John Mueller from Google demonstrates how the Search Console Core Web Vitals report and PageSpeed Insights ratings are different.
Google’s John Mueller explains in an SEO office hour video why the PageSpeed Insights score and the Search Console Core Web Vitals score differ regularly. He shows why it makes logical that two measures of page speed never agree and why it never happens.
PageSpeed Insights vs Search Console CWV Report
It makes sense to inquire as to why two measurements that claim to capture the same phenomenon yield disparate outcomes.
“When I check my PageSpeed Insight score on my website, I see a basic number,” the user asks. Why doesn’t this match the Core Web Vitals data in Search Console?
Which of these figures is accurate?”
Different Page Speed Scores
The scores offered by PageSpeed Insights and Search Console are two separate sets of numbers, as John Mueller first points out.
He stated:
“I believe that the first step should be to state the obvious: there is no precise speed measurement that can be used to gauge how well a website is serving users.
PageSpeed Insights displays a single number I believe it’s a score from 0 to 100 by default. This score is based on several assumptions, including the notion that different factors affect users’ response times differently.
And we create a score based on that. The Core Web Vitals data in Search Console is based on three metrics for speed, responsiveness, and interactivity. Of course, these figures differ significantly because there are three of them rather than just one.”
Following that, Mueller explains how and why the findings of the two measurements PageSpeed Insights and Search Console CWV report are different.
But there is also a significant discrepancy in how these data are calculated, said Mueller.
There is a distinction between the so-called field data and the lab data, specifically.
Field data represents what visitors to your website have viewed. And Search Console makes use of this.
We also employ that for search. While lab data is more of a theoretical perspective of your website, based on certain assumptions made by our systems that the typical visitor is likely to look like this, use this kind of device, and have this kind of connection,
We will then make an educated guess as to what those figures would be for the typical user based on these presumptions. And none of those estimates will ever be entirely accurate.
Mueller is so asserting that Search Console scores accurately reflect the experiences of actual site visitors. Visitors who have chosen to allow Chrome to send anonymized core web vitals data are how Google determines those real numbers. Only those users who have chosen to provide their data to Google are counted when measuring website traffic. PageSpeed Insights is modeling what a user could encounter, as opposed to Search Console’s field data.
Data from Search Console is used to display what actual site visitors are doing. The goal of the PageSpeed Insights data is to give an approximation of what’s happening to give diagnostic comments on what might be causing a slow webpage performance. Mueller continues by pointing out that even though Search Console is based on actual site visitors, the statistics will fluctuate and be inconsistent.
“And similarly, the data that consumers have seen, that will change over time as well, where some users might have a really fast connection or a fast device and everything runs extremely quickly on their website, or when they visit your website,” he said. “And some people might not. And as a result, this variance can always produce various numbers. To understand the general state of a website, we normally advise using the field data, or the data you would see in Search Console. and then optimize your website and make an effort to enhance things using the lab data, namely the particular tests that you may run on your own directly. And after you’re satisfied with the lab data you’re receiving from your new website version, you can gradually collect the field data (which happens automatically) and confirm that users genuinely perceive the website as being quicker or more responsive. In conclusion, none of these measurements have a single, unquestionably accurate value.
There isn’t a single, unquestionably accurate response that you could point to and say, ‘This is how it ought to be.’ Instead, there are several presumptions and data collection methods, each of which is slightly different.”
Two Metrics for Two Purposes
The two measures are valuable for two separate goals, which may be a useful way to think about them. Data from Search Console Core Web Vitals, which is derived from actual users, gives an example of how real users interact with websites (field data0. Actual website visitor statistics might draw attention to problems that a simulated sampling might miss. The goal of PageSpeed Insights scores, which are simulated scores, is to analyze and troubleshoot page speed problems. This program finds issues that slow down page downloads and makes recommendations about how to repair them.
Since the data sources differ, a precise match is impossible. But it doesn’t matter because each instrument has a different purpose.
