32 Page Speed and Core Web Vitals Statistics, Trends, and Data Points for 2026
This briefing compiles verified data on web performance: how fast pages load across the real web, what share of sites pass Google’s Core Web Vitals, how mobile and desktop differ, and how load time moves conversion and bounce. The field measurement comes from Google’s Chrome UX Report (CrUX) as analyzed in the HTTP Archive Web Almanac, the most authoritative public sample of real-user performance. The conversion-impact figures come from separate commercial studies that each use different samples, time periods, and definitions, so they are directional rather than directly comparable to the CrUX pass rates.
Two cautions run through this asset. First, “page speed” is not one number. Field metrics (CrUX, real users) and lab metrics (Lighthouse, WebPageTest) routinely disagree, and Core Web Vitals are three separate metrics with different pass rates. Second, the widely quoted conversion and bounce statistics come from studies conducted between 2016 and 2019 on different brand sets, so they should be cited with their year and sample, not blended into a single claim.
Executive Summary
- 48% of mobile origins and 56% of desktop origins passed all three Core Web Vitals as of July 2025, per the HTTP Archive Web Almanac using Chrome UX Report field data (Source: HTTP Archive Web Almanac, Performance, 2025).
- Mobile good-CWV share rose from 44% (2024) to 48% (2025), continuing a multi-year improvement (Source: HTTP Archive Web Almanac, Performance, 2024 and 2025).
- Largest Contentful Paint is the binding constraint on mobile: only 62% of mobile origins had good LCP in 2025, versus 77% for INP and 81% for CLS (Source: HTTP Archive Web Almanac, Performance, 2025).
- Desktop interactivity is nearly solved: 97% of desktop origins had good Interaction to Next Paint in 2024, against 74% on mobile (Source: HTTP Archive Web Almanac, Performance, 2024).
- A 0.1-second improvement in mobile site speed was associated with an 8.4% lift in retail conversion rate and a 10.1% lift in travel conversion rate, in a Google-commissioned Deloitte and 55 study of 37 brands and over 30 million sessions in 2019 (Source: Deloitte / Google, Milliseconds Make Millions, 2019/2020).
- As mobile page load time rose from 1 to 3 seconds, the probability of a bounce increased 32%; from 1 to 5 seconds it rose 90% (Source: Google / SOASTA Research, 2017).
- 53% of mobile site visits are abandoned if a page takes longer than 3 seconds to load (Source: Google / DoubleClick, The Need for Mobile Speed, 2016).
- A 100-millisecond delay in load time was associated with a 7% drop in conversion rate across roughly 10 billion retail visits (Source: Akamai / SOASTA, State of Online Retail Performance, 2017).
Key Findings
- 48% of mobile origins passed all three Core Web Vitals as of July 2025, up from 44% in 2024 (Source: HTTP Archive Web Almanac, Performance, 2025).
- 56% of desktop origins passed all three Core Web Vitals as of July 2025, up from 55% in 2024 (Source: HTTP Archive Web Almanac, Performance, 2025).
- The desktop-minus-mobile gap in overall pass rate was 8 percentage points in 2025 (56% vs 48%) (Source: HTTP Archive Web Almanac, Performance, 2025).
- 62% of mobile origins had good LCP in 2025, the lowest of the three metrics and the main reason sites fail (Source: HTTP Archive Web Almanac, Performance, 2025).
- 77% of mobile origins had good INP (under 200ms) in 2025 (Source: HTTP Archive Web Almanac, Performance, 2025).
- 81% of mobile origins had good CLS in 2025 (Source: HTTP Archive Web Almanac, Performance, 2025).
- 97% of desktop origins had good INP in 2024, versus 74% on mobile, the widest device gap among the three metrics (Source: HTTP Archive Web Almanac, Performance, 2024).
- Mobile good-INP share rose from 64% in 2023 to 74% in 2024 after INP replaced First Input Delay as a Core Web Vital in March 2024 (Source: HTTP Archive Web Almanac, Performance, 2024).
- A 0.1-second mobile speed improvement was associated with an 8.4% higher retail conversion rate and a 9.2% higher average order value (Source: Deloitte / Google, Milliseconds Make Millions, 2019/2020).
- The same 0.1-second improvement was associated with a 10.1% higher travel conversion rate (Source: Deloitte / Google, Milliseconds Make Millions, 2019/2020).
