The difference between Pagespeed Insights and Lighthouse

There is still some confusion around Pagespeed Insights and Google Lighthouse. Which one is best to audit with and which score is best?

What is Pagespeed Insights

When you perform an audit through Pagespeed Insights, you will see two different scores, one based on real user data and one based on lab data. The lab data score you see within Pagespeed Insights differs from a Lighthouse (lab data) audit, but I'll explain that later.

The real user (CrUX) data

When you enter a link and run a test, you see that the top part is based on your real users. This data is called CrUX data. This is free data that Google gives access to. The six metrics of the CrUX data consist of the three Core Web Vitals and three additional performance metrics.

When should I use Pagespeed Insights?

Pagespeed Insights is useful for getting an indication of how your users experience your website's speed. Its biggest drawback is that it doesn't show many nuances. You won't know where the numbers are coming from, think of circumstances under which your visitors enter the website (device and internet connection). This immediately makes it more difficult to know where you can implement possible optimizations.

Pagespeed Insights is not enough for a website that changes regularly. Imagine you make a change and it turns out to hurt your user experience. With Pagespeed Insights you'll only find out 28 days later, whereas a Real User Monitoring (RUM) tool can give you insights within an hour.

If you are not yet at this level in terms of pagespeed then there are certain situations where Pagespeed Insights can come in handy.

• You need to share a link to the audit report for your website.
• You want a quick indication of your site speed.
• You want a (basic) report of how your real user experiences your website or webshop.

What is Google Lighthouse?

Unlike Pagespeed Insights, Google Lighthouse does not show you data from your real users over the last 28 days. However, you do get to see other interesting things, such as scores on SEO, accessibility, and progressive web apps.

When should you use Google Lighthouse?

Lighthouse is often used by people who are just looking at website performance scores and do not yet know what the possibilities are. Unfortunately, we often see claims that the Lighthouse score would improve SEO, which is not true. The CrUX data plays a role in your rankings.

• Lighthouse is useful to get an indication of certain bottlenecks
• It sparks the conversation around performance
• When you want to take a look at scores other than performance

We have fully explained Lighthouse in another article, in case you are curious about the other scores Lighthouse shows.

The difference in lab data scores at Pagespeed Insights and Lighthouse

When you run a Lighthouse test through your browser, it may give a different result than the score you see within Pagespeed Insights. If I run a Lighthouse test (in my browser) on a new and old laptop, the results will be different. Because of the fact my new laptop is faster.

Even Chrome extensions can affect the score Lighthouse gives. For that reason, we recommend that you always run your audits in an incognito tab, because Chrome disables extensions here by default.

Pagespeed Insights vs. Lighthouse

The main difference between Pagespeed Insights and Lighthouse is that one also displays data from your real visitors. Lighthouse again displays data from other important aspects, which are not related to pagespeed.

The CrUX data that Pagespeed Insights provides you with is more important in performance terms than Lighthouse's lab data. This is because ultimately it's about your real visitors.

Besides the fact that Pagespeed Insights is data from real users, this also comes with some nuances:

• No page-level specific data, there is no way to tell if it's just a specific template that is causing issues or a webshop-wide bottleneck
• Data is more than a month behind, while most merchants are deploying weekly if not daily (server, marketing, third parties). Most merchants want to know about regression or improvements within an hour (or day, when not having enough traffic).
• There is no way to tell what causes certain issues. Knowing the conditions of your users that cause issues, developers will be able to fix issues way faster.

Conclusion

Are you already looking at the performance scores within Google Lighthouse? Good job, but we recommend you look beyond this. For example, start looking at the CrUX data within Pagespeed Insights.

Are you working on your website's performance and using Pagespeed Insights as your only tool to make improvements? Then we recommend you to look for a real user monitoring tool. This will give you the ability to deep-dive into performance data and be able to optimize more specifically for your real users. In addition, with a real user monitoring tool, you don't have to wait 28 days to see the effects of your improvements either.

Read our RUMvision vs. Lightspeed or our RUMvision vs. Pagespeed Insights article.

Jordy Scholing

Customer success & Google Developer Expert

• {label:mailto}
• linkedin-square