What makes a website fast in 2026, and why it matters for SEO
I read what Google, the Chrome team and the largest public dataset of websites say about speed, and put it in plain language.
On this page
The short answer
A fast website loads its main content within 2.5 seconds, responds to a tap or click within 200 milliseconds, and keeps its layout steady while it loads, for at least three out of four page loads. Google says these Core Web Vitals line up with what its ranking systems reward, but it also says relevance comes first and good scores do not guarantee a top ranking. That is what I found after reading the sources below, so I would treat speed as something you do for your visitors first and for search second.
Key takeaways
- Google's Core Web Vitals define “fast”: main content within 2.5 seconds, a response to a tap within 200 milliseconds, and a layout shift score of 0.1 or less.
- Most sites are not there yet: in 2025, 48% of websites had good Core Web Vitals on mobile and 56% on desktop.
- Loading the main content is the hardest part: 62% of mobile sites have a good LCP, against 77% for INP and 81% for CLS.
- Google says these metrics align with what its ranking systems reward, but relevance comes first and good scores do not guarantee a top ranking.
- The one business case I found: in a Vodafone A/B test, a 31% better LCP went with 8% more sales. One company's result, not a promise.
- Slow loading is often not about image size: most sites with poor LCP spend under 10% of that time downloading the image, so the delay before it matters more.
What does “fast” actually mean?
It has a precise meaning now. Google and the Chrome team define it with three measurements called Core Web Vitals.
Softlixx builds websites, so I have a stake in this topic.
I started with the Chrome team's own definition on web.dev. Largest Contentful Paint (LCP) measures loading: the main content should appear within 2.5 seconds of the page starting to load. Interaction to Next Paint (INP) measures responsiveness: pages should respond within 200 milliseconds or less. Cumulative Layout Shift (CLS) measures visual stability: pages should keep a score of 0.1 or less. The team explains in a second article how the “poor” boundaries were set: more than 4 seconds for LCP, more than 500 milliseconds for INP, and 0.25 for CLS.
Measured at the 75th percentile of page loads, across mobile and desktop
Largest Contentful Paint (LCP)
Loading: when the main content appears
Interaction to Next Paint (INP)
Responsiveness: how fast the page reacts to a tap or click
Cumulative Layout Shift (CLS)
Visual stability: how much the layout jumps
The three bands are drawn at equal width so they are easy to read. They are not to scale.
Sources: web.dev (Web Vitals), web.dev (thresholds).
These are measured on real visitors, not on one test. The Chrome team says a good threshold to measure is the 75th percentile of page loads, segmented across mobile and desktop, and that a page passes when it meets the targets at that percentile for all three metrics. In plain words, at least three out of four page loads have to be good.
One change is worth knowing about. INP replaced First Input Delay as a Core Web Vital, and web.dev's announcement, last updated on 12 March 2024, calls it a significant step forward in how interaction responsiveness is measured.
By establishing clear, actionable metrics, Core Web Vitals created a common goal that everyone could rally behind.
How many websites are actually fast?
Fewer than you might think. I found the answer in the HTTP Archive's Web Almanac, which measures millions of real websites.
The Web Almanac 2025 performance chapter, based on July 2025 data from the HTTP Archive and the Chrome UX Report, found that 48% of websites had good Core Web Vitals on mobile and 56% on desktop. On mobile that is up from 36% in 2023 and 44% in 2024. On desktop it rose from 55% in 2024 to 56% in 2025.
Loading is the weak spot. On mobile, 62% of websites have a good LCP, against 77% for INP and 81% for CLS. The chapter compares this with desktop, where 74% of pages reach a good LCP, and describes the gap as consistent with slower networks and less capable devices on mobile.
Websites with good Core Web Vitals, by year (HTTP Archive)
Bars use a scale of 0 to 100%.
Mobile websites with a good score, by metric, 2025
Bars use a scale of 0 to 100%. Data from July 2025.
Source: HTTP Archive.
