl of 16,213,084 websites and the July 2025 CrUX release, per the [Web Almanac 2025 methodology](https://almanac.httparchive.org/en/2025/methodology?utm_source=hostingguider.com). We quote figures as published, without rounding, except where a chapter rounds its own. First-party figures come from accounts we bought between December 2024 and August 2026 and tested under the same [hands-on hosting test rules](https://hostingguider.com/methodology/). **Update cadence:** we recheck these figures when each new Web Almanac is published, and review the case studies every 6 months. **Change log:** - October 4, 2026: Web Almanac 2025 Performance and Page Weight figures rechecked on the live chapters. - October 3, 2026: First published with Web Almanac 2025 data, the Deloitte report, three web.dev case studies, and first-party data from 4 accounts.