Structured Data is an Extremely Light Signal

Google’s John Mueller offered insightful comments in a Twitter conversation about structured data and how Google uses it. His comment touched on the importance of structured data as a Google signal.

Twitter Discussion on Structured Data

The Twitter discussion grew out of a seeming contradiction in a statement made in a recent article, Google Doesn’t Read Unsupported Structured Data.

There was no contradiction and the details are besides the point. Mueller clarified the seeming contradiction and then went on to discuss how Google uses structured data.

What followed are the tweets that matter.

Extra Structured Data

Google’s Mueller began by clarifying what kind of structured data (SD) was considered as extra. In this case, he called attention to structured data types and information that is obvious.

The first issue was about using the WebPage structured data type instead of a more specific data type:

“The thing is a lot of “extra” SD is super obvious. “This is a webpage”, well, that’s shocking, seeing we’re crawling webpages.

Lots of other SD is already clear from the page text (Is it a Ford car or a Ford president? No need for SD unless you’re really creative in writing).”

The thing is a lot of “extra” SD is super obvious. “This is a webpage”, well, that’s shocking, seeing we’re crawling webpages. Lots of other SD is already clear from the page text (Is it a Ford car or a Ford president? No need for SD unless you’re really creative in writing).

— ? John ? (@JohnMu) December 5, 2020

The WebPage structured data type is considered to be very general and according to Mueller, it’s “super obvious.”

Schema.org says this about the WebPage structured data type:

“Every web page is implicitly assumed to be declared to be of type WebPage…”

That seems to mean that you don’t need to use structured data to tell Google that a webpage is a webpage, it’s implied.

So that actually frees you up to use a more specific structured data type.

This is actually a fairly common error. Probably because almost everything you can document with a more specific structured data type can be declared in the more general WebPage structured data.

Structured data for Google can be split into two kinds:

1. Rich Results Structured Data
2. Non-rich Results Structured Data

Rich results structured data can qualify for a search results listing that is enhanced, i.e. a rich result. Non-rich results do not qualify.

But it’s probably best to be more specific. For example, these are rich results structured data types that can be used on a webpage:

So if you’re already using a specific structured data like Article, there is no reason to use the WebPage structured data because it’s superfluous.

There are other structured data types that are non-rich results structured data that won’t show rich results but are more specifically about webpages:

• AboutPage
• CheckoutPage
• CollectionPage

While those exist, they won’t qualify for any kind of rich results. But that doesn’t mean you can’t use the AboutPage structured data to communicate to a search engine what that page is about, although it’s highly likely Google can already tell from the content that it’s an “about page.”

Communicating what a page is about is generally a good idea and if you feel it might help to make something clear, then go ahead and use it.

But if the structured data type is not listed on Google’s developer pages as one that could qualify for a rich result, don’t expect to see that kind of result from it, set your expectations lower.

A rich result is like those featured snippets that show at the top of the page or stars shown in the search results for reviews.

Structured data types that are not listed in Google’s developer pages are highly likely to not quality for a rich result. Those are non-rich results structured data.

Non-rich Results Structured Data

John Mueller next discussed non-rich results structured data and said that it can be helpful but in a limited way.

He used the acronym RR to refer to Rich Results and SD to refer to Structured Data.

This is what he tweeted:

“What about the non-RR SD that’s not absolutely clear from the page? It can be helpful, but it’s also limited in the extra value it provides.”

Structured Data Signal

He finished the above tweet with a statement that seems to say that structured data is a light signal.

Mueller’s tweet:

“How do you rank something purely from SD hints? It’s an extremely light signal. If you’re worried, make the content more obvious.”

So… structured data is an extremely light signal? By signal, did he mean a ranking signal? Or a signal related to what the content was about?

John Mueller didn’t elaborate on those details.

Structured Data and Rich Results Types

Mueller reaffirmed the value of structured data, particularly where it might be difficult to accurately understand specific details.

John tweeted:

“I do see long-term value in SD for RR types where parsing the page is hard. Event dates? Venue phone numbers? Ratings & scale? Article date? It’s possible, but hard + unique per page/site, and embarrassing when we highlight it incorrectly.”

Structured data is useful for communicating information where it’s critical to get it right, like phone numbers, ratings and dates.

But Mueller also said that structured data is “extremely light” as a signal, a statement that might need clarification as to what kind of signal he was referring to.