Do you want to come to Andalu-SEO?
Andalu-SEO

Measuring a website’s speed (WPO)

Este artículo desmonta mitos sobre la velocidad web en SEO Técnico, explicando cómo Google realmente mide el rendimiento a través de las Core Web Vitals

Measuring a website’s speed (WPO)
Author:
Carlos en Dean Romero
Topics:
WPO
Publication Date:
2025-12-04

Last Review:
2025-12-05

When talking about Technical SEO, one of the aspects that stands out the most, is more requested or draws the most attention is WPO (Web Performance Optimization), that is, improving a website’s speed.

The reality is that although it is a factor that affects search rankings both directly (measured by Core Web Vitals) and indirectly (user behavior), it is not the most important part of Technical SEO nor the one with the greatest impact. However, it is one of the most visible and noticeable aspects, because anyone, whether they have knowledge about websites or not, can tell if a page is fast or slow.

The aim of this post is to learn how to really measure and analyze your website’s speed and interpret the correct information to determine whether actions should be taken or not.

Difference between Google and the user

As I mentioned, your website’s speed affects rankings directly and indirectly. For users, it never hurts to squeeze out a bit more speed, but for Google, beyond a certain point it becomes irrelevant. But to know that, we have to understand how Google measures our website’s speed.

Core Web Vitals and how Google measures a website’s speed

There is a lot of misinformation about knowing whether a website is fast or not for Google, and this is due to the PageSpeed Insights tool which, although very complete, people don’t know how to interpret correctly.

The circle with the number and the traffic light in PageSpeed Insights is not used to measure a website’s speed, but to get an estimate of possible improvements based on a generic checklist in Google’s tool that may or may not have an impact on improving speed. Having a 0 or a 100 has no impact at all on SEO or your website’s speed. — Carlos Sánchez

WPO pageSpeed Insights

Here on this example page we can see that there is a performance score of 82, which does not necessarily have any impact; it only shows us possible points for improvement and the specific analysis of our site. But Google does not select a particular load to measure our website’s speed; instead, it uses a sufficiently representative sample of users over the last 28 days in any Chrome browser.

It is true, however, that we don’t know how long this will remain the case, since Google is facing an antitrust trial over Chrome which, depending on the ruling, may mean that in the future it will no longer be able to obtain that information from users. (07/07/2025)

How Google actually analyzes speed

As I mentioned, Google measures speed based on a large sample over the last 28 days of Chrome users who have visited your page. Depending on their user experience and results, Google will understand that your website is fast or slow mainly according to these 3 criteria:

So yes, Google’s perception of our website’s speed depends on the user. It’s not the same to have a site with a rural audience, or in Angola with worse internet, as the same site with the same technology, coding and everything, whose target audience usually browses from a desktop with fiber. The country, device and connection of our users matter, and there is nothing we can do about that.

That said, we can find out these metrics in several ways:

PageSpeed Insights

There is a section in PageSpeed Insights that is the most relevant one and that people overlook, and it’s the first section that says “Discover what real users experience”, only available in the web version (not the one in Devtools). You should analyze both mobile and desktop, and several things can happen.

Test not passed

In this case, Google understands that, based on the experience of its users over the last 28 days, they have had a negative experience. In this case, we should make an effort to improve our website’s speed in order to achieve better results in Google.

Test passed

prueba superada core web vitals

If our main metrics are Passed, any effort we make to further improve speed will be for the user experience, since at the level of Google there will be no further improvement from optimizing it more.

No data

PageSpeed without data

In this case, you should focus on other aspects before checking speed, because although it’s possible that your site is slow, this means that not enough users visit your website per month to have the necessary information.

For example, even if we have a “speed” score of 98, Google is not even taking it into account on this site because not enough users visited it in the last month (something normal on a seasonal website).

good performance score but not in core web vitals

Google Search Console

This should be the best way to measure a website’s speed, since it doesn’t make sense to go page by page or only evaluate the homepage to detect whether our site has speed issues according to Google.

For that we have a free tool that is our ally, Google Search Console.

By accessing the panel where it says “Core Web Vitals”

Core Web Vitals Search Console

Whether we select poor URLs, those that need improvement, or fast ones, we can quickly notice something.

And that is that, in addition to knowing that Google analyzes the site’s speed based on Chrome users, Google does not score pages one by one, but groups the site and subdirectories play a crucial role here.

Core Web Vitals grouping URLs

Thanks to this, Google can interpret that these are different URLs. In this case, on my site Google has interpreted that posts, posts in English and core pages are different types of content. That’s why it calculates an average speed.

However, if everything appears as correct, from Google’s point of view there is no need to improve speed. This is the quickest way to check it (it’s important to check both mobile and desktop).

