Notes from the Martin Splitt’s JavaScript SEO Office Hours on May 6th 2020
LCP, CLS & FID are the Best Metrics to Measure User Experience Page Speed
Google thinks that Largest Contentful Paint, Cumulative Layout Shift, and First Input Delay are the best way to measure the real user experience page speed.
Canonical Tags Added by JavaScript are OK
It’s acceptable to inject canonical tags using JavaScript, provided you can see the correct canonical tag in the head when you render the pages with Google’s testing tools.
JavaScript Served by Google Tag Manager is OK
It’s OK to serve JavaScript on a page using Google Tag Manager, but it will be included in page speed measurement. However you should avoid it if possible and include the JavaScript directly on the page
Serving Non-200 Pages to Users as 200 Status is Not Considered Cloaking
It is not considered cloaking if you serve a 200 status for non-200 URLs on a single page app to users, but serve a non-200 status to Googlebot.
Google Does Not Index Browser State Data
Google only indexes content in the DOM, and won’t take any content from the browser state.
Google’s Testing Tools May Show False ‘Other Errors’ Due to Limited Bandwith
Google’s testing tools don’t cache resources, and have limited bandwidth to fetch a page’s resources which may result in ‘Other Errors’, however, these don’t impact Googlebot. You can use the URL inspection tool in Search Console to see the HTML rendered by Googlebot.
There is No Best JS Framework for Google
There is no specific JS framework which is best for Google. You should choose the framework which provides the tooling you need and suits the project and developers.
Links Added by JavaScript Won’t be Discovered Until Rendered by Googlebot
If you add links to a page using JavaScript, they can’t be used for discovery until Google has rendered the page.
Googlebot Won’t See Any Content Added by Service Workers
Googlebot won’t see any content added by services workers as it doesn’t use them.