Open in new background tab effects on FCP, LCP

[I Biron Ren]

I have the Web Vitals chrome extension installed on my browser and I've noticed LCP increases on pages I open in the background. 

If I open a page in a new tab in the background, but I don't navigate to the page right away, the time elapsed between clicking on the page URL and actually visiting the tab is added to the LCP. 

E.g., I'm viewing /page-a/ and open /page-b/ in the background in a new tab. I wait 90 seconds before navigating to the tab, and the extension reports my local LCP time as 92 seconds. If I open the page in the same tab it reports 2 seconds to LCP. 

The extension does not report on FCP but I would guess it's impacted in a similar way. 

The description in the Chrome Web Store says, "This extension measures the Core Web Vitals ... [and] is consistent with how these metrics are measured by Chrome and reported to other Google tools (e.g. Chrome User Experience Report, Page Speed Insights, Search Console)." 

This would lead me to believe that the inflated LCP is in fact recorded and used by Google as part of the PX ranking factor. 

If that assumption is correct, then isn't this a gross misrepresentation of CWV in those situations, and should be fixed to only start measuring FCP and LCP when the page is actually in view of the browser? 

[Rick Viscomi]

That's a good question. I think the extension might be using "consistent" a little too loosely in that description. I documented this exact case of inflated background LCP values in the v1 announcement post and noted that "these kinds of irrelevant LCP values are ignored in the CrUX dataset."

So it's not the case that inflated LCP values affect a site's CWV as reported by CrUX.

Rick

[Philip Walton]

You can see the logic to handle this in the web-vitals JS library here:

https://github.com/GoogleChrome/web-vitals/blob/6a51624ad49480fff99c2d4919fb0383f827882b/src/getLCP.ts#L38-L40

My understanding is there are some issues specific to Chrome extensions where page visibility state isn't always known, so there are cases where the extension may be inaccurate, but be assured the underlying data reporting is accurate.

--
You received this message because you are subscribed to the Google Groups "Chrome UX Report (Discussions)" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chrome-ux-repo...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chrome-ux-report/a6ef725d-d137-4626-916a-998d7d3303e3n%40chromium.org.