6

Why does the Cumulative Layout Shift metric reported in Google's speed measurement tools, like Lighthouse/PageSpeed Insights differ from what is reported in Search Console?

Rick Viscomi
  • 8,180
  • 4
  • 35
  • 50
addyo
  • 2,798
  • 1
  • 14
  • 21

1 Answers1

3

There is some nuance to how PageSpeed Insights (PSI) and Search Console report on Cumulative Layout Shift. The lab portion (Lighthouse) of PSI measures CLS until Lighthouse considers a page fully loaded. It does not measure beyond page-load.

enter image description here

Search Console and the field portion of the PSI uses Chrome User Experience Report data and measures CLS until unload, stopping reporting after pagehidden (i.e this is inclusive of CLS beyond page-load). What this means is that the reporting you see in different tools can vary based on the window of time we are able to look at.

enter image description here

Lab tools like Lighthouse have to define a shorter window because they are focused on the experience during page load and optimize for delivering information about the experience quickly. Field data is able to take a more holistic view of the user-experience, which could include shifts caused after a page has loaded and the user has scrolled down to other interactions contributing to CLS.

addyo
  • 2,798
  • 1
  • 14
  • 21
  • This is great to know, thanks for taking the time to add this in. The biggest problem is then working out what is causing a large CLS in the field if PSI shows 0. What method would you suggest? I tried with paint flashing switched on and with layout shift regions but still get 0 shift, whereas field data says 0.99! Just wondered what a true expert would suggest :-) +1 to question and answer as people need to understand PSI better. – GrahamTheDev Jul 02 '20 at 19:12
  • @GrahamRitchie That's an excellent question! Would you like to ask that as a new top-level question and I'll answer? There's two parts of an answer I can cover there. – addyo Jul 05 '20 at 23:50
  • thanks, I have added it as a question, hopefully doesn't get closed! https://stackoverflow.com/questions/62751471/how-to-find-cumulative-layout-shift-problems-if-page-speed-insights-says-one-thi – GrahamTheDev Jul 06 '20 at 07:47
  • Has this changed, or does GSC not use the same time frame that PSI's field data uses? I'm seeing perfect field data (100% CLS = 0) on some domains while they're still getting "CLS > 0.1" in GSC. – janh Jul 29 '21 at 11:28
  • For anyone wondering: it seems to just be severely delayed. Pages that had been perfect in field data for a month or so have now finally turned "green" in GSC as well. – janh Aug 09 '21 at 16:05