John Mueller answered a question during the SEO Office-hours session about why core Web Vitals scores keep changing even though the web pages themselves haven’t changed.
There are two basic types of Web Vitals scores:
- Field data
These are the top Web Vitals scores recorded by real visitors visiting a website in the real world.
The purpose of terrain data is to show the actual user experience of actual site visitors on a given site.
- Laboratory data
These are basic Web Vitals scores generated by a simulated visit. A Google Lighthouse bot that uses a user agent identified as Chrome/94.0.4590.2 will visit the page under test and an algorithm will simulate that the visit took place on a mobile phone and over a limited internet connection. Page load information is collected and an artificial Web Vitals base score is generated. The purpose of laboratory data is diagnosis. Lab Data helps a site editor or SEO troubleshoot the performance of a web page. Lab Data provides information that will help a publisher or SEO understand what is slowing down the webpage and provide tips for troubleshooting.
Screenshot of John Mueller discussing Core Web Vitals scores
Lab data is generated using Google’s Lighthouse tool.
In fact, the bot that visits a site when generating the base Web Vitals scores for lab data uses the word Lighthouse in the user agent.
According to the official Lighthouse webpage:
“Lighthouse is an open source, automated tool for improving the quality of web pages.
…It has audits for performance, accessibility, Progressive Web Apps, SEO and more.
You can run Lighthouse in Chrome DevTools, from the command line, or as a Node. You give Lighthouse a URL to audit, it runs a series of audits on the page, and then it generates a page performance report.
From there, use the failing audits as indicators on how to improve the page. Each audit has a reference document explaining why the audit is important, as well as how to fix it.
Core Web Vitals scores keep changing
The question John Mueller answered did not specify whether it was field data or lab data.
The person asking the question was simply asking when the main web vitals scores stop changing, without specifying field data or lab data.
Mueller answered the question from the perspective of field data, which is the main Web Vitals scores based on real-world visitors.
This is the question that was asked:
“When does vital web data provide stable and correct information?
It keeps changing without changing any data on the website.
Core Web Vitals scores are influenced by random factors
The answer focused on the randomness of the data used to generate the basic Web Vitals field data.
Jean Mueller replied:
“So I think it’s probably a side effect of how the main lifeblood of the web and updating the page experience is handled.
And that’s something where I would try to dig into those details to understand a bit more how the field data, … the data that users actually see, kind of plays a role in that.
And that’s something where if users from a wide variety of backgrounds, different locations, and different device types are accessing your pages, you’re likely to see fluctuations over time as well.
Internet speeds vary
A fast server and a fast website are only part of achieving a high vital web score and it is the only part that is under the control of the SEO or site editor.
As John Mueller said, there are a wide variety of other factors that influence base Web Vitals scores and many of these factors cannot always be controlled.
Even if a website is hosted on a fast dedicated server, network congestion, an outdated mobile phone used by a site visitor, and a poor mobile data connection can all significantly impact a website’s essential web performance. otherwise fast website, contributing to a less than good Web Vitals base score on field data.
And some site visitors may be using a newer model cell phone with a powerful processor running on a newer 5G network. This site visitor will generate high scores on the basic data of the Web Vitals field.
Things like network latency (how slow the entire internet is at any given time) can also affect both actual field data and simulated lab data.
So even if the lab data is not a real visitor, a real Chrome browser-based bot travels across the internet to visit the web page under test. And it will also affect the scores of the simulated lab data.
Google’s Web.dev website has published an explainer explaining why field data is constantly changing and not always the same from month to month.
Web.dev explain it like this:
“The most important thing to understand about field data is that it’s not just a number, it’s a distribution of numbers. In other words, for some people who visit your site, it may load very quickly, while for others, it may load very slowly.
the field data for your site is the complete set of all performance data collected from your users.
As an example, CrUX reports show a distribution of performance metrics from real Chrome users over a 28-day period. If you look at almost any CrUX report, you can see that some users who visit a site may have a very good experience while others may have a very bad experience.
Whether you’re testing a site for lab data or looking at real-world field data, core Web Vitals scores will tend to fluctuate and likely won’t provide a consistent score month after month.
When do Core Web Vitals scores stabilize?
Watch John Mueller answer the question at 19:05 minutes: