Page speed Does not pass even after scoring 90 on Page Speed Insight-CodePudding

My webpage is scoring 90 on desktop version but yet it's test result on Field Data show "does not pass". While the same page on Mobile with 70 speed is marked as "Passed"

What's the criteria over here and what else is needed to pass test on desktop version. Here is the page on which I'm performing test: Blog Page

Note: This page speed is on 90 from about 2 months. Moreover if anyone can guide about improving page speed on Mobile in WordPress using DIVI builder, that would be helpful.

CodePudding user response：

Field Data pass or fails a website based on historical data.

Field Data Over the previous 28-day collection period, field data shows that this page does not pass the Core Web Vitals assessment.

So if you have made recent changes to your website to improve your site score you need to wait atleast a month so that Field Data shows result based on newer data. https://developers.google.com/speed/docs/insights/v5/about#distribution

CodePudding user response：

Although 6 items show in "Field Data" only three of them actually count towards your Core Web Vitals assessment.

First Input Delay (FID)
Largest Contentful Paint (LCP)
Cumulative Layout Shift (CLS)

You will notice that they are denoted with a blue marker.

On mobile all 3 of them pass, despite a lower overall performance score.

However on Desktop your LCP occurs at 3.6 seconds average, which is not a pass (it needs to be within 2.5 seconds).

That is why you do not pass on Desktop but do on mobile.

This appears to be something with your font at a glance (sorry not at PC to test properly), causing a late switch out. I could be wrong, as I said, I haven't had chance to test so you need to investigate using Dev Tools etc.

Bear in mind that the score you see (95 on Desktop, 75 on mobile) is part of a synthetic test performed each time you run Page Speed Insights and has no bearing on your Field Data or Origin Summary.

The data in the "Field Data" (and Origin Summary) is real world data, gathered from browsers, so they can be far apart if you have a problem at a particular screen size (for example) etc. that is not picked up in a synthetic test.