My readingThe Almanac measures the whole public web, so it cannot tell you where your own site stands. It does tell me that passing is not a given, and that loading is where most sites have the most room to improve.
Does speed help SEO?
A little, and Google is careful about how it says so. I read its two documentation pages on this.
Google's page experience documentation, last updated on 22 September 2026, says its core ranking systems look to reward content that provides a good page experience. Its Core Web Vitals page says these metrics, along with other page experience aspects, align with what its core ranking systems seek to reward.
Google Search always seeks to show the most relevant content, even if the page experience is sub-par.
The same page adds a warning. Getting good results in Search Console's Core Web Vitals report or in third-party tools does not guarantee that your pages will rank at the top of Google Search results, and there is more to a great page experience than the scores alone.
For the wider SEO picture, including AI search, see how to get cited by ChatGPT and AI search. If you are planning a new site, the research on website redesigns and SEO covers how to keep your search traffic through the move.
My readingI read this as two things at once. Speed will not rescue weak content, and I would not expect a faster site to jump the queue on its own. But Google names it as something its ranking systems are built to reward, and it is one of the few things you can measure and fix yourself.
Does speed help the business, not just the ranking?
The best evidence I found is one company's A/B test, so I would read it as an example, not a promise.
In a case study published on web.dev in March 2021, Vodafone ran an A/B test on a landing page. Version A was optimized for Web Vitals and had a 31% better LCP score in the field than version B, and Vodafone found it produced 8% more sales, a 15% improvement in the lead to visit rate and an 11% improvement in the cart to visit rate.
My readingIt is one company's test on one landing page, published on Google's web.dev. So I would treat the result as an example, not a forecast, and test the effect on your own pages.
Optimizing for quality of user experience is key to the long-term success of any site on the web.
What actually slows most sites down?
It is rarely one thing, and it is not always the image file size. The Chrome team's real-user data points somewhere you might not expect.
Start with what the biggest element on the page usually is. The Web Almanac found that an image is the LCP element on 85.3% of desktop pages and 76% of mobile pages, with text making up much of the remainder.
You might expect the fix to be smaller images. But the Chrome team's guide says an analysis of real-user data from Chrome shows that the majority of origins with poor LCP spend less than 10% of their LCP time, at the 75th percentile, downloading the LCP image. Among pages with poor LCP, loading the image is delayed on the client by 1,290 milliseconds at the 75th percentile, which the guide says is more than half of the budget for a fast experience. The time is going before the download even starts.
The Chrome team's guide to optimizing LCP splits LCP time into four parts and gives rough targets: about 40% waiting for the server to respond (time to first byte), under 10% before the browser starts loading the image, about 40% loading the image, and under 10% before it is drawn on screen. It adds that these are guidelines, not strict rules.
The Chrome team's rough targets for the four parts of LCP time. They are guidelines, not strict rules.
- Time to first byteabout 40%From starting to load the page until the browser receives the first byte of the HTML.
- Resource load delayunder 10%From then until the browser starts loading the main image or other resource.
- Resource load durationabout 40%The time it takes to load that resource itself.
- Element render delayunder 10%From the resource finishing loading until the element is fully drawn.
Source: web.dev.
The Web Almanac authors flag one common mistake in how images are loaded.
Lazy loading is generally a useful optimization, but applying it to the LCP image can be harmful because it delays the main content users are waiting to see.
What fixes each metric?
The Chrome team names one place to start for each metric.
| Metric | What it measures | Where the Chrome team says to start |
|---|---|---|
| LCP | Loading: when the main content appears | Make sure the LCP resource is discoverable from the HTML source and prioritized. |
| INP | Responsiveness: how fast the page reacts | Yield often to break up long tasks. A task over 50 milliseconds is a long task. |
| CLS | Visual stability: how much the layout jumps | Reserve the required space in advance as much as possible, so the layout does not move as content loads. |
Source: web.dev.
