Does The "Accepted Answer" Schema Do Anything for SEO?

Over the decades, Google has created and introduced a lot of different tools and techniques for improving how a site appears in search engine results. Some of them come and go – Authorship is a big one – while others stick around and prove to be very useful.

One of the most useful relatively modern inventions is Schema markup. Schema was actually invented all the way back in 2011 by the combined efforts of Google, Bing, and Yahoo. Over the years, the organizations in control have shifted (it's now Google, Yahoo, Microsoft, and Yandex), but the core functionality of the markup code persists.

The idea behind Schema is simple: code that identifies types of data on a page, so that data can be handled properly, rather than a search engine needing to guess what the data is from the context and position, and possibly get it wrong.

Schema itself is far from simple, though. With over 800 different schema types, there's something for nearly every piece of data on your page. Trying to implement complete Schema for everything that could possibly be applicable is an endless rabbit hole.

As webmasters and marketers, we need to choose where to focus our efforts, and that can mean picking the most relevant and useful schema types to implement. Product listing element markup, local business data markup, review markup, and video markup; these kinds of schema can be very useful for giving Google data to display your page properly in their search results pages.

One of the schema types I've seen mentioned more and more often recently is the Accepted Answer markup. What is it, does it help with SEO, and should you implement it when you're adding Schema to your blog?

What is the Accepted Answer Schema and How Does it Work?

In December of 2017, Google started to show something new in the search results: a question-and-answer card. When you typed in certain kinds of queries, specifically queries that formed questions, Google would look for websites that reflected discussion boards covering that topic. Web forums in particular, but also sites like Quora or Reddit, could be discussions of that question. This also comes up a lot with support forums, like Google's, Microsoft's, and Facebook's support communities.

Google would place a box beneath the page with the question, and that box would contain two or three small sub-boxes, each with a summary of one of the answers to the question. Sometimes, if the board or resource had voted on the question answers, Google could display how many votes the answer received and order them appropriately.

In December of 2018, one year later, Google added an improvement to this system. Where the data was available, such as through community voting or through a highlighted "top answer", Google could flag one of the answers as the best answer.

Generally speaking, the whole answer wouldn't be displayed. Not because Google wouldn't do it if they could, but because the amount of space was limited, so only a dozen words or so could be shown. Still, that was often enough to determine whether or not the answer was good enough to click through the page to see it.

While Google could populate this result with data they index, they also made sure that Schema had a way to specifically flag this data for display. This is the QAPage schema.

Valid and Invalid Uses of the QAPage Schema

Before digging into the specific components of this schema, it's important to consider whether or not it's viable for you to use. A lot of marketers get this wrong, so I wanted to touch on it directly.

The QAPage schema is meant specifically for discussion and support pages. For example, Google's own help center community has pages like this, which is a discussion thread, forum-style, with an original poster posing a question and a series of other users and experts leaving their answers. You have a question, you have a recommended answer, and you have all of the replies.

In the case above, it's a lengthy question with no current clear answer as of this writing, so there's not a flagged top answer. Many other questions have a "relevant answer" but no top answer, and others have a recommended answer. You can even see the number and type of answers from the question index, as well as see how many threads have zero answers at all.

There are a couple of critical attributes here.

• Different users can submit their answers.
• There's a feedback mechanism for surfacing a top answer.
• There's just one question per page.

These are all required for using the QAPage schema.

Other kinds of pages are not eligible. These include:

• FAQ pages with multiple questions on the same page. QAPage is for URLs dedicated to one single question.
• FAQ pages where the site owner writes the question and answer. QAPage is for discussions involving multiple people.
• Non-discussion pages, like essays or blog posts focused on the question, or a how-to guide.

This is where a lot of marketers get tripped up. One of the best ways to find new topics to write about on your blog is to look for questions your users or customers are asking, and write content that answers those questions. That sounds like a perfect use case for schema that flags a top answer, right? Unfortunately, that's just not the purpose, and if you use it that way, Google will either ignore it or even potentially penalize you for trying to game the system.

The Components of the QAPage Schema

You can't just slap a "top answer" marker on any piece of data and assume it's going to help you. It won't pass validation if you do. It needs to be a component of the overall Q&A page schema markup. I'm going to run down the important elements of the schema, but if you want the full rundown, you can view the tag page here.

QAPage

At the top level is QAPage itself. This is the encompassing wrapper that identifies the content within it as a question and answers set.

