Website Performance Statistics (2026): Speed, Core Web Vitals and Conversions

25 website performance statistics for 2026, every one linked to its primary source: Core Web Vitals pass rates, load time vs conversion data, page weight trends and checkout benchmarks.

Website speed statistics are among the most misquoted numbers in marketing. The same recycled figures bounce from listicle to listicle, stripped of their source, their year and their methodology, until nobody can say where they came from. That is a problem when you are using them to justify a redesign budget or a performance sprint.

This page fixes that. Every statistic below links directly to its primary source, with the publisher and year named inline. Most come from the HTTP Archive Web Almanac, the Chrome UX Report data behind it, Google-commissioned research and published vendor studies with real methodology. At the end, we also list the famous "classics" you should stop citing, because their origins can no longer be verified.

If you are weighing up a rebuild, our web design team lives inside this data daily, and there is more practical follow-up across our web design articles.

Key Statistics at a Glance

The ten numbers we reach for most often, each linked to its source:

Only 48% of websites deliver a good Core Web Vitals experience on mobile, per the 2025 Web Almanac by HTTP Archive, up from 44% in 2024.
A site that loads in 1 second converts about 2.5x better than a site that loads in 5 seconds for B2C ecommerce, per Portent's 2022 site speed study.
A 0.1 second mobile speed improvement lifted retail conversions by 8.4% in the Deloitte and Google "Milliseconds Make Millions" study (2020).
The same 0.1 second gain lifted lead generation form-page progression by 21.6% in that Deloitte study (2020).
A 100 millisecond delay can cut conversion rates by up to 7%, per the Akamai and SOASTA Online Retail Performance Report (2017).
53% of mobile visits are abandoned if a page takes longer than 3 seconds to load, per Google's mobile page speed research (2017).
Bounce probability rises 32% as load time goes from 1 to 3 seconds, per the same Google benchmarks (2017).
The median mobile home page has grown 202.8% in a decade, from 845 KB in 2015 to roughly 2.4 MB, per the 2025 Web Almanac page weight chapter.
Vodafone lifted sales 8% by improving LCP 31% in a controlled A/B test, per web.dev's Vodafone case study (2021).
70.22% of online shopping carts are abandoned on average, per Baymard Institute's running meta-analysis of 50 studies (2025).

Core Web Vitals in 2026: Most Sites Still Do Not Pass

Core Web Vitals are Google's field-measured benchmarks for loading (LCP), responsiveness (INP) and visual stability (CLS), drawn from real Chrome users via the Chrome UX Report. The Web Almanac's annual analysis of that data is the closest thing the industry has to a census, and it shows a web that is improving slowly but still failing more often than it passes on mobile.

48% of websites achieve a good Core Web Vitals assessment on mobile, up from 44% the year before, per the 2025 Web Almanac performance chapter. On desktop the figure is 56%.
Largest Contentful Paint is the weakest link: only 62% of sites achieve good LCP on mobile, against 74% on desktop, per the 2025 Web Almanac.
Time to First Byte is worse still: just 44% of mobile origins record a good TTFB (under 800 ms), versus 55% on desktop, per the 2025 Web Almanac. Slow servers cap everything downstream.
Responsiveness has a huge device gap: 97% of desktop sites achieve good INP, but only 77% of mobile sites do, per the 2025 Web Almanac. That 20 point gap did narrow from 23 points in 2024.
Layout stability is the one bright spot: 81% of mobile sites achieve good CLS, per the 2025 Web Almanac.

The practical read: if your site passes all three vitals on mobile, you are ahead of more than half the web. Most of your local competitors will not be.

Load Time vs Bounce and Conversion

These are the studies that connect seconds to dollars, with published methodology behind each.

