37 Page Speed and Core Web Vitals Statistics for 2026

By Christoph Olivier | Published 2026-07-26 | Last updated 2026-07-26

This briefing collects verified statistics on web performance and Google’s Core Web Vitals (CWV). Every figure below carries a number, a timeframe, a geography where the source specifies one, and a citation. The anchor data set is the HTTP Archive Web Almanac, an annual analysis that combines a lab crawl of millions of websites with field data from the Chrome UX Report (CrUX). Speed-to-conversion figures come from Google, Deloitte, Akamai, and Portent. Those conversion studies are correlational: they observe relationships between speed and behavior, not controlled causal experiments.

Executive summary

• 48% of mobile origins and 56% of desktop origins passed all three Core Web Vitals in 2025, per the HTTP Archive Web Almanac using CrUX field data from July 2025.
• Largest Contentful Paint (LCP) is the metric most sites fail on mobile: only 62% of mobile origins scored “good” in 2025.
• Interaction to Next Paint (INP) replaced First Input Delay as a Core Web Vital in March 2024; 77% of mobile origins passed INP in 2025.
• Mobile trails desktop by roughly 8 percentage points on overall CWV pass rate, a gap the Web Almanac attributes to weaker mobile hardware and networks.
• Mobile CWV pass rates rose from 32% (2021) to 48% (2025), but the Web Almanac notes annual gains are slowing.
• A 0.1-second improvement in mobile load time was associated with an 8.4% lift in retail conversion rate in the Deloitte and Google “Milliseconds Make Millions” study (2019, correlational).
• Portent found e-commerce sites loading in 1 second converted at 2.5 times the rate of sites loading in 5 seconds (2022, correlational).
• Google and SOASTA found the probability of a mobile bounce rose 32% as load time went from 1 to 3 seconds (2017).

Key findings

1. 48% of mobile origins passed all three Core Web Vitals in 2025 (global, CrUX July 2025). Source: HTTP Archive Web Almanac, Performance, 2025.
2. 56% of desktop origins passed all three Core Web Vitals in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
3. 62% of mobile origins had “good” LCP (under 2.5 seconds) in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
4. 77% of mobile origins had “good” INP (under 200 ms) in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
5. 81% of mobile origins had “good” CLS (0.1 or lower) in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
6. 74% of desktop origins had “good” LCP in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
7. 97% of desktop origins had “good” INP in 2025 (global). Source: HTTP Archive Web Almanac, Performance, 2025.
8. Mobile CWV pass rate rose from 44% in 2024 to 48% in 2025, a 4-point gain (global). Source: HTTP Archive Web Almanac, Performance, 2025.
9. Mobile CWV pass rates climbed from 32% (2021) to 36% (2023) to 48% (2025), while desktop rose from 41% to 56% over the same span (global). Source: HTTP Archive Web Almanac, Performance, 2025.
10. A 0.1-second cut in mobile load time was associated with an 8.4% lift in retail conversion rate and a 9.2% lift in average order value (Europe and US, 2019, correlational). Source: Deloitte and Google, “Milliseconds Make Millions,” 2019.
11. The same study associated a 0.1-second improvement with a 10.1% lift in travel conversion rate (Europe and US, 2019, correlational). Source: Deloitte and Google, “Milliseconds Make Millions,” 2019.
12. E-commerce sites loading in 1 second converted at 3.05%, versus 1.68% at 2 seconds and 0.67% at 4 seconds (2022, correlational). Source: Portent, “Site Speed Is (Still) Impacting Your Conversion Rate,” 2022.
13. A 100-millisecond delay in mobile load time was associated with a 7% drop in conversion rate (top online retailers, 2017, correlational). Source: Akamai, State of Online Retail Performance, Spring 2017.
14. A 2-second delay in web page load time was associated with a 103% increase in bounce rate (top online retailers, 2017, correlational). Source: Akamai, State of Online Retail Performance, Spring 2017.
15. The probability of a mobile visitor bouncing rose 32% as load time went from 1 to 3 seconds, and 123% from 1 to 10 seconds (2017). Source: Google and SOASTA, “New Industry Benchmarks for Mobile Page Speed,” Think with Google, 2017.

What Core Web Vitals measure

Core Web Vitals are three field metrics Google uses to grade real-user experience. LCP measures loading: the time until the largest content element renders, with “good” under 2.5 seconds. INP measures responsiveness: the latency of user interactions, with “good” under 200 milliseconds. CLS measures visual stability: unexpected layout movement, with “good” at 0.1 or lower. An origin “passes” only when at least 75% of visits meet the “good” threshold on all three metrics. INP replaced First Input Delay as an official Core Web Vital on 12 March 2024, which reset interactivity measurement across the web.

Section 1: Core Web Vitals pass rates, mobile vs desktop

The HTTP Archive Web Almanac is the primary reference for CWV adoption. Its Performance chapter reads CrUX field data, which is collected from opted-in Chrome users visiting millions of origins. The 2025 edition uses CrUX data from July 2025.

Source: HTTP Archive Web Almanac, Performance, 2025 (CrUX, July 2025, global).

Two patterns stand out. First, LCP is the choke point on mobile: at 62%, it is the metric fewest mobile sites pass, so loading speed is where most CWV failures originate. Second, CLS is the one metric where mobile (81%) outperforms desktop (72%). The Web Almanac attributes the desktop CLS gap to larger hero media and more dynamic above-the-fold content on wide viewports.

Section 2: The mobile-desktop performance gap

Desktop origins have led mobile on overall CWV pass rate every year the Web Almanac has measured it. In 2025 the gap was 8 points (56% desktop vs 48% mobile). The Web Almanac attributes the persistent gap to constrained mobile CPU, variable cellular networks, limited device memory, and heavier third-party scripts on mobile pages. The gap has been directional and stable rather than closing quickly.

Source: HTTP Archive Web Almanac, Performance, 2025 (global). The 2023 desktop figure is interpolated from the reported 41% (2021) to 56% (2025) trajectory and is marked approximate.

Section 3: Year-over-year improvement and its slowdown

Mobile CWV pass rate gained 4 points from 2024 (44%) to 2025 (48%). Desktop gained 1 point (55% to 56%). Within the mobile metrics, INP rose from 74% to 77% and First Contentful Paint rose from 51% to 55% between 2024 and 2025. The Web Almanac frames the deceleration plainly: the large early gains from fixing common issues have mostly been captured, so further improvement is harder to win.

Section 4: Aggregate CrUX snapshot, 2026

Beyond the annual Web Almanac, CrUX publishes a monthly aggregate. A published tracker citing the May 2026 CrUX release reports 55.9% of all tracked origins passing all three Core Web Vitals, with 68.6% good LCP, 81.3% good CLS, and 86.6% good INP across 18,445,974 origins. This aggregate blends mobile and desktop into one origin-level figure and is not split by device, so it is not directly comparable to the device-split Web Almanac numbers above. It is included as a secondary reading of CrUX and flagged as such.

Source: digitalapplied, “Core Web Vitals Benchmarks 2026,” citing Chrome UX Report, May 2026 release (Tier 2, reporting CrUX field data).

Section 5: Load time and conversion (correlational)

Multiple studies link faster pages to higher conversion. None are controlled experiments, so they show association, not proof of cause. They are consistent in direction: slower pages correlate with fewer conversions and more bounces.

Deloitte and Google, “Milliseconds Make Millions” (2019)

Google commissioned data firm 55 and Deloitte to analyze more than 30 million mobile user sessions across 37 retail, travel, luxury, and lead-generation brands in Europe and the US over roughly 30 days at the end of 2019. Associated with a 0.1-second improvement in mobile load time: retail conversion rate up 8.4% and average order value up 9.2%; travel conversion rate up 10.1%.

Source: Deloitte and Google, “Milliseconds Make Millions,” 2019 (correlational).

Portent site speed study (2022)

Portent analyzed more than 100 million page views across 20 sites (6 B2C e-commerce, 14 B2B lead generation) covering 5.6 million sessions in a 30-day snapshot updated April 2022.

Portent reported that a site loading in 1 second had an e-commerce conversion rate 2.5 times higher than a site loading in 5 seconds, and a B2B conversion rate 3 times higher.

Source: Portent, “Site Speed Is (Still) Impacting Your Conversion Rate,” 2022 (correlational).