Screaming Frog + PageSpeed Insights API

We have the option of running the PageSpeed Insights analysis across our entire site without going URL by URL, by connecting it via API to Search Console. Screaming Frog and PageSpeed Insights API

If you don’t know how to get the PageSpeed Insights API key, you can get the key here easily.

You connect it in: crawl configuration > connect APIs > PageSpeed Insights

PageSpeed API

This way we can check all the metrics provided by PageSpeed Insights, including Core Web Vitals.

Tools to measure a website’s speed

However, it’s not all about Google, and it may also be that, regardless of how Google perceives our site, our users perceive it as fast or slow.

To analyze a website’s speed, we need to take into account:

Here are different ways to check it independently of Google:

Pingdom

Pingdom Website Speed is a tool that lets you analyze your site from different geographic locations and see information about your page’s performance from those points.

Devtools

Chrome console performance
The Chrome Console is an essential tool for Technical SEO and for analyzing in detail the effects of elements on a site and understanding how to improve its loading.

GTMetrix

I could give more information, but there is an extensive and complete tutorial on GTmetrix by DinoRank. It works like Pingdom but is even more complete. It’s important to have alternatives.

Webpagetest

webpagetest

If you want customization with recurrence and the option of a free account with periodic reports, webpagetest is a very good option.

Quick tricks to improve WPO

HTTP version

The HTTP version depends on your server. You can ask your hosting provider because you should be using at least the http/2 protocol.

Compression

How to check your site’s compression type:

If you don’t have any kind of compression, you can always ask me on LinkedIn.

Minification

There are multiple plugins and tools that allow you to minify your code. Minifying code means removing spaces, indentations and unnecessary characters so that machines can still read it the same way even though humans find it harder to understand. The goal is to reduce the file size of code files (HTML, CSS, JS) while keeping the same result.

File size

Images and videos should have an optimal quality/size ratio. For images I’ll leave you an article by Dean Romero on how to compress images for a blog. I also recommend compressor.io or automatic plugins like webp express.
As for videos, nothing beats the free software Handbrake, it works wonders.

More advanced tricks

And here are some other quick tricks that require a bit more knowledge to implement.

CDN

A CDN is not always an effective measure; sometimes it can even be counterproductive. It depends on the load your usual server can handle, your target audience, the CDN and the specific plan you choose within the CDN, and even on how it’s implemented.

I usually recommend Cloudflare, but in Spain it is having many issues due to the controversy with the football league. So while that is being resolved, I can recommend Fastly, and if you are less advanced, BunnyCDN.

To avoid leaving any affiliate links here, what I’ll leave is a discount coupon that only benefits you:

#SAASMASTER

It’s important to check the POPs that the CDN has to see whether it has geographic locations close to your region. You can check this on CDNPlanet.

Preloads and deferred loading

The order of the code matters as much as the code itself. Although you have to be careful when changing the order of JavaScript (EXTREMELY IMPORTANT if you use plugins without deep knowledge of the topic) because you can break functionalities.

We can use attributes such as FetchPriority, which WordPress includes by default but which can always be further optimized.

We can make use of different preloads in our favor, such as preload, preconnect or prefetch.

We can use deferred loading such as lazy loading, load on click, async or defer.

By keeping control over the order in which our resources load, we can improve the perceived speed of our page, which ultimately means improving the page’s speed.

Cache

Cache is a form of temporary storage aimed at improving a system’s speed. In fact, in the web world itself there are many types of cache that affect SEO.

There are mainly two types that are relevant for SEO:

  1. Browser cache, which prevents repeated resources (such as CSS or JS) from being downloaded from the server again,
  2. Server cache, which stores already generated versions of pages so they don’t have to be processed every time.

Both are configured from the server and require a strategy to avoid issues when making changes to the site. In addition, there are other types such as DNS, CDN or application cache, which also influence load speed and site behavior.

Key points

You don’t need to become an expert overnight, so let’s start with the key points that should stick in your memory after reading this article:


Author’s notes

I must thank Dean Romero for his initiative for the industry in bringing together different professionals and for having included me in this project.

I appreciate the space and I have written this post with great care and affection for anyone who has wanted to read it.

Also, with his permission, I would like to mention that anyone who wants to learn all these technical aspects in much greater depth, both about WPO and other aspects of Technical SEO, will always be welcome in the Technical SEO master’s program by Asdrubal.

This post has been written for its indexation on Dean Romero’s blog.

If you like this article, you would help me a lot by sharing my content:
Interested in Advanced SEO Training?

I currently offer advanced SEO training in Spanish. Would you like me to create an English version? Let me know!

Tell me you're interested
You might be interested in other articles:
SEO Articles
Usamos cookies para asegurar que te damos la mejor experiencia en nuestra web. Aquí tienes nuestra política de Cookies.