[Support request] Differences between rendering of Index/Archive and Page/Post?

[acela]

Hi GP Team,

I have a strange issue I hope you can help me figure out. I have a blog site and Google PageSpeed Insights is failing all of my index & archive pages on CLS (lab data = 0 but field data = 0.66, for example), on both mobile and desktop.

However, all of my single posts & pages are passing Core Web Vitals just fine – including CLS (Cumulative Layout Shift lab data = 0 and field data = 0).

As far as I’m aware, there is no difference in what my site is doing on either. So I ran this technique, slowing down my site loads and watching for Layout Shift Regions, to try to diagnose the problem.

And although all pageloads seemingly have the same surrounding content and site structure, this technique shows a difference (in Chrome’s eyes) when loading new Index/Archives vs. Page/Posts.

The former shows a blue box around the header area. The latter does not. This is really odd, because all header content is identical and there is no discernible shifting going on. But Chrome seems to think there is CLS happening (not sure why!), but only for Index/Archives.

My Home/Archive pages simply show 10 posts in blog style, each with the post title, featured image, excerpt, and Read More link.

So my question is, are there any differences in GeneratePress between the loading of the site components (in particular in the header region) for different page types?

I can’t imagine why there would be, but I’m not sure what other avenues to pursue because everything else checks out.

(My site is using a Flexbox structure with Separate Containers and Primary/Sticky Navigation plus Mobile Header, but Top Bar and Secondary Navigation disabled.)

Thanks very much!

[Tom]

Hi there,

There shouldn’t be any difference – it’s the same HTML/CSS regardless of pages. Do you have a caching plugin that might be working on archives but not single posts?

Your lab data is perfect, but the field data isn’t. It’s possible that at some point in the last 28 days there was CLS, but it’s been resolved now (according to the lab data).

I also throttled Chrome to “Slow 3G” and did some hard refreshes on mobile – didn’t see any CLS issues at all.

I would keep an eye on it and hope that the field data catches up.

[acela]

Hi Tom, thanks so much for your fast response! That’s what I thought, but I really appreciate the confirmation. It’s certainly a bit of a mystery as to the origins of the CLS, because everything to the naked eye seems perfect.

The origin hosting provider is Kinsta, with Cloudflare APO at the edge – so everything should be cached, in practice.

I’ll give the field data some time to see what happens, and hopefully it will shake out. I’ll also start a new validation in the Google Search Console.

Thank you as always for your amazing support! So glad I found GeneratePress.

[Tom]

No problem at all! Thanks! 🙂

[Author]

Posts