Akamai State of Online Retail Performance (Spring 2017)

SOASTA, now part of Akamai, analyzed roughly one month of anonymous data from top online retailers, on the order of 10 billion user visits. A 100-millisecond delay in load time was associated with a 7% drop in mobile conversion rate. A 2-second delay was associated with a 103% increase in bounce rate.

Source: Akamai, State of Online Retail Performance, Spring 2017 (correlational).

Google and SOASTA mobile benchmarks (2017)

Using a machine-learning model reported at 90% prediction accuracy, Google and SOASTA modeled bounce probability against mobile load time. As load time went from 1 to 3 seconds, bounce probability rose 32%; from 1 to 5 seconds, 90%; from 1 to 6 seconds, 106%; from 1 to 10 seconds, 123%. Separately, Google reported that 53% of mobile visits are abandoned if a page takes longer than 3 seconds to load.

Source: Google and SOASTA, “New Industry Benchmarks for Mobile Page Speed,” Think with Google, 2017.

Original synthesis

Synthesis 1: LCP is the single highest-impact fix for mobile CWV

Logic: among mobile CWV metrics in 2025, LCP has the lowest pass rate (62%) versus INP (77%) and CLS (81%). Because an origin must pass all three to pass overall, the metric with the lowest pass rate is the binding constraint. Improving LCP therefore moves the most origins across the overall threshold. Inputs: HTTP Archive Web Almanac, Performance, 2025 (all three metric pass rates). Limitation: this is a cross-sectional inference from aggregate pass rates, not a study of individual sites; some sites fail on two or three metrics at once, so LCP fixes alone will not lift every failing origin.

Synthesis 2: The mobile-desktop gap has narrowed from 9 to 8 points over four years

Logic: the overall CWV gap was 9 points in 2021 (41% desktop minus 32% mobile) and 8 points in 2025 (56% minus 48%). Mobile gained 16 points over the span while desktop gained 15, so mobile closed the gap by roughly 1 point in four years. Inputs: HTTP Archive Web Almanac, Performance, 2025 (historical pass rates). Limitation: year-to-year definitions shifted when INP replaced FID in 2024, so the pre-2024 and post-2024 series are not perfectly comparable; the 1-point convergence is small enough to fall within measurement noise.

Synthesis 3: Speed-conversion studies agree in direction across a decade and three vendors

Logic: Akamai (2017), Google and SOASTA (2017), Deloitte and Google (2019), and Portent (2022) span four independent data sets, different sectors, and different metrics, yet all report the same direction: faster pages correlate with higher conversion and lower bounce. Convergent findings from independent sources raise confidence in the direction of the effect even when each study alone is correlational. Inputs: the four studies cited above. Limitation: the magnitude figures are not comparable across studies because sample sizes, sectors, geographies, and load-time definitions differ; none isolate speed from confounders such as site quality, so none establish causation.

Methodology

Source selection: primary field and crawl data was prioritized (HTTP Archive Web Almanac and the underlying Chrome UX Report). Vendor studies (Deloitte and Google, Akamai, Portent) were included only for the speed-conversion relationship, where no neutral primary equivalent exists, and each is labeled correlational.

Inclusion rules: a statistic was included only if it carried an identifiable source, a year, and a stated population. Figures were excluded when the original methodology could not be confirmed or when only third-hand blog restatements were available without a traceable origin.

Conflict handling: where sources disagreed on the same figure, the value from the source closest to the raw data was used and the discrepancy noted. The 2026 monthly CrUX aggregate (55.9% passing) is reported by a secondary tracker and is labeled Tier 2; it is not blended with the device-split Web Almanac figures because it is not device-split.

Limitations: CrUX covers only opted-in Chrome users and origins with sufficient traffic, so very low-traffic sites are underrepresented. The conversion studies are correlational and several are several years old. Load-time definitions differ across studies. Last updated 2026-07-26.

Source quality ranking

Tier 1 (primary field or crawl data): HTTP Archive Web Almanac (Performance, 2025), which combines a lab crawl of millions of sites with CrUX field data; Google Chrome UX Report (CrUX) field data; Google and SOASTA “New Industry Benchmarks for Mobile Page Speed” (Think with Google, 2017), based on modeled data from real mobile sessions.