- As mobile load time went from 1 to 3 seconds, bounce probability rose 32%; 1 to 5 seconds, 90%; 1 to 6 seconds, 106% (Source: Google / SOASTA Research, 2017).
- 53% of mobile site visits are abandoned when a page takes longer than 3 seconds to load (Source: Google / DoubleClick, The Need for Mobile Speed, 2016).
- A 100-millisecond delay was associated with a 7% conversion drop, and a 2-second delay with a 103% increase in bounce rate, across about 10 billion retail visits (Source: Akamai / SOASTA, State of Online Retail Performance, 2017).
- In Portent’s study, B2C e-commerce sites loading in 1 second converted at roughly 2.5 times the rate of sites loading in 5 seconds (Source: Portent, Site Speed Is (Still) Impacting Your Conversion Rate, 2022).
- B2B lead-generation sites in Portent’s sample converted at about 40% at 1 second, falling to 29% at 3 seconds (Source: Portent, 2022).
Core Web Vitals Pass Rates: What Share of the Web Passes
Core Web Vitals are three field metrics that Google treats as a page-experience signal. An origin “passes” only when the 75th-percentile real-user value clears the “good” threshold for all three at once: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift under 0.1. The HTTP Archive Web Almanac computes these from the Chrome UX Report, which aggregates anonymized field data from opted-in Chrome users across millions of origins.
As of July 2025, 48% of mobile origins and 56% of desktop origins passed all three Core Web Vitals (Source: HTTP Archive Web Almanac, Performance, 2025). That is up from 44% mobile and 55% desktop in the 2024 edition (Source: HTTP Archive Web Almanac, Performance, 2024). The mobile figure has climbed steadily: the Almanac reported 31% of mobile origins passing under the older FID-based definition in 2022 and 37% in 2023. The trend means a majority of mobile sites still fail at least one metric, while desktop has crossed into majority-passing territory.
The Three Metrics: LCP, INP, and CLS Distributions
Aggregate pass rates hide where sites actually break. Splitting Core Web Vitals into their components shows that loading speed, not interactivity or layout stability, is the dominant failure point on mobile.
In 2025, 62% of mobile origins had good LCP, 77% had good INP, and 81% had good CLS (Source: HTTP Archive Web Almanac, Performance, 2025). Because an origin must pass all three at once, the lowest of those three (LCP at 62%) effectively caps the mobile pass rate. On desktop in 2024, the distribution was inverted toward interactivity: 97% good INP and 74% good LCP (Source: HTTP Archive Web Almanac, Performance, 2024). The 2024 device breakdown for LCP was 59% good on mobile versus 72% on desktop, with 14% of mobile origins in the “poor” band (Source: HTTP Archive Web Almanac, Performance, 2024). The practical takeaway: most failing sites fail on LCP, which is driven by server response time, render-blocking resources, and large hero images.
Mobile vs Desktop: A Persistent Gap
Mobile performance trails desktop on every aggregate measure, and the gap is structural: weaker device CPUs, variable cellular networks, and heavier mobile page weight all push mobile metrics down.
The overall pass-rate gap was 8 percentage points in 2025 (56% desktop vs 48% mobile) (Source: HTTP Archive Web Almanac, Performance, 2025). The single largest metric-level gap is interactivity: in 2024, desktop INP was good on 97% of origins versus 74% on mobile, a 23-point spread (Source: HTTP Archive Web Almanac, Performance, 2024). The shift from FID to INP in March 2024 widened the visible mobile gap because INP is a stricter measure of full interaction latency. One metric runs the other way: in 2024, mobile CLS was good on 79% of origins versus 72% on desktop, because desktop layouts with sidebars and ad slots shift more (Source: HTTP Archive Web Almanac, Performance, 2024).
Speed and Conversion: What the Commercial Studies Show
The conversion-impact evidence comes from a handful of commercial studies, each with a different sample and method. They consistently point one direction (faster pages convert better) but their magnitudes are not interchangeable, and several are now several years old.
