INP appears to have begun failing across 98% of my website

[Simon Lovejoy]

Hello everyone,

I was hoping that this group may be a better place to be ask questions regarding issues around the Core Web Vitals.

Since July 20, 98% of my site's pages at https://www.wanstor.com, with the exception of the homepage itself, have begun to fail on the Mobile INP Core Web Vital - slipping from 189 ms to 239 ms.

An example of this would be as follows:

Homepage (passing):

https://pagespeed.web.dev/analysis/https-www-wanstor-com/vt8l0rw9tf?form_factor=mobile

Credits page (failing):

https://pagespeed.web.dev/analysis/https-www-wanstor-com-why-wanstor-awards-accreditations-html/hcc3gvihnz?form_factor=mobile

I understand that the cause of this could be attributed to more than just interaction delay and may also occur as a result of elements within the body as well as, unless I'm wrong, a DOM that is too large amongst other possibilities. 

I've attempted to audit this using PSI but as i have found in the past, the recommendations appear to line up closely between a pass (the homepage) and a fail (almost all other pages), with the DOM size reported in many of the fail cases to be smaller or less than the homepage.

i would greatly appreciate any input from the community around where i could be looking next or whether there is anything that this behaviour may align with.

[Barry Pollard]

Credits page (failing):
https://pagespeed.web.dev/analysis/https-www-wanstor-com-why-wanstor-awards-accreditations-html/hcc3gvihnz?form_factor=mobile

It's important to note that PSI is NOT saying the Credits Page itself is failing and is instead saying there is insufficient URL-level data and so it's dropping back to overall origin-level data:

Therefore, this page may be fine, and it may be other, heavier pages pulling down the aggregate origin-level score.

CrUX data (as shown in PSI) is a high-level summary and you should combine this with other analytics to see what pages users are visiting on our site most as those are the pages to concentrate on. The best way is to collect the Core Web Vitals data through a RUM solution so you can see exactly what pages are contributing what INP values to the overall score. It looks to me like you have a debugbear script on there so maybe you already have this data?

Alternatively Google Search Console can give a ranked listing of the pages impacting a Core Web Vitals group.

Large, complex pages with large DOMs or lots of complex CSS can indeed mean more work for the browser, but it's far from the only reason. JavaScript is often a cause too and there's lots on your site that requires a fair bit of processing (Consent Providers, Tag Managers, Ads). These are part and parcel of running a website for many, but that doesn't negate the fact they slow your website down.

Looking at the history of your site's INP scores on mobile, it does look like something changed this month:

There was a similar change in March, so does that give clues?

If your site has not changed, then it could be a result of something else changing (e.g. Heavier ads being displayed on the site?) Or the user population (e.g. if you have an influx of users on slower devices, perhaps due to a marketing campaign?). The latter should be easier to confirm with Google Analytics.

Hopefully that gives you some ideas about where to look next to explain this. I would also say that 239 ms, is just outside the 200 ms recommendation so is not a terrible score. Still it would be great to get under that threshold and understand why it fluctuates some months to help you do that!

--
You received this message because you are subscribed to the Google Groups "web-vitals-feedback" group.
To unsubscribe from this group and stop receiving emails from it, send an email to web-vitals-feed...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/e0b1113b-21fa-41a3-8c4c-2eeca1ad3cb8n%40googlegroups.com.

[Simon Lovejoy]

This is great - thanks Barry.

i have kept the debugbear script resident on that subdomain as i can't afford to fund it as an ongoing business expense, but may take another month subscription to see what data this brings back. Ideally this would be set and forget in order to collect long-term information, but the function to which i answer will sadly not make the business case for this cost.

 I can't see any reason as to these changes over time as the site had been passing since the introduction of INP as a metric - i have not made any major structural changes that i can recollect based on changelogs, so ads is one thing i am going to have to try and get more information out of the digital marketing manager here if i am able to.

I am also wondering if there have been any changes to onetrust's consent module deployment, as the company appears to be restructuring parts of its business and pricing models at this point.

Many thanks for your help with this!