What “Fast” Actually Means to a Real Website Visitor

When did you last check your website the way your customers actually see it?

Not on your laptop at home, on fast broadband, where everything loads in a blink. But on a phone. On a real connection. In the middle of a working day, when signals are patchy, and patience is short.

Most people who ask me about website speed want one thing: a green score on Google PageSpeed Insights. They think, get it above 90, and the job is done. What they do not realise is that a PageSpeed score is a lab measurement, taken under controlled conditions on a simulated device. It tells you something useful. But it does not tell you what a real visitor on a patchy 4G connection on a Tuesday afternoon is actually experiencing.

The gap between scoreboard speed and felt speed is where a lot of small business websites are losing people.

What Google Actually Measures vs What You See

PageSpeed Insights gives you a number between 0 and 100. That number is generated by a tool called Lighthouse, simulating your page under standardised conditions. It is good for spotting problems. But not the same as a real visitor experience.

Google’s ranking system evaluates real-world field data from the Chrome User Experience Report, not lab simulations. A page scoring 75 in the lab can rank perfectly well if its Core Web Vitals are in the good range. Real user data and lab data can look completely different; one site recorded an LCP of 1.8 seconds in real user data while the lab test showed 8 seconds for the same page. 

So a 95 on PageSpeed does not automatically mean your site feels fast. And a score below 80 does not necessarily mean visitors are suffering. What matters is what is actually happening in the real world. And that is measured differently.

What Visitors Actually Notice

Visitors do not experience your website as a score. They experience it as three moments.

Does something appear on screen quickly? When they tap a button, does it respond? Does the page hold still while it loads, or do things keep jumping around?

These three experiences map to what Google calls Core Web Vitals.

• LCP – Largest Contentful Paint is the moment a visitor feels the page has loaded. Good is under 2.5 seconds. Poor is over 4 seconds. If your hero image takes four seconds to appear, the visitor has already decided your site is slow, even if everything else loads instantly.

• INP – Interaction to Next Paint replaced First Input Delay in March 2024. It measures the responsiveness of every interaction across the entire visit, not just the first tap. Good is 200 milliseconds or less. When INP is too high, the site feels sluggish after it loads. Visitors notice this. They just cannot name what is wrong.

• CLS – Cumulative Layout Shift measures visual stability. Good is 0.1 or below. This is what happens when a button shifts just as someone goes to tap it, or a paragraph jumps because an image loaded above it. It causes real errors; someone taps what they think is a link and hits something else entirely.

Why a 95 Score Can Still Feel Slow

A score of 95 means the lab simulation went well. But if your LCP is sitting at 3.8 seconds for real users on mobile, visitors are staring at a blank screen for nearly four seconds before anything meaningful appears. That does not feel fast, regardless of what the lab said.

Mobile scores are significantly harder to achieve because PageSpeed tests against a simulated slow 4G connection and a low-end device. Most business owners test their site on a fast laptop at home and think it loads fine. Their customer on a mid-range phone in a less connected area is having a completely different experience.

The score is a starting point. The Core Web Vitals in field data are what actually matter.

Here is a simple way to think about it. A site might load its navigation and footer within a second. But if the main image, the one a visitor actually came to see, takes four seconds, the experience feels broken. The footer loading quickly is irrelevant to someone staring at a blank hero section.

The Two Metrics Worth Focusing On

For most small business sites, two things move the needle more than anything else.

LCP first. It is the metric most directly tied to how fast your site feels. The usual culprits are unoptimised hero images, render-blocking code, and slow server response times.

Beyond three seconds, the data shows users start leaving in significant numbers. Compress your images properly, remove unnecessary code that delays the page from displaying, and make sure your hosting is not the bottleneck.

Worth noting here, page speed also affects how AI search engines evaluate and cite your content. Slow sites are less likely to be referenced by generative engines when they construct answers to user queries. If you want to understand how that works, it is worth reading what GEO is and why your content strategy needs to know about it.

INP second. A site that loads quickly but feels unresponsive is still a bad experience. Heavy JavaScript is almost always the culprit, too many scripts running in the background, third-party tools loading things visitors never asked for. Auditing what is on the page and removing what is not earning its place is usually where the gains come from.

CLS matters too, but for most small business sites, it is the easiest to fix. Declaring image dimensions in the code, loading fonts properly, and keeping late-loading scripts from pushing content around usually handles it.

What to Do If You Only Have an Afternoon

Open Google PageSpeed Insights. Run your homepage on mobile specifically, not desktop. Ignore the overall score and look at the Core Web Vitals section. Are LCP, INP, and CLS all in the good range?

If LCP is flagged, start with your hero image. Is it unnecessarily large? Is it loading before anything else can display? Compressing it and loading it efficiently is often a meaningful fix with very little technical effort.

If INP is the issue, look at your scripts. Google Tag Manager, chat widgets, social media embeds, each one adds to the load. Some are worth keeping. Some are not.

If CLS is flagged, the fix is usually simpler than it sounds. Add dimensions to your images so the browser knows how much space to reserve before they load. Check if any scripts are injecting content above existing text. Small changes here make the experience noticeably more stable.

The right way to use PageSpeed Insights is not to chase a perfect 100, but to find and remove bottlenecks degrading real user experience. What matters most is whether LCP, INP, and CLS are all in the good band at the 75th percentile of field data.

That is the actual goal. Not a number. A site that feels fast to the people using it.

The Bottom Line

PageSpeed scores are a diagnostic tool, not a finish line.

What your visitors actually experience depends on three things: how quickly the main content appears, how responsive the page feels when they interact with it, and whether the layout stays stable while it loads. A number on a screen tells you none of that on its own. Fix LCP first. Address INP second. Check CLS third. Use field data, not lab scores, to judge whether it is working.

A site that feels fast is not the same as a site that scores well. The goal is always the first one. If you are not sure where yours stands, we are happy to take a proper look.