Have you ever entered a query into Google only to be presented with the answer directly via the search engine results page (SERP)?
Search results of this nature of often referred to as “answer boxes”, or “quick answer boxes”.
Try entering common questions in to Google and you will be surprised at just how many results return answer boxes. A couple of specific searches I tested were:
“how to boil an egg”
“how to grow a beard”
You will notice that the information returned can be presented either as a concise paragraph or as a numbered list of steps. The way in which the answer boxes are displayed depends on the source from which Google extracted the content, and the way the content is structured on that web page.
You will also notice that Google doesn’t always extract the answer from the top ranking result. In the second example, beards.org is actually in 5th organic position for the query “how to grow a beard”.
Google also recognises and appreciates that they don’t always return the most appropriate results in answer boxes. Users can therefore provide feedback should they see a potential issue with the SERP.
In the example below, the broad term “changing address” is used. However, Google has presumed that this query specifically relates to changing the address on an individual’s driving licence, deeming the content provided by the DVLA as the most helpful. Although the DVLA site does rank highly for this term, all of the other organic listings are actually to do with moving house, so one would argue that this answer isn’t the most helpful to my query. So if I wanted to, I could provide feedback to the search engine:
A brief introduction to Knowledge Graph Optmisation (KGO)
The existence of answer boxes is once again a result of Google’s drive to return the most useful and appropriate information to their users for any given query. This has only been intensified by the rapid increase in mobile and multi device search.
In essence, answer boxes are powered by Google’s Knowledge Graph, a system launched by Google in the summer of 2012 which is able to understand facts about people, places, and things – and how all these entities are related to one another.
Search Engine Land have an ever increasing library of articles relating to the knowledge graph, and state that “Google’s Knowledge Graph is used both behind-the-scenes to help Google improve its search relevancy and also to present Knowledge Graph boxes, at times, within its search results that provide direct answers.”
With the SERPs seemingly drawing from the Knowledge Graph on an ever increasing basis, both web masters and SEOs alike are now considering how they might construct and optimise their websites in order take advantage of entity based search.
The subject of the Knowledge Graph is vast and complex, so within this post I will attempt to keep things simple and outline some of the latest best practice techniques that can be used to “optimise” the Knowledge Graph, with a specific focus on what I will call “answer boxes”. For further reading on optimising a brand for the Knowledge Graph, I’d encourage you to check out this post.
The term “Knowledge Graph Optimisation” (KGO) was coined AJ Kohn, who has since produced a comprehensive list of “tactics” which he suggests “can help you optimise your site’s connection to the Knowledge Graph”.
Some of his tactics include:
- Implementing structured data on your website where appropriate
- Using nouns and entities in your writing and writing in a natural manor
- Carefully considering your external linking activities
- Managing external data on profiles such as Wikidata and Freebase
- Striving to be featured on powerful sources such as Wikipedia
Creating content with answer boxes in mind
Google encourages website owners to make their content easily accessible and engaging for their users. Not only should web masters structure and optimise their content in line with traditional best practice onsite SEO, they should also strive to create great content and use structured data to enable the search engine to fully understand what their content means.
I appreciate that creating “great content” is easier said than done, so at this point I’d like to suggest what great content looks like, particularly when considering trying to influence answer box results.
Great content always starts with a genuine purpose, and more specifically, a topic that is of genuine interest to a specific set of users. There’s no point in creating content for the sake of targeting keywords, and the value added to the end user must always take priority.
But that’s not to say that keyword research shouldn’t shape or influence how content is titled and written, but the needs and interests of your target audience should be considered above all else.
Having a thorough understanding of your websites audience will make writing great content easier. Creating research based content based customer profiling and buyer personae will ensure that you are able to specifically addresses and answer questions regularly posed by different user types.
In terms of technical implementation, content should always be structured with user experience in mind. Ensure that your content is clearly presented on your website, making sensible use of formatting such as headings and bullet points, and attempt to present any key points contained within your copy in a straightforward fashion. The content should also always align with best optimisation practices, making use of meta data, headings, alt tags and URL structures. Also ensure that structured data is implemented within the page content wherever appropriate.
Although creating useful content is well under a website owner’s control, there are other factors which need to be considered when striving for content to be featured within Google’s answer boxes, because not all website’s have the ability to return results of this nature.
The authority of the website’s domain and associated pages plays a significant role in the level of trust that Google places in a particular website or webpage, and this can influence answer box results. Remember, though, that this doesn’t mean that you have to rank in first position for the search query. Good user engagement metrics such as visitor bounce rate, average time on site and pages per visit also have some impact. It’s therefore also vital that your website is both desktop and mobile friendly.
Optimising content for Google’s Knowledge Graph is a relatively new concept that’s subject to ongoing testing. There are no set guidelines, and thought leadership on the subject is ever evolving.
Now is the perfect time for website owners and search marketers to start considering ways in which they answer the questions posed by their users, and to ensure these answer are provided in the form of rich, well-structured, and well optimised content.