Website speed optimization.
Overview
A slow website costs you in two places. Visitors leave before the page has finished loading, most of them on a phone and many of them paying for their data. And Google measures how real visitors experience your pages, through Core Web Vitals, and uses that as one of its ranking signals.
Speed work is also one of the few parts of a website where the result is plainly measurable. A page either appears in 1.6 seconds or in 6. So we measure before we touch anything, fix the things that cost the most first, and measure again, so you can see exactly what changed.
Where the time goes
Most slow sites are slow for a handful of reasons, and they’re rarely the ones a speed plugin can fix on its own.
Before the page even starts
- Hosting that takes a second or more to send the first byte, before anything else can load
- No page caching, so every visit rebuilds the page from the database
- An out-of-date PHP version, noticeably slower than a current one as well as unsupported
- Database tables clogged with years of leftovers from removed plugins
- No CDN, so visitors far from the server wait on every file
- Hosting that copes with a normal day but not with a busy one
While the page loads
- Images uploaded straight from a camera or phone, several megabytes each
- A page builder or theme loading the code for every feature on every page
- Chat widgets, trackers and social feeds, each one adding requests
- Fonts that block the text from appearing, or that load in five weights nobody uses
- Adverts and embeds that push the page around as they load
- Plugins loading their scripts site-wide when they are used on one page
Real visitors, not just a score
The score from Google’s PageSpeed test is a useful starting point, but it’s a single simulated visit on a throttled connection. What Google actually uses is data from real visitors in Chrome, collected over 28 days. We work from both: the lab test to find the causes, and the real-visitor numbers to judge whether the fix worked for the people who matter.
How we speed up a site
- 01
Measure
Step 1We record where the site stands today: real-visitor Core Web Vitals from Search Console, lab tests on a mid-range phone, server response times, and the page weight of your most visited templates. These are the numbers we will be judged against.
A baseline you can check - 02
Diagnose
Step 2For each slow template we work out where the time goes: the server, the database, the images, the scripts or the fonts. Every finding is ranked by how much time it saves against how much work it takes, so the biggest wins come first.
A ranked list of fixes - 03
Fix the server side
Step 3Page caching, a current PHP version, an object cache where the site needs one, a cleaned-up database and a CDN in front. Changes are made on a staging copy first and checked before they reach the live site.
A fast first byte - 04
Fix the front end
Step 4Images resized and served in modern formats, scripts deferred or removed, fonts trimmed, and space reserved for adverts and embeds so the page stops jumping. Nothing a visitor uses is taken away without asking you first.
Pages that appear quickly and stay still - 05
Keep it fast
Every monthSites slow down again as plugins update and new content goes up. On a care plan we check the numbers every month and fix a regression before visitors notice it.
Speed that lasts
We measure before and after on the same tests, so you can see exactly what changed.
What we will not do
We won’t stack three caching plugins on top of each other and call it done. We won’t chase a perfect 100 by removing things your visitors actually use. And we won’t promise a score, because some of what slows a site down, like a required advertising network, may be yours to decide on rather than ours to remove. We’ll show you what each one costs and let you choose.
What you get
- A before-and-after report on the same tests, with real-visitor data where the site has enough traffic
- Hosting and caching set up properly, or a clear recommendation if the hosting is the problem
- Images, fonts and scripts optimized on the templates that get the most visits
- Layout shift fixed, so buttons and text stop moving as the page loads
- Every change made on a staging copy first, then checked on the live site
- Monthly monitoring on a care plan, so the site stays fast
Questions we get asked
Can a speed plugin fix it?
A good caching plugin helps, and we often use one. But if the hosting takes two seconds to respond, the images are enormous or a page builder loads everything on every page, a plugin is working around the problem rather than fixing it. The biggest gains usually come from the causes underneath.
Do we need to change hosting?
Sometimes, but not always. We measure the server first. If it responds quickly once caching is set up properly, your hosting is fine. If it is slow even for a cached page, we will tell you, recommend something that suits your traffic, and can handle the move without downtime.
Why does my PageSpeed score change every time I test?
The lab test simulates one visit, and small differences in network and server timing move the score from run to run. That is why we compare the median of several runs, and why the real-visitor data in Search Console matters more than any single score.
Does speed affect Google rankings?
Yes, but as one signal among many. A fast page will not outrank a slow one that answers the question much better. Speed matters most where the competing pages are similar, and it always matters to the visitors who give up waiting.
Will the site look or work differently afterwards?
It should look the same and simply load faster. Where a feature is expensive, like a slider on the home page or a third-party widget, we show you what it costs and you decide whether it stays.
Can you speed up a site built with Elementor or another page builder?
Yes. Page builders add weight, but there is usually plenty to gain from caching, images, fonts and unused scripts without rebuilding anything. If the builder itself turns out to be the main cost, we will say so and explain what a rebuild would gain.