What counts as a good website speed score in 2026

What is a website speed score in 2026

A website speed score in 2026 is a single number, usually out of 100, that summarises how quickly a page loads and becomes usable for visitors. The most widely used system is Google’s Core Web Vitals, which combines three measurements: Largest Contentful Paint (how long until the main content appears), Interaction to Next Paint (how long until the page responds to the first click), and Cumulative Layout Shift (how much the page jumps around while loading). These three scores are averaged into an overall score for mobile and a separate one for desktop. Anything below 60 on either device is a signal that visitors are likely to leave before the page finishes loading. A score of 90 or above on mobile and 85 or above on desktop is the point where speed stops being a ranking risk and becomes a performance lever instead.

How many visitors leave if a site is slow

Sites that load in under 2.5 seconds convert 17% more visitors than those taking over 4 seconds, according to aggregated data from thousands of ecommerce stores. A 0.1 second improvement can raise conversion rates by up to 8%, and most users expect pages to load within 2 seconds. Those figures come from real transaction data, not marketing claims. The correlation holds across sectors, with the biggest drops appearing on sites where the first meaningful paint takes more than 3 seconds. If your site’s Largest Contentful Paint is above 3 seconds on mobile, expect roughly 10% more visitors to abandon the page before it finishes loading.

What the Core Web Vitals score actually measures

Core Web Vitals measures three technical events on every page load. Largest Contentful Paint (LCP) is the moment when the biggest block of text or image appears on screen. Interaction to Next Paint (INP) is the delay between a user’s click or tap and the page responding. Cumulative Layout Shift (CLS) is the total amount of unexpected movement of page elements while they load. Google’s scoring system gives a green rating when LCP is under 2.5 seconds, INP is under 200 milliseconds, and CLS is under 0.1. Anything outside those ranges moves the overall score down. The scores are calculated separately for mobile and desktop because phones have slower processors and less stable connections, so the thresholds are slightly looser on desktop.

What counts as a good score in 2026

A good website speed score in 2026 is 90 or above for mobile and 85 or above for desktop. Anything below 60 on either device typically needs work to avoid ranking and conversion losses. Sites scoring between 60 and 90 usually rank normally but miss out on the full benefit of speed-related ranking factors. Sites below 60 risk losing positions in search results and are less likely to be cited by AI answer engines. Most small business websites fall into the 60, 80 range, which is not disastrous but leaves speed-related gains on the table. If your score is already in the high 80s, further optimisation may not be worth the effort unless you are in a highly competitive sector.

The four reasons most sites score badly

Nearly 70% of slow sites are held back by oversized images or uncompressed assets. A typical product photo shot on a modern phone can be 4 to 6 megabytes, but compressing it to 200 kilobytes does not change how it looks on screen. The second most common problem is render-blocking JavaScript or CSS that forces the browser to pause while it downloads and parses files. Deferring non-critical scripts until after the page has rendered can shave a second or two off load time. Server response time is the next bottleneck. If your site takes more than 200 milliseconds to start sending data, the issue is usually hosting. Shared servers with no caching optimisation are the usual culprit. Code bloat is the final frontier: a bloated theme or plugin library can add hundreds of kilobytes of unnecessary JavaScript and CSS. Removing unused code and consolidating plugins will cut load times by 30% to 50% in many cases.

How to improve your score in two weeks

Start by compressing images and deferring non-critical JavaScript and CSS. Tools such as ImageMagick or Squoosh can reduce file sizes without visible quality loss. For JavaScript, use the defer or async attribute on scripts that do not need to run immediately. Next, enable server-side caching and a CDN with edge locations. Most hosting plans include basic caching, but enabling it can halve server response times. If your site still scores below 60, move to a plan with built-in optimisation or consider a lightweight theme. A full code audit is usually only necessary when the site is custom-built or uses dozens of plugins. At that point, the time and cost often exceed the benefit for small businesses.

When to stop optimising your speed score

Stop optimising when the business impact no longer justifies the effort. If your mobile score is already 90 or above and desktop is 85 or above, further tweaks are unlikely to move the needle. If you are between 60 and 90, a 10-point gain is usually worth doing but a 20-point gain is rarely justified unless you are in a high-competition sector. If you are below 60, prioritise the fastest wins: images and render-blocking resources. Once those are fixed, decide whether the remaining gap is worth the extra cost of server upgrades or code refactoring. If your site is already ranking well and conversions are stable, keep monitoring but do not treat a score of 85 as a failure.

What happens if you ignore your speed score

Sites that ignore speed scores gradually lose traffic as search rankings drift and AI answer engines stop citing them. A drop of one position in search results can cost up to 6% of organic traffic, and slow pages are less likely to appear in featured snippets. Visitors also abandon pages that take too long to load, which shows up as a higher bounce rate in analytics. The combined effect is measurable: sites with LCP above 3 seconds on mobile see roughly 10% fewer transactions. If your competitors optimise while you do not, the gap widens over time. The risk is not immediate, but it accumulates month by month until the cost of fixing the problem becomes larger than the cost of prevention.

Should you pay for a speed optimisation service

A speed optimisation service is worth paying for when your in-house team lacks the tools or time to compress images, defer scripts, and enable caching. Typical costs are between £300 and £1,200 for a one-off audit and fixes. If your site scores below 60, expect the work to take two to four weeks and yield a 15 to 25 point lift. If you score between 60 and 90, the gains are smaller and may not justify the cost unless you are in a highly competitive sector. DIY tools such as GTmetrix, WebPageTest, and Cloudflare’s free plan can achieve most of the gains for no cost. Reserve paid services for cases where the site is complex, the team lacks time, or the technical debt is high.

How to check your site’s speed score

Use free tools such as PageSpeed Insights, GTmetrix, or WebPageTest. Enter your URL and run the test on both mobile and desktop. The results will show your Core Web Vitals score, a breakdown of LCP, INP, and CLS, and a list of suggested fixes ranked by impact. These tools update their scoring algorithms roughly every six months, so run the test again after any major site change. If you use Google Analytics 4, the Site Speed reports provide real-user measurements alongside lab scores. Combine both sources to see whether your lab score matches actual visitor experience.

What to do next

Run a speed test on your site today and note the mobile and desktop scores. If either is below 60, prioritise image compression and script deferral. If you score between 60 and 90, focus on server response times and caching. If you are already above 90 on mobile and 85 on desktop, set a schedule to re-test every six months. Speed is not a one-off project but a continuous process of monitoring and small improvements.