Home / Journal / How fast should a website load?
How fast should a website load?
Under 2.5 seconds. That is the threshold Google uses: your largest piece of visible content should finish rendering within 2.5 seconds on a mid-range phone over 4G. Past 4 seconds, Google classes the page as poor — and most of the people who tapped your link have already gone.
What counts as fast in 2026?
Google measures three things, collectively called Core Web Vitals. Each has a threshold for good, one for needs improvement, and everything beyond that is poor. These are measured on real visits to your site, not in a lab.
| Metric | What it measures | Good | Poor |
|---|---|---|---|
| Largest Contentful Paint (LCP) | When the main content appears | Under 2.5 s | Over 4.0 s |
| Interaction to Next Paint (INP) | How fast the page responds to a tap | Under 200 ms | Over 500 ms |
| Cumulative Layout Shift (CLS) | How much the page jumps while loading | Under 0.1 | Over 0.25 |
To pass, 75 per cent of your real visits must hit the good threshold. An average is not enough — a fast experience for half your visitors and a miserable one for the rest is a fail.
What does a slow website actually cost?
The most-cited figure comes from Deloitte's Milliseconds Make Millions study with Google in 2020: improving mobile load time by just 0.1 seconds lifted retail conversions by 8.4 per cent and travel conversions by 10.1 per cent. Google's own earlier research found that 53 per cent of mobile visits are abandoned when a page takes longer than three seconds.
Both numbers are a few years old and both come from companies with an interest in fast websites, so treat them as direction rather than gospel. The direction has never been in dispute. Speed is not a technical nicety — it is the first thing that decides whether anyone sees your work at all.
Why is my website slow?
We run this test on a lot of sites. Four causes account for nearly all of it.
1. Images sent at full size to phones
A 3000-pixel-wide photograph downloaded in full and then squeezed into a 375-pixel screen. The visitor pays for every pixel they never see. Converting to WebP or AVIF and serving correctly sized versions typically halves total page weight with no visible difference.
2. JavaScript nobody uses
Plugins, tracking tags, chat widgets, an A/B testing tool somebody installed once and forgot. Every one of them is downloaded, parsed and executed before your page settles. On a mid-range Android this is usually the single biggest cost.
3. Render-blocking stylesheets and scripts
The browser has everything it needs to show your page but is made to wait for files that are not needed yet. Deferring what is not required for the first screen usually buys back a second or more.
4. Slow server response
If your host takes 800 ms to send the first byte, you have spent a third of your budget before the browser has drawn anything. Cheap shared hosting on a single oversold server is the usual culprit. A static site on a CDN answers in tens of milliseconds, from the country the visitor is in.
The order matters. Compress the images first, remove the scripts you do not need second, and only then argue about hosting. Most sites we test could halve their load time without touching a line of code.
How do I test my own website?
Run it through Google PageSpeed Insights, which uses the same Lighthouse engine Google uses to judge you. It is free and it is the source of truth.
It also produces a wall of jargon. That is why we built the free audit on our homepage — it runs the same test and hands back the five things we would fix first, written in English, with no email address required.
Frequently asked questions
How fast should a website load in 2026?
Your largest piece of content should appear within 2.5 seconds on a mid-range phone on 4G. That is Google's threshold for a good Largest Contentful Paint. Between 2.5 and 4 seconds needs improvement; over 4 seconds is classed as poor.
What are Core Web Vitals?
Three measurements Google uses to score real-world experience: Largest Contentful Paint (how soon the main content appears), Interaction to Next Paint (how quickly the page responds when tapped), and Cumulative Layout Shift (how much the page jumps around while loading).
Do Core Web Vitals affect Google rankings?
Yes, but as a tie-breaker rather than a main factor. Google has been clear that relevant content outranks a fast empty page. Where two pages are comparably useful, the faster one wins, and speed also affects how many visitors stay long enough to read either.
Why is my website slow?
In our experience four causes account for nearly everything: oversized images sent at full resolution to phones, JavaScript from plugins and tracking tags that is downloaded but never used, stylesheets and scripts that block the first paint, and slow server response from cheap shared hosting.
How do I test my website speed?
Run it through Google PageSpeed Insights, which uses the same Lighthouse engine Google uses to judge your site. The audit tool on the Howdy homepage runs that test and translates the result into plain English, free and without an email address.
Does page speed matter more on mobile?
Considerably. Most traffic is on phones, phones have slower processors and less reliable connections, and Google's index is mobile-first, so the mobile version of your page is what gets measured and ranked.
Want to know where yours stands?
Paste your address into the audit on our homepage. Twenty seconds, no email, and you keep the answer whether or not you ever speak to us.
Run the free audit →