Ecommerce sites loading in 1 second converted at 3.05%; the rate fell to roughly 1% at 5 seconds, a 2.5x difference, per Portent (2022), based on more than 100 million page views across 20 sites.
B2B lead generation sites loading in 1 second converted about 3x better than at 5 seconds, and 5x better than at 10 seconds, per Portent (2022).
Ecommerce conversion rates dropped by an average of 0.3 percentage points for every additional second of load time, with the best rates landing between 0 and 2 seconds, per Portent (2022).
A 100 millisecond delay hurt conversion rates by up to 7%, and a 2 second delay increased bounce rates by 103%, per the Akamai and SOASTA Spring 2017 Online Retail Performance Report, which analysed roughly 10 billion retail visits over one month.
The probability of a bounce increases 32% as page load time goes from 1 second to 3 seconds, and 123% from 1 second to 10 seconds, per Google's neural-network analysis of 900,000 mobile landing pages (2017).
In a controlled A/B test, a 31% LCP improvement produced 8% more sales, a 15% better lead-to-visit rate and an 11% better cart-to-visit rate for Vodafone, per web.dev (2021). The two page versions were visually identical; only the performance work differed.

We break down what these numbers mean for tradies and local operators in our guide to service business website speed optimisation.

What 0.1 Seconds Is Worth: The Deloitte Numbers

The most rigorous public study on marginal speed gains remains "Milliseconds Make Millions" (2020), commissioned by Google and conducted by Deloitte and agency fifty-five. It tracked 37 European and American brand sites, over 30 million user sessions, hour by hour for 30 days, with no design changes allowed during the study. For a 0.1 second improvement in mobile site speed:

Retail conversions rose 8.4% and average order value rose 9.2%.
Travel conversions rose 10.1%.
Lead generation sites saw 21.6% more users progress to the form submission page.
Luxury sites saw a 40.1% increase in users moving from product detail pages to the basket.

The finding that matters for service businesses is the lead generation one. If a tenth of a second moves a fifth more visitors toward your enquiry form, speed is not a technical nicety. It is a lead-cost lever.

The Mobile Performance Gap

Most Australian local search happens on phones, and phones are exactly where the web performs worst.

Good LCP: 62% of sites on mobile vs 74% on desktop. Good TTFB: 44% vs 55%. Good INP: 77% vs 97%, all per the 2025 Web Almanac.
53% of mobile site visits are abandoned when pages take longer than 3 seconds, per Google research (2017).
For 70% of the mobile landing pages Google analysed, above-the-fold visual content took nearly 7 seconds to display, per the same Google benchmarks study (2017). The hardware has improved since; the pages have got heavier.

Test on a mid-range Android over 4G, not on your office fibre. That is where your customers are.

Page Weight Keeps Climbing

The reason mobile performance stays stubborn is simple: pages keep getting heavier faster than networks get faster.

The median mobile home page grew 202.8% in ten years, from 845 KB in July 2015 to about 2.4 MB in 2025, per the 2025 Web Almanac page weight chapter.
Images remain the heaviest payload at a median of roughly 911 KB per mobile home page, with JavaScript next at around 632 KB, per the 2025 Web Almanac.
The median page now ships 315 KB of video, a 28% jump in a single year from 246 KB, per the 2025 Web Almanac.
At the 90th percentile, mobile pages weigh more than 8.3 MB, per the 2025 Web Almanac. One in ten pages is heavier than that.

Every kilobyte is a choice. Most of the weight above is decorative, not functional, which is why performance-first design keeps beating retrofitted optimisation.

Forms, Checkouts and Conversion Friction

Speed gets people to the page; friction decides whether they finish. Baymard Institute's checkout research is the reference dataset here.

The average documented online cart abandonment rate is 70.22%, per Baymard Institute's meta-analysis of 50 studies (updated 2025).
17% of US online shoppers have abandoned an order because the checkout process was too long or complicated, per Baymard Institute (2025).
Fixing documented checkout usability issues can lift conversion rates by 35.26% for a typical large ecommerce site, per Baymard Institute (2025).

For a service business the equivalent is your enquiry form: every unnecessary field is a step in someone else's checkout research come to life on your quote page.