The Google-commissioned Deloitte and 55 study analyzed 37 European and American brand sites and over 30 million sessions in 2019. It found a 0.1-second mobile speed improvement associated with an 8.4% higher retail conversion rate, a 9.2% higher retail average order value, and a 10.1% higher travel conversion rate (Source: Deloitte / Google, Milliseconds Make Millions, 2019/2020). Akamai’s 2017 report, drawn from about 10 billion retail visits, found a 100-millisecond delay associated with a 7% conversion drop (Source: Akamai / SOASTA, 2017). Portent’s 2022 analysis of 20 sites and over 100 million page views found B2C e-commerce conversion averaging 3.05% at a 1-second load time, declining to 0.67% at 4 seconds (Source: Portent, 2022). These are correlations from observational data, not controlled experiments, so they show association rather than proven causation.
Speed and Bounce
Bounce and abandonment data show the cost of slow loads on the front end of the funnel, before any conversion step.
Google and SOASTA’s 2017 machine-learning analysis found that as mobile load time went from 1 to 3 seconds, bounce probability rose 32%; from 1 to 5 seconds, 90%; and from 1 to 6 seconds, 106% (Source: Google / SOASTA Research, 2017). Google and DoubleClick’s 2016 report found 53% of mobile site visits are abandoned when the page takes longer than 3 seconds (Source: Google / DoubleClick, 2016). Akamai reported a 2-second delay associated with a 103% increase in bounce rate (Source: Akamai / SOASTA, 2017). These three figures come from different samples and years and should be cited individually, not summed.
Reference Tables
| Metric (mobile, good %) | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| All three CWV pass | 31% | 37% | 44% | 48% |
| LCP good | – | – | 59% | 62% |
| INP good | – | 64% | 74% | 77% |
| CLS good | – | – | 79% | 81% |
Sources for table: HTTP Archive Web Almanac, Performance, 2024 and 2025 (CrUX field data). Dashes indicate values not reported in the cited editions; 2022 and 2023 used the FID-based CWV definition.
| Metric, 2024 (good %) | Mobile | Desktop | Gap (pp) |
|---|---|---|---|
| All three CWV pass | 44% | 55% | 11 |
| LCP good | 59% | 72% | 13 |
| INP good | 74% | 97% | 23 |
| CLS good | 79% | 72% | -7 |
Source for table: HTTP Archive Web Almanac, Performance, 2024 (CrUX field data). A negative gap means mobile outperformed desktop.
| Study | Year | Sample | Headline finding |
|---|---|---|---|
| Deloitte / Google, Milliseconds Make Millions | 2019/2020 | 37 brands, 30M+ sessions | 0.1s faster: retail conversion +8.4% |
| Akamai / SOASTA, Online Retail Performance | 2017 | ~10B retail visits | 100ms delay: conversion -7% |
| Google / SOASTA Research | 2017 | Mobile e-commerce (ML model) | 1 to 3s load: bounce probability +32% |
| Google / DoubleClick, Need for Mobile Speed | 2016 | Mobile web sessions | 53% abandon if load > 3s |
| Portent, Site Speed and Conversion | 2022 | 20 sites, 100M+ page views | B2C conversion 3.05% at 1s vs 0.67% at 4s |
Sources for table: linked in the sections above. Samples differ in geography, vertical, and method and are not directly comparable.
Original Synthesis
1. The LCP-limited mobile pass rate. Because an origin must clear all three thresholds simultaneously, the aggregate mobile pass rate is bounded above by its weakest metric. In 2025, mobile LCP-good was 62% while INP-good was 77% and CLS-good was 81%; the observed all-three pass rate was 48%. The 14-point gap between LCP-good (62%) and the actual pass rate (48%) quantifies how many sites clear LCP yet still fail on a second metric. Inputs: HTTP Archive Web Almanac, Performance, 2025. Limitation: the metrics are correlated, so the bound is a ceiling, not an additive decomposition.
2. Where the device gap actually lives. Ranking the 2024 mobile-vs-desktop gaps by metric (INP 23 points, LCP 13 points, CLS minus 7 points) shows interactivity, not layout stability, drives the device divide. CLS is the one metric where mobile leads desktop. Inputs: HTTP Archive Web Almanac, Performance, 2024. Limitation: gaps are origin-level percentage-point differences, not weighted by traffic.
3. Improvement velocity is slowing as the web matures. Mobile all-three pass rate moved +6 points (31 to 37) from 2022 to 2023, +7 points (37 to 44) from 2023 to 2024, and +4 points (44 to 48) from 2024 to 2025. The most recent year-over-year gain (4 points) is the smallest in the series, consistent with diminishing returns as the easiest wins are captured. Inputs: HTTP Archive Web Almanac, Performance, 2022 through 2025. Limitation: the 2022 to 2023 figures used the FID-based definition, so the early series is not perfectly continuous with the INP era.