The guide, last updated in October 2024, also says which one is hardest. LCP is the Core Web Vital that developers tend to struggle with the most, and 40% of sites in the Chrome UX Report do not meet its recommended threshold. CLS is the metric most sites do well on, yet about a quarter of them still miss the recommended threshold.
How do you measure it?
The Chrome team measures it on real page loads at the 75th percentile, split by mobile and desktop.
Google's page experience documentation mentions Search Console's Core Web Vitals report and third-party tools. The Chrome team's LCP guide adds that the Performance panel of Chrome DevTools shows your local LCP experience next to the page's or origin's Chrome UX Report LCP, and that a performance trace includes a breakdown of the LCP subpart timings.
My readingI would start with real-user data, because it is what the thresholds are defined on, and I would look at mobile first, because that is where the gap is.
What should a business do first?
If I were starting today, I would work in this order: measure, fix loading, then responsiveness, then layout stability. It is my reading of the sources above.
- Measure real visitors on mobileUse the Core Web Vitals report in Search Console and the Chrome UX Report, at the 75th percentile.
- Find your LCP elementIt is an image on about three out of four mobile pages, so check what yours is.
- Make it discoverable and prioritizedMake sure it can be found in the HTML source and give it priority. Do not lazy-load it.
- Look at the server tooTime to first byte is about 40% of the LCP budget in the Chrome team's guide.
- Break up long scriptsA task over 50 milliseconds is a long task, so yield often.
- Reserve space in advanceHold the room a piece of content needs before it loads, as much as you can, so the layout does not jump.
- Re-measure, then test the effectCheck the change on real visitors, and if you can, run a small test on your own pages.
Questions people ask
Are Core Web Vitals a Google ranking factor?
Google says its core ranking systems look to reward content that provides a good page experience, and that Core Web Vitals align with what those systems seek to reward. It also says it always seeks to show the most relevant content, even if the page experience is sub-par, and that good scores do not guarantee a top ranking. So I treat them as helpful, not decisive.
What is a good Core Web Vitals score?
The Chrome team's thresholds are an LCP within 2.5 seconds, an INP of 200 milliseconds or less, and a CLS of 0.1 or less, measured at the 75th percentile of page loads across mobile and desktop. In plain words, at least three out of four visits should meet all three targets.
What replaced First Input Delay?
Interaction to Next Paint (INP) replaced First Input Delay as a Core Web Vital. web.dev's announcement, last updated on 12 March 2024, calls it a significant step forward in how interaction responsiveness is measured. The Chrome team says pages should respond within 200 milliseconds or less.
Will making my site faster increase sales?
It did for Vodafone, in an A/B test where a 31% better LCP went with 8% more sales. That is one company's result, so I would not count on a specific number. I would test it on my own pages.
Sources
- Web Vitalsweb.dev (Philip Walton). Published 4 May 2020, last updated 31 October 2024.
- How the Core Web Vitals metrics thresholds were definedweb.dev. Documentation, read 28 September 2026.
- The most effective ways to improve Core Web Vitalsweb.dev (Chrome team). Last updated 31 October 2024.
- Optimize Largest Contentful Paintweb.dev (Chrome team). Documentation, read 28 September 2026.
- Interaction to Next Paint is officially a Core Web Vitalweb.dev (Rick Viscomi). Last updated 12 March 2024.
- Understanding Google Page ExperienceGoogle Search Central. Updated 22 September 2026.
- Understanding Core Web Vitals and Google search resultsGoogle Search Central. Updated 10 December 2025.
- Performance | 2025 | The Web AlmanacHTTP Archive. 2025 edition, data from July 2025.
- Vodafone: A 31% improvement in LCP increased sales by 8%web.dev. Published 17 March 2021.
- The History of Core Web VitalsAddy Osmani (addyosmani.com). 2 October 2025.
How I checked this. I opened every source linked here and checked each figure and each quotation against the source page, most recently on 3 October 2026. Anything in a box marked "My reading" is my own interpretation, not a finding of the studies. Studies change and new ones appear, so I will update this piece and change its dates when the evidence does.