Yes, if it is slow on a phone, and most are. On 19 October 2026 we ran Google's Lighthouse mobile audit on the home pages of 24 web design and marketing agencies in the Illawarra, our own included. The median performance score was 62 out of 100. Nineteen of the 24 took longer than Google's 2.5-second target to show their main content, and 17 took longer than four seconds. If the people who sell websites for a living score like that, the average local business site is slower. This article explains what that costs you in enquiries, with the evidence, how to test your own site in ten minutes and read the result, and what the fixes usually are.
Does website speed really affect conversions?
Yes, measurably. The most careful public study is Milliseconds Make Millions, run by Deloitte for Google in 2020 across retail, travel, luxury and lead generation sites in Europe and the US. A 0.1-second improvement in mobile site speed lifted retail conversions by 8.4%, travel conversions by 10.1%, and for lead generation sites, the ones most like a service business, improved the bounce rate on information pages by 8.3%. That is one tenth of a second. The gap between a 2.5-second site and a 7-second site is forty-five of them.
Google's own case studies say the same in business terms: Vodafone improved its largest-content load time by 31% and saw 8% more sales; redBus cut layout shift from 1.65 to zero and saw mobile conversion rates rise 80 to 100% across its properties. The mechanism is not mysterious. A visitor on a phone with a search result open gives a page a few seconds before going back to the results and tapping the next business.
What is a good page speed?
Google's Core Web Vitals set the thresholds, measured for real visitors at the 75th percentile:
- Largest Contentful Paint (LCP), how long until the main content shows: 2.5 seconds or less.
- Interaction to Next Paint (INP), how quickly the page responds to a tap: 200 milliseconds or less.
- Cumulative Layout Shift (CLS), how much the page jumps around while loading: 0.1 or less.
Google recommends meeting all three "for success with Search", and its page experience signals favour pages that do. Rankings are the smaller half of the story, though. Our earlier article covers site speed as a ranking factor; this one is about the visitors you already have.
What we found on 24 Illawarra agency websites
We tested each home page once with Lighthouse 12 on a simulated mid-range phone and slow 4G connection, the same method as Google's PageSpeed Insights lab test. Single lab runs vary, so treat individual scores as approximate and the medians as the finding.
- Median performance score: 62 out of 100. Two sites scored under 50, eighteen scored between 50 and 89, and four scored 90 or more.
- Median time to show the main content: 6.5 seconds. Nineteen of 24 missed Google's 2.5-second target; 17 took longer than four seconds; the slowest took 29 seconds.
- Median page weight: 2.5 MB. Eight home pages weighed more than 3 MB; the heaviest was 47 MB, almost all of it video.
- Layout shift: three sites moved their content around enough to fail the 0.1 threshold, one scoring 0.88, which means buttons jumping away from thumbs.
The pattern behind the numbers was consistent: full-screen background videos that start downloading before anything else, hero images served at desktop size to phones, and page builders loading a dozen scripts before the first word appears. None of it is exotic. All of it is fixable, and the four sites that scored 90 or better were not the ones with the biggest budgets; they were the ones with the least in the way.
How to test your own website in ten minutes
1. Open PageSpeed Insights (pagespeed.web.dev), paste your home page address and run it. Use the Mobile tab; that is where your enquiries come from. 2. Look for the real-visitor data first. If your site has enough traffic, the top section shows what actual visitors experienced over the last 28 days for LCP, INP and CLS, with a pass or fail against the thresholds. This is the number that matters. If it says there is not enough data, use the lab score below it. 3. Read the lab score as a band, not a mark. Under 50 is a problem costing you enquiries. 50 to 89 is typical and improvable. 90 and above is good. Run it three times; lab scores move by ten points between runs. 4. Scroll to the opportunities. The report names the heaviest images, the scripts that block rendering, the elements that shift, and roughly how much time each would save. The top three items are usually most of the problem. 5. Test the page that earns the money. Your contact page, your most-visited service page and a blog post. Home pages are often the best-tuned page on a site and the worst guide to the rest.
If you want the report read for you, we do that as part of a site health check.
Why websites are slow: the usual five causes
Oversized images and videos. The single biggest cause in our sample. A 4,000-pixel photo scaled down in the browser, or a hero video that starts downloading on a phone that will never show it well. Fix: properly sized, compressed images in modern formats, videos that load after the page, posters instead of autoplay on mobile.
Page builders and plugin weight. On WordPress particularly, a visual builder plus a dozen plugins can load 50 to 100 separate files before the page is usable. Fix: fewer plugins, a lighter theme, caching, or a rebuild on something that ships less.
Slow hosting far away. A $5-a-month shared host in another country adds a second before the page even starts. Fix: managed hosting with Australian servers or an edge network such as Cloudflare that serves the page from Sydney.
Scripts loading before content. Chat widgets, tracking tags, fonts and sliders that all insist on loading first. Fix: defer anything the visitor does not need in the first second; load tracking after the page.
Layout shift. Images without dimensions, late-loading fonts and banners that push content down. Fix: reserve space for everything, which is a build-quality issue rather than a hosting one.
What fixing it costs, and when a rebuild is cheaper
Image compression, caching and script deferral on an existing site is usually a day or two of work, often covered by a care plan, and it can move a site from the 40s into the 70s. Getting from the 70s to the 90s on a page-builder WordPress site is harder, because the weight is structural. At that point a rebuild on a lean theme, on Webflow, or on a modern static framework often costs little more than the optimisation work and removes the problem rather than managing it; our website redesign page explains how we scope that decision. Our own site was rebuilt on Astro with a headless CMS in October 2026 for exactly this reason, and in the same test its service and article pages load their main content in about three seconds on the simulated phone, with desktop scores of 98 to 100.
Frequently asked questions
How much does a slow website cost a business?
It depends on your enquiry value, but the Deloitte study's figure for lead generation sites, an 8.3% improvement in bounce rate for a 0.1-second gain, gives a way to estimate. If 100 people a month visit a service page and a slow site loses an extra 20 of them before the page shows, and one in ten of those would have enquired, that is two lost enquiries a month. For a trade or clinic each enquiry is worth hundreds or thousands of dollars.
What is a good PageSpeed Insights score?
Ninety or above is good, 50 to 89 needs improvement and below 50 is poor, which are Google's own bands. The real-visitor Core Web Vitals assessment at the top of the report matters more than the lab score: a site can score 70 in the lab and still pass for real visitors, or the reverse. Aim to pass all three vitals for real visitors first.
Does website speed affect Google rankings?
Yes, as part of Google's page experience signals, though content relevance matters more. Google recommends good Core Web Vitals for success in Search and uses them to help choose between otherwise similar pages. Speed also affects rankings indirectly: slow pages lose visitors, which weakens the engagement signals Google can see.
Is my website slow or is it just my internet connection?
Test it from outside your own connection with PageSpeed Insights or Lighthouse, which simulate a mid-range phone on a slow 4G connection, and look at the real-visitor data if your site has it. If the lab test shows the main content taking more than four seconds, the site is slow regardless of your office Wi-Fi.
Can a slow WordPress site be fixed without rebuilding it?
Often, yes. Image optimisation, caching, removing unused plugins, deferring scripts and better hosting can halve load times on many WordPress sites. Where the theme or page builder is the weight, those fixes plateau in the 60s or 70s, and a rebuild on a lighter base becomes the cheaper path to a fast site.
Find out what your site is costing you
Run the ten-minute test above, or send us the address and we will read the report with you, show the three biggest fixes and what each would cost. Ask for a speed check, or see what a web care plan covers each month.
