Send half as much data for the same page

Page weight is the one thing on this site we measure end to end rather than model: bytes are bytes. Below is what happened to them across the sites we audited — what came off, what it means for your hosting bill and your visitors’ data, and where it did not work.

No signup. See your before/after PageSpeed score in about a minute.

Measured, not modelled

What we measured

Our own audits — the original page against the same page served through WebSpeed. The sample and the date are part of the result, not small print.

45 audited sites that cleared the publishing gates measured 2026-08-04

median weight cut, mobile 46.1%
total across the sample, before 293 MB
the same pages, after 117 MB
sites that got heavier 5

What this does not say. This is one load of one page per site, measured in a lab, not your monthly transfer. What it scales to depends on traffic, on how much of your audience is returning (their browsers cache), and on how much of your bill is pages at all rather than video, downloads or API traffic we never touch. It also says nothing about your origin bandwidth on its own: cached delivery changes how often we ask your server for anything, and we do not currently publish a measured figure for that — so we do not claim one.

What WebSpeed does about it

Each one is independent and can be switched off for your domain — nothing here is all-or-nothing.

Modern image formats, without touching your library

Images are most of the weight on most pages. We convert them to WebP on the fly for browsers that accept it, and animated GIFs — routinely 10–15 MB each, and the single largest thing on a page when present — to animated WebP. Your files and your markup stay exactly as they are.

Lazy loading

Send only the CSS the page uses

Stylesheets ship for the whole site and get used a screen at a time. We inline what the first screen needs and stop the rest from blocking it, so the bytes that matter arrive first and the rest arrive when they are wanted.

Critical CSS Render-blocking resources

Subset the icon fonts

A full icon font is often several hundred kilobytes per variant for a handful of glyphs actually used. We cut it down to the glyphs on the page — typically a few kilobytes.

Don’t fetch what nobody scrolls to

Images and embeds below the fold are deferred until they are approached, so a visitor who reads the top of the page never pays for the bottom of it.

Lazy loading

Bytes are the honest part

Most claims on a performance site are one step removed from what was measured. Rankings, conversions, emissions — each is a model applied to a measurement, and the model is where the overclaiming happens.

Page weight is not like that. We fetch the page as it was, we fetch it as we serve it, and we count. There is no coefficient in between, nothing to argue about, and nothing that stops being true if the industry benchmark it rested on turns out to have measured something else.

That is why this page exists as its own thing rather than as a bullet somewhere: it is the claim we can make most plainly.

What came off

The numbers above are one mobile load per site, across the audit base, gated the same way everything else on this site is gated — both devices held up, the proxy genuinely served the page rather than bouncing back to the original, the render was not degraded into a page that looks light because half of it never arrived. That last gate matters most here: a broken page is very light indeed, and without checking what was actually delivered, “we halved the weight” and “we failed to load your site” produce the same number.

Roughly half, at the median. Most of it is images — WebP instead of JPEG and PNG, animated WebP instead of GIFs that routinely run to ten megabytes each. The rest is the accumulation of small things: icon fonts cut to the glyphs in use, stylesheets that stop shipping the whole site to render one screen, below-fold media that waits until someone scrolls.

And on a handful of sites the number went the other way. That count is printed with the median, not hidden under it. Optimizing delivery means adding a little in order to remove a lot, and on a page that was already lean, there is little to remove and the small addition shows. It is a real limit of the approach, it is visible in your own before/after, and you should know it before you are asked to pay for anything.

Where the saving actually lands

For your visitors, immediately and entirely: half the data spent on your page, most sharply felt on the connections that were already struggling with it.

For your bill, less directly than a vendor would like to imply. Visitors are served from our cache, so what changes on your side is how often we come to your origin at all — which depends on your traffic pattern and how much of your transfer is pages rather than video, downloads and API traffic we do not touch. We do not publish a number for that, because we do not measure it yet. When we do, it will be on this page with the sample beside it, like everything else here.

Bandwidth & weight — questions

Will this reduce my hosting bill?

Possibly, and by less than the headline suggests. Bytes to the visitor come from our cache, not your server, so what changes on your bill is the traffic we no longer ask you for. How much that is depends on your cache hit rate and how much of your transfer is pages rather than video, downloads or API calls. We would rather set that expectation than quote you a saving we have not measured.

Does the page look worse?

No. WebP at our quality settings is visually indistinguishable at normal viewing size, and nothing is resized or re-cropped — the same image, encoded better. If you can see a difference on your own site, that is a bug and we want to hear about it.

What about visitors on limited data plans?

This is where the effect is most directly felt. Half the bytes is half the data allowance spent on your page, and it is the same halving on an old phone over a weak connection, where it matters most and where the original page hurt worst.

Why did some sites get heavier?

Because optimizing delivery means adding a little — a small runtime, inlined critical CSS — in exchange for removing a lot, and on a page that was already lean there may be little to remove. The count is on this page rather than in a footnote for that reason. Your own before/after shows which side yours lands on.

Pricing

Pick how hands-off you want it

The before/after check is free to run; it isn’t one of these plans. Choose a plan when you’re ready to make the speed-up permanent — or buy the report once and apply it yourself.

SaaS subscription

Growth

from €79 · /mo
For several sites at once
14-day free trial
  • Everything in Fast, across multiple sites
  • Priority cache — your pages stay warm
  • Support with a named contact
  • One invoice for the whole portfolio
Request access
Agencies and multi-site owners.
SaaS + setup service

Custom

from €149 · setup, then monthly
Hand-tuned for your site
  • Everything in Growth
  • A config built by hand for your site
  • What only you can change, written up for your team
  • Set up, measured and maintained by us
Get a plan for my site
We look at your site by hand first. One-time setup, then a monthly plan.

Subscriptions start with a 14-day free trial (Fast and Growth), then run for a minimum of 6 months — you commit only after the trial. Fast is priced by monthly visits; indicative launch pricing.

Downloadable report

Not ready for a subscription?

Buy the Audit once and apply it yourself — the same findings the pipeline acts on, written up as instructions.

  • The same fixes Fast applies, written up as a guide you run yourself
  • Server settings, code changes and ready-to-paste head snippets
  • Before/after PageSpeed from the free check (PDF / HTML / MD)
  • Re-issued free for a year as your site changes
€108 · one-time
Priced at a year of Fast (€9 × 12)
DIY: you apply it
Buy the audit See an example report →
Not useful? Email us within 14 days and we refund it. A document, not a service — you apply it; Custom is the same work done for you.

See your own before / after

Paste your URL for a free PageSpeed comparison on mobile and desktop. No signup, about a minute.

No signup. See your before/after PageSpeed score in about a minute.

Other outcomes: Agencies & studios , Search traffic , Conversion , Carbon footprint

By platform: WordPress , Duda , WooCommerce , Wix , Tilda , Bitrix , Squarespace