In a hangout, during the submitted Question and Answer segment, an SEO professional asked John Mueller about using a chat tool as a help service.
They have it implemented in such a way that the button is a small bubble in the bottom left corner on mobile.
This covers a bit of content, because the bubble is an overlay. Will they have a ranking problem because of the usability problem?
They asked if they should have an additional close button? The bubble is visible on all pages, also when users change pages.
John explained that he doesn’t know about the usability side. He can’t really give any advice there.
But, regarding the kinds of searches in general ranking guidelines, there are two aspects that could come into play. And it’s something where they probably have to make a judgment call on their side.
On one hand, there is the intrusive interstitial guidelines that Google has, where if you have an interstitial on your pages that is intrusive for users, then that’s something that they would recommend avoiding.
That’s part of the page experience ranking factor.
However, they don’t have a fixed number of pixels or anything like that, which they would say would be intrusive.
And John’s guess is if they have a chat bubble in the corner, then this would not be considered intrusive by their users.
The other aspect is Core Web Vitals. In particular, John thinks the LCP and perhaps the CLS could apply here.
The time it takes to load the page, and if the content shifts around while the page is loading, depending on how the chat bubble is implemented in the corner, this is possibly something that could also be playing a role.
And this is something that they can test where you turn that on and off, and you try this out in the browser, to see what the effect is there.
The effect that you see in your browser is kind of a different effect than users may see.
But, this gives you a bit of guidance.
In particular, for Core Web Vitals, they use the metrics that users see which they call “field data” in the CrUX report, or Chrome User Experience report.
If you test it yourself, then they would consider it to be lab data, and this is slightly different than from what an average user would see.
Primarily, this is due to the different connections that users have, or different devices, the capabilities, how fast they are, how much RAM they have – those types of differences. But, usually, if you turn it on and off, you will quickly see whether there is an actual difference or not.
And, if there is a difference, you would have to make a judgment call here.
Is this something I want to worry about, or not?
This happens at approximately the 34:24 mark in the video.