Stats to Stop Citing

Some famous performance statistics no longer meet a basic sourcing test. We exclude these from client work and suggest you do the same:

"Amazon found every 100ms of latency cost 1% of sales." This traces to a 2006 conference presentation by a former Amazon engineer, not a published study. Amazon has never released the methodology, and the figure is two decades old. Directionally plausible, unverifiable in practice.
"40% of users abandon a site that takes more than 3 seconds." Usually attributed to a 2009-era Gomez/Forrester survey that is no longer publicly available. The modern, traceable replacement is Google's 53% mobile abandonment figure linked above.
"The average mobile page takes 22 seconds to load." Real Google finding, but from 2017 hardware and networks. Quoting it as current overstates the problem; use current CrUX and Web Almanac data instead.

Methodology and Sources

Statistics on this page were compiled in mid-2026. Every figure links inline to its primary source, and we recorded publisher and year at each citation. Primary datasets and studies used:

The Web Almanac 2025, performance chapter (HTTP Archive, built on Chrome UX Report field data from millions of real origins)
The Web Almanac 2025, page weight chapter (HTTP Archive lab crawl)
"Milliseconds Make Millions" (Deloitte and fifty-five for Google, 2020; 37 sites, 30M+ sessions, 30 days)
Site speed and conversion research (Portent, 2022; 20 sites, 100M+ page views)
Spring 2017 Online Retail Performance Report (Akamai/SOASTA; ~10 billion retail visits)
Mobile page speed industry benchmarks (Google, 2017; 900,000 landing pages)
Vodafone Web Vitals case study (web.dev, 2021; controlled A/B test)
Cart abandonment rate meta-analysis (Baymard Institute, updated 2025; 50 studies)

Where a source publishes updated editions annually, we cite the most recent edition available at the time of writing. Older studies (2017, 2020) are retained only where no newer study of comparable rigour exists, and are always dated inline.

FAQ

What percentage of websites pass Core Web Vitals in 2026?

Per the most recent Web Almanac by HTTP Archive (2025), 48% of websites deliver a good Core Web Vitals experience on mobile and 56% on desktop. Mobile pass rates have improved four percentage points year on year, but the majority of the mobile web still fails at least one vital, most commonly LCP.

How much does one second of load time cost in conversions?

Portent's 2022 study found ecommerce conversion rates fall by roughly 0.3 percentage points per additional second, with sites loading in 1 second converting about 2.5x better than sites loading in 5. For smaller margins, the Deloitte study (2020) found even a 0.1 second improvement lifted retail conversions 8.4%.

Is website speed a Google ranking factor?

Yes, though a modest one. Google confirmed page experience signals, including Core Web Vitals, feed into ranking. The larger commercial effect is behavioural: the bounce and conversion data above dwarfs the direct ranking impact for most local businesses.

What is a good page load time in 2026?

Aim for a Largest Contentful Paint under 2.5 seconds on real mobile devices, which is Google's "good" threshold, and note that Portent's data shows conversion rates peak when pages load within about 2 seconds. Field data from the Chrome UX Report is the measure that counts, not a lab score on office wifi.

Why is mobile performance so much worse than desktop?

Two compounding causes: pages have grown heavier (the median mobile home page is up 202.8% in a decade, per the Web Almanac 2025) and mobile devices have less CPU to parse the median ~632 KB of JavaScript. That is why only 77% of sites achieve good INP on mobile versus 97% on desktop, per the Web Almanac 2025.

Cite These Statistics

You are welcome to cite any statistic on this page in your own articles, decks or reports. Please credit Orkkid and link to this page (https://orkkid.com/blog/website-performance-statistics), and keep the original source attribution intact, since every number here belongs to the researchers who produced it. If you spot a figure that has been superseded by a newer edition of its source, tell us and we will update it.

Want your own site on the right side of these numbers? Our web design work is performance-first from the initial wireframe, and you can dig deeper in the rest of our web design category.