Tier 2 (credible vendor research and public trackers citing primary data): Deloitte and Google “Milliseconds Make Millions” (2019); Akamai State of Online Retail Performance (Spring 2017); Portent site speed study (2022); digitalapplied 2026 benchmark tracker reporting the May 2026 CrUX release.

Tier 3 (reputable journalism and expert commentary): The Drum and Marketing Dive coverage of the above studies, used only for context and not as a figure source.

Most quotable statistics

• “Only 48% of mobile sites passed all three Core Web Vitals in 2025.” (HTTP Archive Web Almanac, Performance, 2025)
• “LCP is the metric most mobile sites fail: just 62% passed in 2025.” (HTTP Archive Web Almanac, Performance, 2025)
• “A 0.1-second mobile speed gain was associated with an 8.4% lift in retail conversion.” (Deloitte and Google, 2019)
• “E-commerce sites loading in 1 second converted 2.5 times better than sites loading in 5 seconds.” (Portent, 2022)
• “53% of mobile visits are abandoned when a page takes longer than 3 seconds to load.” (Google and SOASTA, Think with Google, 2017)

Data limitations

CrUX and the Web Almanac reflect opted-in Chrome users and traffic-qualified origins, not the entire web. The four speed-conversion studies are correlational and cannot prove that speed causes conversion; site quality and other confounders are not isolated. Akamai and Think with Google figures date to 2017 and describe that period’s web. Magnitudes are not comparable across the conversion studies because populations and load-time definitions differ. The 2026 CrUX aggregate is a secondary reading and is not split by device.

Downloadable dataset: recommended fields

A companion dataset for this asset should include: metric_name, device (mobile/desktop), year, geography, pass_rate_percent, threshold, data_source, methodology (lab/field/modeled), sample_size, correlational_flag, and source_url.

Press summary

Web performance improved again in 2025 but remains uneven. The HTTP Archive Web Almanac, drawing on Chrome UX Report field data from July 2025, found that 48% of mobile sites and 56% of desktop sites passed all three Core Web Vitals. Largest Contentful Paint is the weak point on mobile, where only 62% of sites scored “good,” making loading speed the main reason sites fail. Mobile pass rates have risen from 32% in 2021 to 48% in 2025, though annual gains are slowing. The business case for speed comes from four independent, correlational studies spanning 2017 to 2022. Deloitte and Google linked a 0.1-second mobile speed gain to an 8.4% retail conversion lift. Portent found sites loading in 1 second converted 2.5 times better than sites loading in 5 seconds. Google reported 53% of mobile visits are abandoned past 3 seconds.

Suggested headlines

• Only Half of Mobile Sites Pass Core Web Vitals in 2026
• Largest Contentful Paint Is Where Most Mobile Sites Still Fail
• Four Studies, One Answer: Faster Pages Convert Better
• The Mobile Speed Gap Has Barely Closed in Four Years
• 53% of Mobile Visits Vanish After 3 Seconds

Frequently asked questions

Sources

• HTTP Archive Web Almanac, Performance, 2025: https://almanac.httparchive.org/en/2025/performance
• Google Chrome UX Report (CrUX): https://developer.chrome.com/docs/crux
• Deloitte and Google, “Milliseconds Make Millions” (web.dev case study): https://web.dev/case-studies/milliseconds-make-millions
• Portent, “Site Speed Is (Still) Impacting Your Conversion Rate,” 2022: https://portent.com/blog/analytics/research-site-speed-hurting-everyones-revenue.htm
• Akamai, State of Online Retail Performance, Spring 2017: https://www.akamai.com/newsroom/press-release/akamai-releases-spring-2017-state-of-online-retail-performance-report
• Google and SOASTA, “New Industry Benchmarks for Mobile Page Speed,” Think with Google, 2017: https://business.google.com/think/marketing-strategies/mobile-page-speed-new-industry-benchmarks/
• digitalapplied, “Core Web Vitals Benchmarks 2026” (reporting CrUX May 2026): https://www.digitalapplied.com/blog/core-web-vitals-benchmarks-2026-pass-rate-reference