Page 8 - SEO Mad Scientist 2020 Report

Would Schema Result in an Advantage if Google Displayed the Snippet?

Test Setup:

We wondered, would video (or any) Schema markup result in an SEO advantage IF Google
displays the snippet? Because we had yet to find a Schema markup that actually pushed rankings
on it’s own, our theory was that Schema may have SEO value, but only if Google is actually using
(or trusting) that specific Schema. So we needed to test Schema that was actively being used by
Google to dive deeper into Schema’s SEO advantage when Google displayed the video snippet.

Schema doesn't seem to help on its own but how, when and why may need to be considered. To
take it further, we set up a test of multiple kinds of Schema, testing the removal of Schema that is
actively being displayed in the SERPs, as well as adding FAQ Schema to multiple pages. Then,
we checked the rankings of those pages and how they correlated to which FAQ page Google
displayed for that search term.

Test Results: February 21, 2020

In our initial findings, every case where we added FAQ Schema and were able to see it indexed
in the SERPs, we noticed a slight fluctuation in upward SERP movement outside of the normal ups
and downs. BUT, when removing the FAQ Schema from a stable ranking page, we did not see
any downward movement for the listing. This makes sense since Google says they use Schema
to help determine what a page is about, but if they already have that understanding then why
would removing the Schema make a difference.

Update: February 28, 2020

We reviewed other SERP snippets and found some had markup, but many did not. This leads us
to believe that Google isn't actually using Schema to determine snippets, but instead using
machine learning to better understand media, which then triggers these snippets. We believe
Google was attempting a similar approach with their new link attributions. Google even stated
that these will not impact how links are treated at the moment, which leads us to believe they are
using that data to further their machine learning to better understand link variations.

8