Speed up your site to win back search traffic
Page speed affects rankings, but not evenly across the scale: the sharp edge is the “Poor” band, and most advice never says so. Below is what we measured on our own audit base — what moved, by how much, and the part the numbers do not prove.
That doesn’t look like a website address. Enter a domain — for example example.com or https://example.com/page.
No signup. See your before/after PageSpeed score in about a minute.
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
What this does not say. These are lab scores, and a lab score is not traffic. We measured what Google’s ranking signal reads — Core Web Vitals, on the sites we audited — not rankings and not sessions. Nobody can honestly promise the second from the first: speed is one signal among many, and a page that is fast and irrelevant still loses to a page that is slow and exactly right. What the numbers below support is narrower and worth more: the technical reason a page is held back can be removed, and we can show it being removed.
What the industry benchmark says
Quoted in the unit its source used, linking the page that states it. An industry benchmark is a direction, not a forecast for one site — the per-site numbers are the audits above, and yours is a free check away.
What WebSpeed does about it
Each one is independent and can be switched off for your domain — nothing here is all-or-nothing.
Get the page out of the “Poor” band
The band is where the ranking treatment bites hardest, and it is usually not one big fault but a pile of small ones: render-blocking stylesheets, scripts running mid-paint, the hero image queued behind assets nobody sees. We fix them in delivery, so nothing in your source changes.
Make the largest paint arrive first
LCP is the vital most audits fail on. We find the element the visitor is actually waiting for and give it priority — preloaded, high fetch priority, never lazy — instead of letting it queue behind third-party scripts.
Stop the layout jumping
CLS counts against ranking and is invisible in a screenshot. We reserve the real space for images from their measured dimensions rather than guessing, so the page settles instead of shifting under the reader’s thumb.
Keep crawlers seeing the same page
Everything above happens in delivery, and visitors and crawlers get the identical document. Nothing is cloaked, nothing is served conditionally by user agent — which is what makes this a technical fix rather than a risk to your indexing.
The band is what matters, not the number
Most speed advice treats the PageSpeed score as a dial: higher is better, keep turning. That is true in the sense that no one prefers a slower page, and misleading in the way that costs the most time — because the scale is not flat. The consequential distinction is whether a page sits in the “Poor” band, where Google’s treatment of Core Web Vitals is at its harshest.
That is also the band where the fixes are cheapest. A page scoring in the thirties is rarely suffering from something subtle. It is usually a stylesheet blocking the first paint, a set of scripts executing while the browser is trying to draw, and the one image the visitor came for queued behind assets that are nowhere near the screen. None of those require a redesign, and none of them are visible to the person reading the page — which is exactly why they survive for years.
What we found on our own base
The figures at the top of this page come from our own audits, and the sample is stated with them for a reason: they describe the sites we measured, not the web. Every site in it cleared the gates that keep a measurement honest — both mobile and desktop held up, the proxy was genuinely serving the page rather than redirecting back to the original, and the render was not degraded into a page that scores well because half of it never loaded.
Within that sample, the median mobile score moved from the low fifties into the seventies. The result worth pausing on is narrower: of the sites that began in the “Poor” band, every one of them left it. Not because the band is easy to leave in general, but because the faults that put a page there are mechanical, and mechanical faults are the kind a delivery layer can fix without touching what you publish.
What this does not tell you
It does not tell you what happened to anyone’s traffic, and we have not measured that. Search position depends on relevance, links, intent, competition and a long list of other things, and speed is one input among them. A page that nobody wants to read does not start ranking because it paints quickly.
What the measurement does support is the claim we are actually making: if a technical bottleneck is holding a page below the threshold Google reads, it can be removed, and you can see it removed on your own site before deciding anything. The check is free and takes about a minute. If your site is already out of the “Poor” band, the honest answer is that the ranking upside here is small — and the report will say so rather than sell you a plan.
Search traffic — questions
Does page speed actually affect Google rankings?
Yes — Core Web Vitals have been a confirmed ranking signal since 2021. The effect is not linear across the scale, though. The meaningful move is out of the “Poor” band, where Google can hold pages back in favour of faster competitors; going from good to excellent buys far less.
Will faster pages bring me more traffic?
Not on their own, and anyone promising that is guessing. Speed removes a technical reason you are held back; it does not make a page more relevant to a query. Where speed does show up reliably is on pages that already rank and already get impressions — there, a faster page loses fewer of the visitors it has already earned.
How long until rankings change?
Google has to re-crawl and re-assess the pages, and Core Web Vitals in Search Console are reported on 28 days of field data — so changes show up there over weeks, not days. Our before/after is a lab measurement, available in about a minute, and it tells you whether the technical problem is gone. The two answer different questions.
Is this the same as cloaking?
No. Cloaking is serving different content by user agent. We serve the same document to everybody, optimized in delivery — the difference is how the page is packaged, never what it says.
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.
Fast
- The whole optimization pipeline, live
- One site
- Proxy or on-site SDK, no code changes
- Cached delivery keeps your site up during origin outages
- Priced by monthly visits, scales with you
Growth
- Everything in Fast, across multiple sites
- Priority cache — your pages stay warm
- Support with a named contact
- One invoice for the whole portfolio
Custom
- 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
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.
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
See your own before / after
Paste your URL for a free PageSpeed comparison on mobile and desktop. No signup, about a minute.
That doesn’t look like a website address. Enter a domain — for example example.com or https://example.com/page.
No signup. See your before/after PageSpeed score in about a minute.
Other outcomes: Agencies & studios , Conversion , Bandwidth & weight , Carbon footprint
By platform: WordPress , Duda , WooCommerce , Wix , Tilda , Bitrix , Squarespace