Last updated: October 2026.
Website speed optimisation is the work of making every page load and respond fast enough that people stay and Google trusts the experience. DigiFix builds and maintains sites for Australian businesses, and this guide covers what speed is worth, how Google scores it and which fixes come first.
Key takeaways
- Google sets three Core Web Vitals thresholds: LCP within 2.5 seconds, INP of 200 milliseconds or less and CLS of 0.1 or less, judged at the 75th percentile of real page loads (web.dev, Web Vitals).
- Interaction to Next Paint replaced First Input Delay as a Core Web Vital on 12 March 2024, so any guide still quoting FID is describing a retired metric (web.dev, Chrome team announcement).
- Speed is a tiebreaker, not a trump card: Google states it always seeks the most relevant content even when page experience is sub-par (Google Search Central, Understanding page experience, updated 22 September 2026).
- Mobile is where the problem hides. StatCounter measured mobile at 48.57 per cent of Australian platform share in September 2026, against 48.48 per cent desktop and 2.95 per cent tablet (StatCounter Global Stats, September 2026).
- Brotli has replaced gzip as the compression to ask for, and it is supported by every current browser (web.dev, Reduce network payloads with text compression).
Definition: website speed optimisation is the practice of reducing the time between a visitor tapping your link and your page being usable, then keeping that time low as content and plugins are added. It covers hosting, caching, images, third-party scripts, compressed delivery and the code the browser has to process before anything appears.
Why does website speed affect your leads?
Two separate things are happening at once. The first is human: a slow page loses people before they read your offer, and the faster the connection the less patience they have for a slow site. The second is mechanical: Google measures how real visitors experienced your pages and uses that as one signal among many when choosing between pages that answer a query equally well.
Google is explicit that relevance comes first. Its page experience guidance says Search always seeks to show the most relevant content, even if the page experience is sub-par, but that where several pages are equally helpful, a great page experience can contribute to success (Google Search Central, Understanding page experience, updated 22 September 2026). Read that carefully and the strategy writes itself: speed rarely rescues a thin page, and it regularly decides a close contest.
The revenue side is better documented than most marketing claims. web.dev’s roundup of published case studies records the BBC losing an additional 10 per cent of users for every additional second their site took to load, Rakuten 24 lifting revenue per visitor by 53.37 per cent after core web vitals work, and Vodafone turning a 31 per cent improvement in LCP into an 8 per cent sales increase (web.dev, Why does speed matter?). Those are large businesses on large traffic, so treat the percentages as direction rather than forecast.
What it means for a small business: you do not need a 53 per cent revenue lift. If your contact form converts at 3 per cent and a faster page keeps one more visitor in twenty, the same ad budget produces more enquiries. Speed is a multiplier on traffic you already paid for.
How does Google measure website speed?
Google measures three things, together called Core Web Vitals, and it scores them from real visits rather than from a lab test. All three are judged at the 75th percentile of page loads, which means three quarters of your visits must hit the target. A good average with a bad tail still fails (web.dev, Web Vitals).
| Metric | What it times | Good | Poor |
|---|---|---|---|
| LCP Largest Contentful Paint | How long until the biggest element in view is drawn | 2.5 seconds or less | Over 4.0 seconds |
| INP Interaction to Next Paint | The delay between a tap or click and the screen responding | 200 milliseconds or less | Over 500 milliseconds |
| CLS Cumulative Layout Shift | How much the page jumps around while it loads | 0.1 or less | Over 0.25 |
The middle band between good and poor is labelled needs improvement, and it is where most small business sites sit. Thresholds above are the ones Google publishes on its own Core Web Vitals page (Google Search Central, Core Web Vitals).
LCP in plain English. When a visitor lands, the browser paints the page in stages. The largest block of content in the first screen, usually your hero image or main heading, is the LCP element. If it takes four seconds to appear, the visitor spent four seconds looking at nothing useful.
INP in plain English. INP times the gap between a real interaction and the next frame the browser draws. A page that looks loaded but freezes for a second when someone taps the menu has a poor INP. First Input Delay measured only the first interaction and only the input delay, so Google replaced it with INP on 12 March 2024 (web.dev, Interaction to Next Paint becomes a Core Web Vital on March 12).
CLS in plain English. CLS adds up how far visible elements move unexpectedly. An ad slot that loads late and shoves your headline down counts against you, and it is the metric most often broken by a chat widget or a cookie banner.
How do you test website speed in five minutes?
Three free Google tools answer three different questions, and knowing which is which stops you chasing the wrong number. Field data is what real visitors experienced. Lab data is a single simulated visit on a throttled connection. Only field data feeds the Core Web Vitals assessment.
- PageSpeed Insights at pagespeed.web.dev shows both. The top block is field data from the Chrome UX Report for your real visitors over the last 28 days. The bottom block is a Lighthouse lab audit. Fix the field data problems first (PageSpeed Insights).
- The Core Web Vitals report in Search Console groups your URLs into good, needs improvement and poor, by device type. This is the report that mirrors what Google records against your site (Search Console Help, Core Web Vitals report).
- Chrome DevTools with device emulation and network throttling shows you the page as a mid-range phone on a slow connection experiences it, which is where LCP problems usually reveal themselves.
Send us your PageSpeed Insights link and we’ll tell you the 3 fixes that matter most. No audit document, just the three changes with the biggest effect on your enquiries, ranked.
What slows a website down?
Almost every slow WordPress site we open has the same four causes, in roughly the same order of impact. The symptom tells you which one you are looking at before you open a single file.
| Symptom | Likely cause | First fix to try |
|---|---|---|
| Time to first byte over 600 milliseconds on every page | Slow shared hosting, no page cache, or the server sits overseas | Turn on full page caching, then move to an Australian or CDN-fronted host |
| Hero image arrives last, LCP over 4 seconds | Multi megabyte JPEG or PNG served at full size, lazy loaded by mistake | Convert to WebP or AVIF, size it to the slot, never lazy load the LCP image |
| Page looks loaded but taps do nothing for a second | Long JavaScript tasks from sliders, chat widgets, tag managers and page builders | Remove the plugin, or defer the script and load the widget on interaction |
| Content jumps as ads or embeds appear | Images and iframes without width and height, fonts swapping late | Reserve the space with explicit dimensions, preload the main font |
| Slow only for logged out visitors after a few months | Bloated database, thousands of autoloaded options, revisions piling up | Clean revisions and transients, review autoloaded options, keep plugins current |
One piece of old advice deserves retiring. Combining every CSS and JavaScript file into one request was correct for HTTP/1.1, which allowed a handful of connections at a time. HTTP/2 multiplexes parallel requests over one connection, so concatenating files now costs you caching granularity for no gain (MDN, Evolution of HTTP). The same applies to GZIP: it still works, but Brotli compresses text assets further and every current browser advertises support for it (web.dev, Reduce network payloads with text compression).
Which fixes make the biggest difference first?
The five stages of a page load. Stage definitions follow MDN’s guide to the critical rendering path and Google’s explanation of Core Web Vitals.
Order matters more than effort. A CDN in front of a slow origin hides nothing, and compressed files that are still four megabytes are still four megabytes. Work down this list and stop when the numbers are good, because the bottom items cost the most for the least.
- Hosting and server location. An Australian audience served from a European data centre pays 200 to 300 milliseconds before a single byte of content arrives. Choose a host with Sydney or Melbourne presence, or front the site with a CDN.
- Caching. Full page caching turns a database round trip into a static file. This is the single highest return change on a WordPress site and usually a plugin setting.
- Images. Serve WebP or AVIF instead of JPEG and PNG, size each image to the slot it fills, and use srcset so phones download a smaller file (web.dev, Serve responsive images). Images are consistently the largest transfer on a content page.
- Third-party scripts. Every chat widget, pixel, heatmap and review widget is a separate company’s code running on your page. Audit them quarterly and delete what nobody reads.
- CDN and Brotli. A CDN puts your files close to the visitor and compresses text on the way out. Both are configuration, not development.
- Code. Minify CSS and JavaScript, defer what is not needed for the first screen, and remove unused plugins rather than optimising them.
- Database. Clear revisions, expired transients and orphaned tables. Do this last, because it rarely moves a metric on a small site.
Want the numbers before you book anything? Our free checker grades any URL on load time, page weight and mobile layout, with no email required.
If you are on WordPress and want the click by click version of each step, we keep that in a separate walkthrough: mastering page load speed for better UX. This guide is the strategy; that one is the spanner.
Why is mobile speed the one that counts?
Because the phone is where your slowest visitor is, and the gap between desktop and mobile on the same page is usually two to four times. A mid range Android phone has a fraction of a laptop’s processing power and often a weaker connection, so the same JavaScript takes far longer to execute. StatCounter put mobile at 48.57 per cent of Australian platform share in September 2026, with desktop at 48.48 per cent and tablet at 2.95 per cent, which means roughly half your visitors meet your site on the harder device (StatCounter Global Stats, September 2026).
That is also why Google assesses the metrics separately for mobile and desktop and reports them separately in Search Console. If your desktop numbers are green and mobile is amber, you have a real problem that a fast office connection will never show you. Test on an actual phone, on mobile data, not on your desk wifi.
What does this look like for an Australian business?
A Melbourne professional services firm we would typically see: eight page WordPress site, a page builder, two sliders, a chat widget, a booking embed and images straight from the photographer. Mobile LCP sits at 5.4 seconds, INP at 340 milliseconds, CLS at 0.19. Enquiries are steady but the cost per enquiry has crept up for two quarters.