Charts to build
- Mobile CWV pass rate, 2022 to 2025 (line chart). Data: 31%, 37%, 44%, 48%. Source: HTTP Archive Web Almanac. Insight: steady improvement with a slowing last step. Citation-worthy because it is the cleanest single trend line for “is the web getting faster.”
- 2025 mobile metric distribution, LCP vs INP vs CLS (grouped bar, good/needs-improvement/poor). Data: LCP 62% good, INP 77%, CLS 81% good. Source: HTTP Archive Web Almanac, 2025. Insight: LCP is the choke point. Citation-worthy for diagnosing where sites fail.
- Mobile vs desktop gap by metric, 2024 (diverging bar). Data: INP +23, LCP +13, CLS -7. Source: HTTP Archive Web Almanac, 2024. Insight: interactivity drives the device divide; CLS reverses it.
- Bounce probability vs load time (line). Data: +32% at 3s, +90% at 5s, +106% at 6s relative to 1s. Source: Google / SOASTA, 2017. Insight: bounce risk accelerates with each second. Citation-worthy and highly quotable.
- B2C e-commerce conversion vs load time (bar). Data: 3.05% at 1s declining to 0.67% at 4s. Source: Portent, 2022. Insight: the conversion penalty of slowness in absolute rates.
Inline rendered chart, mobile CWV pass rate by year (good %):
Source: HTTP Archive Web Almanac, Performance, 2022 to 2025 (CrUX field data).
Methodology
Source selection prioritized primary field data. The HTTP Archive Web Almanac is the lead source for pass rates and metric distributions because it is built on the Chrome UX Report, Google’s real-user dataset spanning millions of origins, and is published with open methodology. For each pass-rate figure, the year and device (mobile or desktop) are stated, since they materially change the number. The conversion and bounce studies (Deloitte/Google, Akamai/SOASTA, Google/SOASTA, Google/DoubleClick, Portent) are included as secondary evidence of business impact; each is attributed to its publisher, year, and sample. Where studies disagreed or used different definitions, both the FID and INP eras of CWV are flagged rather than blended. Conflicting numbers were resolved by preferring the primary publisher’s own document over third-party recaps; figures that could only be traced to aggregator blogs were excluded. No numbers were estimated or invented. Last updated June 2026.
Source Quality
Tier 1 (primary field data and platform owners): HTTP Archive Web Almanac (Performance, 2024 and 2025), built on the Google Chrome UX Report; Google / DoubleClick and Google / SOASTA research published via Think with Google and web.dev.
Tier 2 (credible vendor and consultancy research using large samples): Deloitte and 55, Milliseconds Make Millions (Google-commissioned); Akamai / SOASTA State of Online Retail Performance; Portent site-speed study.
Tier 3 (reputable journalism and recaps used only to locate primary documents): Marketing Dive and trade-press coverage of the DoubleClick and Akamai reports.
Most Quotable Statistics
- “As of July 2025, 48% of mobile sites and 56% of desktop sites passed all three Core Web Vitals.” (HTTP Archive Web Almanac, 2025)
- “Only 62% of mobile sites had good LCP in 2025, making load speed the main reason sites fail Core Web Vitals.” (HTTP Archive Web Almanac, 2025)
- “A 0.1-second mobile speed improvement was tied to an 8.4% lift in retail conversions.” (Deloitte / Google, 2019/2020)
- “As mobile load time goes from 1 to 3 seconds, bounce probability rises 32%.” (Google / SOASTA, 2017)
- “53% of mobile visits are abandoned if a page takes more than 3 seconds to load.” (Google / DoubleClick, 2016)
Data Limitations
- CrUX covers opted-in Chrome users with sufficient traffic, so very low-traffic origins and non-Chrome users are underrepresented.
- Core Web Vitals thresholds and the metric set changed (FID to INP in March 2024), so pre-2024 and post-2024 pass rates are not perfectly continuous.
- The conversion and bounce studies are observational correlations on different samples and years (2016 to 2022); they should not be summed or treated as controlled causal estimates.
