How to optimise for voice search

The near ubiquity of voice-activated smart speakers and phones means that users’ entire interactions with search engines can now happen without ever opening a browser. 

Spurred on by the ease and efficiency that comes with being able to operate devices simply by talking to them, Juniper Research has predicted there will be over eight billion digital voice assistants in use globally by 2023. 

That’s enough, based on population growth forecasts, for the world’s population to have roughly one each.

Voice search, therefore, is not only here to stay, but looks set to become the norm. 

But what is required to optimise content for voice search? And how much does it differ from accepted SEO best practices? We take a look below at voice search optimisation.

How does voice search work?

To gain a better understanding of how to optimise for voice, let’s first look at how the process works.

Voice search converts speech into text and then runs this through a search engine in the same way it would with a typed query. 

The major departure from a standard query is that we then do not get to see and scroll through every single result. Instead, the featured snippet from the SERPs is read back to us by default. 

While on some devices we can then cycle through to other results if this is not satisfactory, both Google Home and Amazon’s Echo Dot only ever provide the featured snippet. 

Both this and the fact we are using our voices rather than hands to make searches throws up optimisation considerations, which we’ll discuss next.

Voice searches are much more conversational

Google claims that up to 70% of searches on Google Assistant use what is known as ‘natural language’. This differs from the traditional way we use search engines because it directly mimics human speech.

For example, what was once ‘omelette best way’ now becomes ‘Hey Google/Siri/Alexa, what is the best way to make an omelette?’. 

This has the biggest impact on keyword research. We now need to judge the suitability of keywords based on how natural they sound, not just traditional metrics like average search volume. 

It also means question-based keywords are more important than ever, as almost every voice search query involves a question. These can’t be filtered for on Google’s Keyword Planner, so it’s best to use a search listening tool like AnswerThePublic.

The further knock-on effect of this is that what would previously be thought of as overly long keywords are now relevant. This is because searches this long are both easier to make and much more natural when using our voices. 

To address this when optimising for voice search, don’t discredit keywords from keyword research based on length alone. 

Next, we’ll look at how these subtle tweaks to keyword research can be applied to content.

How has voice search changed content optimisation?

Overall, the rise of voice search hasn’t meant radical changes to how to best optimise content. Instead, it is a case of incorporating the keyword research discussed above.

An immediate way to address this is to include FAQ pages. They are the perfect forum for posing and then succinctly answering a huge range of questions, with voice searches 1.7x more likely to come from an FAQ page than a desktop search. You can also rank for hundreds of keywords on the same page, catering for a huge range of potential queries.

In addition, make sure to include the kind of long keywords that are typical of a voice search throughout your copy. As with any content, don’t keyword stuff with these, but use them naturally and wherever possible. This will also help appeal to the natural language element of voice search.

The featured snippet is key

But perhaps the most crucial way to optimise for voice search is to directly target the featured snippet. Have your content rank in this position, and it’s guaranteed to be presented to users first after a voice search. 

The featured snippet has also become highly sought after for desktop searches, as well as being responsible for the rise of the ‘no click’ search. This is where a user’s query is answered by the snippet without them having to leave the SERPs. 

Thankfully, best practice for ranking in the featured snippet mirrors what is needed to optimise content for voice search. 

Be sure to create content that is highly informative and provides detailed answers to likely questions that arise from your subject matter. If these have already been uncovered as part of voice search keyword research, this becomes even easier.

Another step is to provide a section of copy that can be directly lifted and used as the featured snippet from within a longer piece. This is typically around 40-60 words and can be clearly signposted by using ‘What is…’ or ‘What are…’ in the heading. 

If going down this route, it’s vital that your definition is objective. Google is keen to avoid serving opinion as fact, so will never include a specific point of view in a featured snippet. 

Not every keyword can rank for a featured snippet, so use a tool like AhRefs to find out which do and then apply the above. 

Optimising for voice search: small adjustments that can have a big impact

The sudden prevalence of voice search is unlikely to lead to an enormous overhaul of how to optimise content for search engines.

There are still, however, several key factors to consider when producing new content or updating existing content now that voice search is so popular. Many of these already fall under traditional SEO best practice, such as providing information rich content.

The featured snippet looks set to become even more crucial, and is the best section of a SERP to target when optimising for voice search.