In preparation for the next Core Web Vitals update, I am working on decreasing CLS across my companies site. When I run the Lighthouse audit in Chrome, the page I've updated gets a 0.05, under the max 0.1 Google recommends.
However, when my companies SEO specialist runs an audit, she gets scores > 0.3 on the same page. I'd thought maybe it could be a speed issue, but throttling my internet speed didn't change my results at all (at least where CLS is concerned). For that matter, the audit on her end is calling out high level parent div elements as causing high CLS, rather than specific elements within the parent div.
Is there any reason why her scores would be so much worse than mine on the same page? Are there any known issues with CLS testing in Lighthouse I should be aware of?
If two different people are getting two different results there are numerous things to check.
The first one is to check that you are both testing at the same display size (for desktop tests - all mobile tests are same size) as otherwise one of you may get CLS at one breakpoint that does not occur at another breakpoint (or items may be on the screen on one device and not another that contribute to CLS).
The next thing to check is that the computers have a similar CPU power. Lighthouse uses a 4x slowdown when running the mobile test, if you have an i3 vs an i9 this can make quite a difference!
There is a useful document that explains where variability can come from when testing with Lighthouse.