Understanding URL Groups in GSC

[Matt See]

Hi all,

Love this forum, and it's been great to read through. I have a question about how GSC uses the CRUX data.

In GSC, we have a URL group of 24 URLs that are scoring "Poor" in CLS. However, one of the pages (our homepage) gets 90% of the traffic, and through any CRUX report, we see that the p75 its 0.00 or 0.01. This hasn't fluctuated in months, and we can't find anything wrong with it.

In the the GSC report, there are only 3 URLs with enough data: the homepage and 2 others. They all have 0.00 CLS next to them. The report is showing "Poor," with a 0.36 for the URL group. 

A basic interpretation is that one or more of the other 21 URLs in the group is scoring a p75 of 0.36, and we should fix that. 

This would make some (or all) of this true:

• GSC groups similar URLs, but these URLs may have drastically different CWVs. This similar URLs are based on layout, features, deliverability, and other shared technical attributes
  
• The metrics of a group in GSC are only as strong as the weakest page within the group

I know this might be the wrong side to ask, but this will really help us solve the issue. Thanks in advance

[Barry Pollard]

I afraid that GSC is managed by the search team, and they classify, and group the URLs based on their own criteria and not CrUX's so we don't have the expertise to explain this.

I would check out: https://support.google.com/webmasters/answer/9205520

In my experience when there are only a small number of URLs that make up a site (as appears to be the case here), they are commonly grouped together regardless of how similar or different they are.

Also, CrUX reports numbers in it's origin group by page views, and it appears from the above link that GSC does the same. This means if your home page really does get 90% of your traffic, and it is that group of 24, then it's surprising the group does not effectively report the home page's Core Web Vitals.

[Matt See]

Really appreciate you getting back. It's clearly a GSC issue, so thanks for even responding. I should have shared this screenshot previously.

We have some 9K other URLs that make up 80%+ of the total domain traffic. 

Homepage is simply 90%+ of traffic within this group of 23.

While it's not be in your domain, would you presume that the [0.37 poor] rating is coming from a single page in the group that may not have enough data, or would you guess that the 0.00 for homepage is not telling the full story? It's been extremely puzzling for us. We know there are some CLS issues due to fallback font, but we'd like to know what to contain and where to look. 

Thanks again, Barry.