Speed and technical SEO

A slow site on mobile: where the problem comes from and how to read the PageSpeed report

Published approx. 6 min read

A phone showing a slowly loading page and a laptop with a speed report with colour-coded indicators, on a wooden desk.
Editorial illustration generated with AI

The question

Our site opens slowly on mobile, and the PageSpeed report has dozens of recommendations we don't understand. Where does the problem come from, and what is worth fixing first?

A site that loads slowly on a phone loses visitors before they see the offer. Many site owners run a PageSpeed Insights test, get a low score and a long list of technical recommendations, and don't know what really matters or who to ask.

This guide explains what the report actually measures, why results differ between phone and computer or from one run to the next, what the usual causes of a slow site are and how to decide the order of interventions. The goal is not a perfect score, but a site that opens well for the people who use it.

If you want us to handle this for you, see our service Google PageSpeed analysis and optimisation or our service Web and cloud hosting.

What PageSpeed Insights measures

PageSpeed Insights is a free tool from Google that analyses a single page at a time. The report usually has two different parts, which are often confused.

The first part shows data from real visitors, collected from the Chrome browsers of people who visited the site, when Google has enough such data. The second part is a lab test run with Lighthouse: a repeatable simulation on a device and connection set by the tool.

The real-world data shows how visitors experience the site. The lab test is useful for finding the causes. A new site or one with little traffic may not have real-world data yet, in which case the report relies on the simulation alone.

Core Web Vitals, explained simply

Core Web Vitals are the measurements Google uses to describe the loading and usage experience of a page. The thresholds considered good are published and updated by Google in its official web developer documentation, so they are worth checking there rather than taking from older articles.

There are currently three such measurements:

  • LCP (Largest Contentful Paint): how long it takes for the main element of the page to appear, usually the large image or the heading at the top.
  • INP (Interaction to Next Paint): how quickly the page responds when the visitor taps a button, opens a menu or fills in a field.
  • CLS (Cumulative Layout Shift): how much elements move around on screen during loading, for example a button that jumps just as you are about to tap it.
Visual flow: Core Web Vitals, explained simply
Practical steps from the “Core Web Vitals, explained simply” section.

Why results differ between mobile, desktop and runs

The mobile test simulates a slower phone and a weaker connection than an ordinary computer has, which is why the mobile result is almost always worse. This is not an error: many visitors really do use modest phones and unstable mobile networks.

Lab results also vary from one run to the next, because of the server, the network or external scripts that respond differently. That is why a fair before-and-after comparison is made over several runs, on the same pages and under the same conditions, not on a single test.

Test pages other than the home page. In an online shop, category pages, product pages and the cart often have different problems from the home page.

Common causes of a slow site

Most speed problems recur from one site to another and have fairly predictable causes:

  • Images that are too large, uploaded at the original size from a camera or a stock image library, without resizing or compression.
  • Multiple fonts, or fonts loaded from external sources, which delay the display of text.
  • Themes or visual page builders that load a lot of code on every page, including pages where it is not used.
  • Extensions or plugins added over time, some unused, others that load scripts everywhere.
  • External scripts: chat, analytics, advertising pixels, maps or embedded videos.
  • No caching, or incorrectly configured caching.
  • A server or hosting plan that responds slowly, especially at busy times.
  • Elements without reserved dimensions, such as banners, images or ads, which push content around during loading.
Visual flow: Common causes of a slow website
Practical steps from the “Common causes of a slow website” section.

What is worth fixing first

The report does not order its recommendations by how much effort they require in your case. A long list does not mean that all items carry the same weight. A practical way to set the order is to put the estimated impact next to the effort required.

The steps usually look like this:

  1. Identify the important pages: those that receive the most traffic or bring in enquiries and orders.
  2. Run the tests on mobile and desktop and note the starting results.
  3. Fix first what is visible at the top of the page: the main image, the fonts, the elements that jump.
  4. Reduce what is loaded unnecessarily: images, unused extensions, scripts that do not need to run on every page.
  5. Check the cache and the server response.
  6. After each change, check that the site still works: forms, cart, payment, menus.
  7. Run the same tests again and compare the results over several runs.

When speed conflicts with other needs

Some elements that slow the page down are there on purpose: customer chat, the advertising pixel, an introductory video or a slider with offers. Removing them may improve the test result, but it may affect sales or campaign measurement.

These are business decisions, not just technical ones. Middle-ground options are often better: a script loaded only on the pages where it is needed, a video that starts only on request, a static image instead of a slider. Whoever does the optimisation should explain the trade-off and leave the decision to you.

What to ask for before and after an optimisation

Whether you work with the team that built the site or with someone else, a few clear questions prevent misunderstandings:

  • Which pages are tested and under what conditions, on mobile and on desktop?
  • What can actually be changed: the theme, the extensions, the server? What remains outside access or belongs to another provider?
  • Is there a backup before the changes, and a way to roll back?
  • How are the important functions checked after each change?
  • Will you receive the before and after results, on the same pages and over several runs?
  • Which recommendations remain unresolved, and why?
  • Who monitors speed over time, after updates or after new extensions are added?

Speed, Google ranking and realistic expectations

Google has stated publicly that page experience, including Core Web Vitals, is one of the signals used in ranking, alongside many others. Relevance and content quality remain essential, so a faster site does not automatically move up in the results.

The surest gain is for visitors: pages that open faster, buttons that respond immediately and content that does not jump around the screen. The evolution of real-world data can be tracked over time in the Core Web Vitals report in Google Search Console, which needs a collection period after changes.

Be wary of promises of a guaranteed maximum score. A very good lab score achieved by removing features the business needs is not a useful result.

Services related to this guide

Let's talk

Want to apply this in your business?

Tell us what you have now and what you want to solve; we will reply with concrete steps.

CallRequest a quote