This is a reasonable cause of concern for SEOs and site owners, and the topic is discussed at length during the Google Search Central live stream with John Mueller recorded on December 4.
A question is asked related to Google using aggregate data to measure the Core Web Vitals. Google takes a group pages, analyzes them, and uses the data to calculate Core Web Vitals representative of the entire site.
That’s how Google arrives at the data found in Search Console, and the same evaluation method will be used when the Web Vitals ranking factor rolls out next year.
The question directed at Mueller asks how Google chooses the groups of pages used to calculate Core Web Vitals:
“Core Web Vitals will count for noindexed pages and things blocked in robots.txt. It’s quite interesting because obviously Search Console is aggregated, you get a group of pages… How do you understand that this is a group if these are noindexed and you’ve not got the context? Or is it just based on URL path
The idea Google could be grabbing any URLs to evaluate a site’s Web Vitals is worrying for sites with slow pages that are deliberately noindexed. Even though they’re excluded from search results, the slow pages may still bring the site down in Google’s evaluation of Core Web Vitals.
Mueller on Why Noindexed Pages Count Toward Core Web Vitals
Mueller offers the rationale that Core Web Vitals are designed to measure page experience, and noindexed pages are still accessible to users on the site, therefore they should be included in the overall evaluation.
However, he’s not sure how Search Console chooses which pages to evaluate.
“My general feeling is [noindexed pages are] something that’s also part of your website. So if you use some extra functionality, and that’s noindexed, people still see it as part of your website and say this website is slow, or this website is fast kind of thing. I don’t know how Search Console reports on that in particular. I know within the Chrome User Experience report data in the Chrome developer site they have some information of how that grouping is made, but I don’t know how detailed that is there.
One of the tricky parts is also it’s very hard for us to understand when a page is something that is not meant to be indexed because all of the canonical decisions and all of that. Just looking at an individual page on its own, it’s sometimes not absolutely clear is this something that can be accessed directly, or does that cookie that was set in the beginning need to be set to access the page… like what all is involved there. So I imagine that’s always kind of tricky to balance out.”
Mueller is then asked a follow-up question from one of the live attendees regarding whether there’s a way to exclude pages from Google’s Core Web Vitals evaluation.
Can Pages Be Excluded From Core Web Vitals Evaluation?
A live attendee quickly jumps in with an appropriate follow-up question:
“Is there any way to ensure that particular pages are not being used to assess how well a site passes Core web Vitals?”
In response, Mueller plainly states: “I don’t have any great answers for you at the moment.”
Not quite satisfied with that response, the attendee wants to know whether there will be answers in the future.
Mueller says he’s not up-to-date with all the information out there about grouping, so he’s hesitant to offer any suggestions. Though, in general, Google tends to sort pages into groups based on URL and page content.
So it may be helpful to use folders as a way of signalling to Google which pages should be recognized as a group.
Here’s the full quote from Mueller:
“Maybe. I don’t know what the information out there is on grouping at the moment, so it’s really hard for me to say exactly what you’d need to watch out for or what you can ignore.
In general, when it comes to grouping across websites, when we try to do grouping we try to do that on the one hand by URL pattern, and on the other hand by the content of the site. And those are, I think, the way we do grouping in Search Console, for example. So if you have parts of your website that you want to be seen as belonging together then I would definitely make sure from a URL pattern point of view it’s clear these belong together.
So if you’re specifically worried about Search pages, for example, then putting those in a folder with /search/ makes it a little bit easier for us to understand all of these search pages belong together, all of these product pages belong together, and all of these blog posts belong together. We might be able to treat them individually when it comes to Core Web Vitals.”
Google Uses Limited Data to Assess Core Web Vitals
The last thing Mueller adds on the subject is Google doesn’t have a lot of aggregated data for each website. For some sites Google may just have a few data points to go off off.
In cases where Google used a limited amount of site data, grouping pages by URL structure may have no effect on which ones are chosen as a group.
“The other thing to keep in mind is we don’t have a ton of aggregated data for every website. It might be that for some sites we just have few data points. So we just have for the website overall (information on the Core Web Vitals), and in that case even if you use URL structures to split things up cleanly, we might not be able to use that grouping because we just don’t have enough data for those individual groups.”
Mueller says more information about page grouping may be available at a later time. To manage expectations, he says there will 100% never be a way to pick and choose which individual pages should be used to calculate Core Web Vitals.
“So that’s something I suspect over time we’ll probably be able to have a little bit more information, but it will still be something that is not 100% such that you can say ignore this page and do count that page kind of thing on a website.”
Hear the several questions and answers about page grouping in the video below: