How to test website speed for a client

Knowing how to test website speed for a client properly, rather than just loading the site and guessing, is the difference between a pitch that lands and one that gets picked apart. Plenty of freelancers open a prospect's homepage, notice it takes a beat to appear, and lead with “your site is slow” on that basis alone. That works until the client asks how you know, or has a nephew who checked it on his own phone and it loaded fine. A proper test gives you a specific, repeatable answer instead of an impression.

Start with PageSpeed Insights, not a guess

Google's PageSpeed Insights is free, takes a URL and nothing else, and is the tool worth reaching for first. Paste in the client's homepage and it returns a score out of 100 for mobile and a separate one for desktop, along with a list of specific issues it found on that exact page. It is not the only speed tool that exists, but it is the one most people, including the client if they go looking, will recognise and trust, which matters when you are using the result to make a case rather than just satisfy your own curiosity.

What the score is actually measuring

The single number out of 100 is a summary, not the whole story, and it is worth understanding roughly what feeds into it before you quote it to anyone. PageSpeed Insights weighs things like how long it takes before the biggest visible piece of content appears on screen, how long the page takes to respond to the first tap or click, and how much the layout jumps around while everything finishes loading. A site can score badly because of one dominant problem, usually an oversized hero image or a chunk of script that has to run before anything else can, rather than being slow everywhere at once. That is useful to know because it means the fix is often narrower than the headline score suggests.

Reading the report properly

Resist the temptation to screenshot the top number and stop there. Scroll down and the report lists the specific opportunities it found, usually things like unoptimised images, render-blocking resources, or fonts loaded from an external service before the page can display anything. Each one names the actual file or resource causing the delay, which turns a vague complaint into a concrete line you can read back to the client: this one image is costing you this many seconds. That is a more persuasive way to open the conversation than a bare score, because it shows you looked at their site specifically rather than running a stock line past every prospect. The causes behind those flagged issues, and what typically drives them on a small business site, are covered in more depth in why slow websites quietly lose local businesses work.

Test on a real phone too

A lab report is useful but it is not the same thing as watching the site load. Open it yourself on an ordinary phone, over mobile data rather than office wifi, the way an actual customer searching for a plumber in Bolton on a Tuesday afternoon would encounter it. This catches things a score alone can miss, such as a slider that stutters into place or a booking form that takes a visible pause to appear after the rest of the page has loaded. Watching it happen is also simply more convincing to describe to a client than reading out a number, since most business owners have felt a slow page load on their own phone before, even if they never connected the feeling to their own site.

Test more than the homepage

It is easy to run one test on the homepage and assume the result applies to the whole site, but a contact page, a services page or a gallery can behave very differently, particularly if it carries more images or a booking widget the homepage does not. Running the same test against the page a visitor would actually land on, whether that is a services page linked from a Google search or the contact page a customer reaches right before enquiring, gives a more honest picture of what a real visit looks like than checking the front door alone.

Turning the result into a pitch

A speed score only earns its keep once it becomes part of a conversation with the business, not a fact you keep to yourself. Screenshot the report, or better, load the site on your phone in front of them or on a call and let them watch it happen, then point at the specific delay rather than the abstract number. That approach, and how to frame findings like this so they read as help rather than criticism, is covered more fully in using a website audit as a sales tool. A speed test on its own is one data point; it works best as part of that wider audit rather than a single fact fired off in isolation.

Running a proper speed test by hand works well for one prospect at a time, but it does not scale to checking every trade in a town before you decide who to contact first. Patchscout runs this test as part of every site it audits, alongside mobile behaviour, SSL and the rest of the checks that make up a full picture, so you can see which businesses have a genuine speed problem before you open a single page yourself. The three free searches at app.patchscout.co.uk/signup are enough to see it against a trade and town you already know.