This overall attribute only has one required property, and that's Question. This is because there can be a lot of discussions out there with a question but no answers. Google's own support forum, which I linked above, has tons of these. They don't want to deal with an infinite sea of "errors" for questions no one has answered, so the answer is actually not a required part of the schema for it to be valid.

Question

Question is an attribute within the QAPage schema. It identifies the question central to the page. For a discussion forum, this might be the title or first post of the thread. There can only be one Question per URL.

Question has three required properties and several recommended properties you can add to it.

The required properties are:

• answerCount. This is a number and represents the number of answers there are to the question, from 0 to however many there are. This is total answers, so if you have 20 answers paginated across two pages of 10 each, the number is still 20.
• Answer. This can be either acceptedAnswer or suggestedAnswer, and represents the best answer from among the answers given to the question. I'll discuss the difference between these down below.
• Name. This is the text of the short form of the question. Often, this is just the title of the post, though not always.

What's the difference between an accepted answer and a suggested answer?

A suggested answer is one of the possible answers to the question. All answers to the question should be tagged with this flag by default. There can be zero suggested answers, though. For example, if the question only has one answer and that answer is flagged as the accepted answer, there can be zero suggested answers.

The accepted answer is the top answer, the one the site, the community, or the question asker has flagged as the best answer. The mechanism you use to flag the answer can vary, and different discussion boards use different options. A question also doesn't have to have an accepted answer.

Generally speaking, the accepted answer is a special flag for a verified best answer. Whether that means an expert checked and validated the answer as objectively correct, or the answer comes from a product expert who added it in and closed the question, or even if the question asker validated the answer and reported that it was correct for them, it doesn't matter. Google doesn't care how the top answer is chosen.

To recap, for answers, a question does not need to have any. If there is an answer, though, it needs to either be flagged as a suggested answer or as an accepted answer. Untagged answers result in invalid schema.

What about those recommended but not required properties for the schema? There's a bunch of them, most of which are self-explanatory.

• Author, information about who asked the question.
• Author.url, usually the author's user account profile page link.
• Comment, generally the text of the post that clarifies the question and is not an answer.
• DateModified and datePublished, for the question.
• Image, if there are images in the question.
• Text, for the full text of the question.
• UpvoteCount, for the number of votes a question gets.
• Video, for if there are videos embedded in the question.

Answer

Within the Question property, each answer is tagged as an Answer, and either as a suggested answer or an accepted answer, as I just described. The Answer markup also has one required property, which is text, just the text of the answer itself.

It also has recommended optional properties, almost all of which are identical to the question properties listed above.

The one major addition is URL. Google recommends that each answer be given its own on-page permalink that goes directly to the answer. This way, they can include it in the SERPs, so if a user clicks on the rich result, they're taken to the top answer directly rather than having to scroll through a discussion page.

Comment

Another optional addition to the Q&A paradigm is the comment schema. This can be used to flag clarifications and additional discussion, and is nested within either the question or answer attribute.

If you think about a typical support forum thread, you have the OP asking a question, then an expert asking clarifying questions, and the OP responding with more information. Neither of those is an answer to the question. So, rather than tagging them as answers, they should be tagged as comments.

Is the Accepted Answer Flag Valuable for SEO?

Yes, but with a strong caveat.

If you're running a blog or website, there's a pretty good chance nothing on your site qualifies for the QAPage schema.

Up above, I mentioned that the QAPage markup is meant for discussions, where one page is centered around one question and multiple users contribute to answering the question. It's very narrow, and a lot of websites don't have ways to use it.

I've seen a few marketers here and there try to use this by flagging their blog post as a question-and-answer set, and flagging blog comments as other non-top answers. This is a pretty clear attempt to exploit the system, and Google will generally ignore it, if not find it invalid or even lightly penalize you for using it.

IF your site has a discussion forum that fits the paradigm for a question, discussion, and answer system, then using the QAPage markup can help boost your SEO for those pages.

Within that paradigm, using the Accepted Answer flag is part of it. Just having answers and getting the QAPage result card in Google Search is good for SEO, and having a flagged top answer can also be beneficial to help it stand out that tiny bit more.

In my experience, most of the site owners who ask this question do not have sites that can take advantage of it. Those who can take advantage of it, meanwhile, are usually using a plugin that manages the discussion format for them, and that plugin will handle the schema automatically, so knowing how it all works and needs to be implemented isn't necessary.

So, that's your answer. Yes, it can help your SEO in the relatively narrow event that you have pages that can qualify for using it. Most of you reading this? Probably don't.