There are a lot of tools out there that measure your page speed, numerous tools from Google and many tools from third party developers. But not all of them measure the same thing. Some measure when the page starts to load, some measure when first party stuff loads on the page and some measure other aspects of how fast the page is.
Ultimately, what is important to the user is when they can start using the page they are looking at. If there is a tool on the page or links and those featured do not work until the page is fully loaded, then that is what your measurement and benchmark should be for that page.
John Mueller from Google said he'd be more interested in looking "at the page as a whole" at the time the page is "ready to interact" with. Here is his tweet:
I'd look at the page as a whole (until ready to interact), and look for low-hanging fruit there. It might be the initial HTML response (usually where app-level speed plays in), it might be somewhere else.
— John ☆.o(≧▽≦)o.☆ (@JohnMu) November 26, 2017
Now, is Google measuring this and does this impact your Google rankings? Probably not fully. Google has said you need a really slow page for it to impact your Google rankings.
Of course, we all know how slow pages can devastate conversion rates - so stop just thinking about Google and think about the users you have already.
Forum discussion at Twitter.