- The Deloitte and Portent samples are small (37 and 20 sites) relative to CrUX, so their magnitudes carry wider uncertainty.
- Some headline conversion figures are now five to ten years old and predate large shifts in mobile networks and device speed.
Recommended Dataset Fields
- year (integer)
- device (mobile / desktop)
- metric (cwv_all / lcp / inp / cls)
- good_pct, needs_improvement_pct, poor_pct (numeric)
- data_source (e.g., HTTP Archive Web Almanac)
- field_dataset (e.g., CrUX July 2025)
- cwv_definition_era (fid / inp)
- study_name, study_year, sample_size, metric_type (field / lab / conversion / bounce)
- source_url
Press Summary
Most of the web still fails Google’s speed test on phones. As of July 2025, 48% of mobile sites and 56% of desktop sites passed all three Core Web Vitals, per the HTTP Archive Web Almanac built on Google’s Chrome UX Report field data. The mobile figure is improving, up from 31% in 2022, but the gains are slowing. The choke point is loading speed: only 62% of mobile sites had good Largest Contentful Paint in 2025, versus 77% for interactivity and 81% for layout stability. The business stakes are documented in older commercial studies. A Google-commissioned Deloitte study of 37 brands found a 0.1-second mobile speed gain associated with an 8.4% lift in retail conversions, and Google/SOASTA research found bounce probability rising 32% as load time moved from one to three seconds. Those conversion figures come from different samples and years and should be cited individually rather than blended. For founders, the practical reading is that LCP, not flashy interactivity work, is where most speed failures live.
Suggested Headlines
- Less Than Half of Mobile Sites Pass Google’s Core Web Vitals in 2025
- Loading Speed, Not Interactivity, Is Where Most Sites Fail Core Web Vitals
- The 8-Point Mobile-Desktop Performance Gap That Will Not Close
- What a Tenth of a Second Is Worth: Speed and Conversion, by the Numbers
- 53% of Mobile Visitors Leave After 3 Seconds: The Data Behind the Stat
FAQ
What share of websites pass Core Web Vitals?
As of July 2025, 48% of mobile origins and 56% of desktop origins passed all three Core Web Vitals (Source: HTTP Archive Web Almanac, Performance, 2025).
Which Core Web Vital do sites fail most often?
Largest Contentful Paint, with only 62% of mobile origins rated good in 2025 versus 77% for INP and 81% for CLS (Source: HTTP Archive Web Almanac, Performance, 2025).
How much faster is desktop than mobile?
The overall pass-rate gap was 8 percentage points in 2025 (56% desktop vs 48% mobile) (Source: HTTP Archive Web Almanac, Performance, 2025).
What is a good INP score and how many sites hit it?
Good INP is under 200 milliseconds; 77% of mobile and 97% of desktop origins met it (mobile 2025, desktop 2024) (Source: HTTP Archive Web Almanac, Performance, 2024 and 2025).
Is web performance improving over time?
Yes; mobile all-three pass rate rose from 31% in 2022 to 48% in 2025, though the most recent annual gain was the smallest (Source: HTTP Archive Web Almanac, Performance, 2022 to 2025).
How much does speed affect conversions?
A 0.1-second mobile speed improvement was associated with an 8.4% higher retail conversion rate in a 2019 Google-commissioned study of 37 brands (Source: Deloitte / Google, Milliseconds Make Millions, 2019/2020).
How much does a 100-millisecond delay cost?
Akamai’s 2017 report associated a 100-millisecond delay with a 7% drop in conversion across about 10 billion retail visits (Source: Akamai / SOASTA, 2017).
What happens to bounce rate when pages get slower?
As mobile load time went from 1 to 5 seconds, bounce probability rose 90% (Source: Google / SOASTA Research, 2017).
How many mobile users abandon slow pages?
53% of mobile site visits are abandoned when a page takes longer than 3 seconds to load (Source: Google / DoubleClick, The Need for Mobile Speed, 2016).
Are these conversion statistics directly comparable?
No; they come from different samples, verticals, and years (2016 to 2022) and should be cited individually, not summed or averaged (Source: see individual studies above).
For teams turning these benchmarks into a remediation plan, our technical SEO checklist for founders maps the LCP, INP, and CLS fixes that move the needle. CO Consulting runs performance and growth research like this; if you want help applying it to your own field data, you can book a consultation.