What we would do, in order. Week one: turn on page caching and move the site behind a CDN, which typically takes mobile LCP under 4 seconds on its own. Week two: convert the hero and gallery images to WebP at the right dimensions, remove both sliders in favour of a static hero, and defer the chat widget until interaction. Week three: clean the plugin list, replace the booking embed with a link, add explicit dimensions to every image and iframe to settle CLS, then re-measure. Four weeks after that, one maintenance check.
What waiting costs. Not a cliff, a slope. Every month at 5.4 seconds keeps taxing paid traffic that already converted once into a click, and competitors who fix theirs take the comparison. The work above is three weeks of part time effort and no redesign.
Our website care plans exist for exactly this. Caching, image optimisation, plugin updates, monthly Core Web Vitals checks and a plain English report, so the site does not drift back to slow.
How do you keep a website fast each month?
Sites get slower on their own, because every new plugin, campaign tag and uploaded image adds weight that nobody removes. A twenty minute monthly routine is enough to hold the line.
- Open the Core Web Vitals report and note any URL that moved into needs improvement.
- Check the plugin list for anything installed and deactivated, then delete rather than deactivate.
- Look at uploaded images from the past month and compress anything over 200 kilobytes.
- Review third-party scripts against the campaigns actually running.
- Run PageSpeed Insights on the homepage, the top service page and the contact page.
- Confirm backups ran, and that the site is on a supported PHP version, since both affect performance and security together. Our notes on WordPress security risks for small business cover the same upkeep list.
When should you rebuild instead of fix?
Fix first, almost always. A rebuild is the right call in four situations: the theme or builder is abandoned and no longer receives updates, the plugin stack has grown past what the site actually uses, the design cannot be made responsive without duplicating every template, or the hosting contract cannot deliver a reasonable time to first byte at any price. If none of those apply, you are buying a new site to solve a caching problem.
When a rebuild is genuinely the answer, we build on a lightweight stack with performance as a design constraint rather than an afterthought, which is what our custom website speed and SEO service covers. You can see the standard we build to at websites.digifix.com.au, and measure any URL you like with the free checker at seotools.digifix.com.au.
Not sure whether to fix or rebuild? Send us the URL. We will tell you which one the numbers support, including when the answer is leave it alone.
What else do people ask about website speed?
Is website speed a ranking factor in Google?
Yes, with a caveat. Google states that Search always seeks the most relevant content even when page experience is sub-par, and that a great page experience can contribute to success where several pages are equally helpful (Google Search Central, Understanding page experience, updated 22 September 2026). So speed decides close contests rather than lifting an irrelevant page. It will not rescue a page that fails to answer the query, and it will beat a competitor who answers it just as well but loads in five seconds.
What is a good page load time in seconds?
Judge it on LCP rather than on total load time. Google’s threshold for a good LCP is 2.5 seconds or less at the 75th percentile of real page loads, with anything over 4.0 seconds rated poor (web.dev, Web Vitals). Total load time is a misleading number because the page can be perfectly usable long before the last tracking pixel fires. If your LCP is under 2.5 seconds on mobile field data, you are in the good band and further work has diminishing returns.
Why is my website slow only on mobile?
Usually because the same JavaScript bundle runs on a device with a fraction of the processing power, and because mobile networks add latency that office wifi hides. Sliders, page builders and chat widgets all cost more on a phone. Test on a real handset on mobile data, then look at INP specifically: a slow-to-respond page is nearly always a long JavaScript task rather than a large image. Removing one unused plugin often beats months of tuning.
Does a CDN fix a slow website?
It fixes delivery, not computation. A CDN caches your static files at edge locations near the visitor, which cuts latency for images, CSS and JavaScript, and it can serve a cached HTML page for anonymous visitors. What it cannot do is make a heavy theme render faster on the visitor’s phone, or speed up a page that must query the database for every request. If time to first byte is high but the CDN is in place, the origin is the problem.
How much does website speed optimisation cost in Australia?
It scales with how much of the site needs to change. Caching, compression and image conversion are configuration work and are usually a few hours. Replacing a page builder or removing a plugin stack that the design depends on is development work and can run to days. Ask for the fix list before the price, then ask which three items account for most of the gain, because on most small business sites two or three changes do the heavy lifting and the rest is tidying.
Can a fast website still rank badly?
Easily, and it happens constantly. Speed is one input among many, and a fast page that targets a phrase nobody searches, or answers a question a different page answers better, will not rank. The reverse is also true: a slow page with genuinely useful content can rank well, which is why Google’s own wording puts relevance ahead of page experience. Treat speed as the multiplier you apply after the content and the search intent are right.
DigiFix is a Melbourne web design and digital marketing team. You can see the sites we have built in our portfolio, read about how we work, or browse the DigiFix blog for more guides. Our digital marketing services cover the build, the speed work and the ongoing measurement together. If mobile layout is the part you are unsure about, start with mobile optimisation for mobile users.