How to Optimise for Voice Search

• Voice search ranking factors
• Understanding user intent
• Add schema markup to web pages
• How AI is evolving voice search
• Not all content needs to be optimised for voice search

Voice search isn’t exactly a new feature, however with the rise in popularity of smart home devices, such as Amazon’s Alexa and Google Home, voice search is quickly becoming an aspect of marketing that can’t go ignored if you want to have the edge over your competitors (or keep up with them!).

But even at this moment in time, there are no clear guidelines from Google themselves about optimising for voice search. So that’s when the keen, smarty-pants SEO experts step in to be ahead of the curve.

Who is using voice search?

Understanding the demographic behind voice search usage is the very first step. Luckily, Stone Temple Consulting have released their findings after surveying 1,000 people about their use of voice commands:

• People are becoming more comfortable using voice search in public.
• The 35-to-44 age group is the largest segment using voice search.
• The 25-to-34 age group is most comfortable using voice search in public.
• The heaviest users of voice search have an income above $50,000 per year.

Familiar with Alexa and Google Home? Gartner’s research below predicted that by 2021-end, 75% of homes in the US would have a smart speaker. Meanwhile, data from Statista projects the smart speaker market hitting $35.5 billion by 2025 as more people purchase voice-based speakers for their homes and the device becomes further integrated into business and consumer settings.

According to Marketing Dive, the US smart speaker market will grow by 46% this year, and we can assume this trend will be very similar over here in the UK, too:

Voice search ranking factor

It’s difficult to precisely predict how users will interact with their devices regarding voice search, and it may broadly differ between those devices.

A study by Backlinko, based on 10,000 Google Home search results, has some fascinating findings:

• Answers are 29 words on average. So, when structuring the data you want to become a voice “answer,” make sure it’s short and to the point. This means formatting the page so an answer can be quickly drawn from it and understood as a complete answer to the question.
• The average writing level of results was ninth-grade reading level, so keep it simple.
• Presently, voice search results seem to serve a more generic audience. I don’t expect this to last long; ranking for the present requires writing to the masses.
• Google may eventually cater the reading level to the individual searching and implied education level of the query.
• The average word count of pages used to draw voice search results was 2,312 words, implying that Google wants to pull results from authoritative pages.

So it’s important to note that when creating content, we need to bear in mind the entity being discussed and the intent we need to satisfy when optimising for both voice and general search.

What’s an Entity?

To put it simply, an entity is a noun connected by relationships. When somebody makes a search query, Google looks up their database of entities and tries to determine the most relevant answer to the searcher’s intent while comparing it with other entities to assess various traits. For example, when searching a particular celebrity’s name, this entity could be linked to multiple other entities by way of family relationships and the celebrity’s birthday or any work they have been involved in (such as particular movies, bands, albums, etc.)

But this is when, for example, searching for somebody who has the same name as a celebrity, the result with more prominence will be served – most likely the celebrity in this instance – particularly when there is a relevant Wikipedia page available.

For example, If I were to say, “Ok Google, who is David Foster?” I’d be presented with a result from the Canadian musician’s Wikipedia page instead of our colleague’s profile here on Soar Online. The reason is simply that the Wikipedia article will have sufficient content on the musician’s page itself and supporting articles linked to that entity such as associated acts, albums, nominations, etc.

So understanding how entities relate to each other and giving concise and easily digested information on as many related topics as possible will ensure that Google sees us as the authoritative answer.

User Intents

So far, it’s safe to say that optimising for as many related entities, questions, and related content as possible is a critical factor when trying to rank for voice search. When Google determines which entity to serve to the user, it primarily comes down to the user’s intent, which is based on a combination of related factors from previous queries. It also uses a system of metrics related to authority and relevance to determine which would win in a generic environment.

If we assume that Google uses their patent of “Ranking Search Results Based On Entity Metrics”, we can determine that Google uses these four metrics to determine the strength of an entity:

• Relatedness. As Google sees relationships or entities appear relatedly on the web, they will connect these entities.
• Notability. This relates to notability in the field, taking into account the popularity of the entity in question and also the popularity of the field as a whole.
• Contribution. Google will weight entities by reviews, fame rankings and similar information.
• Prizes. Google adds more weight to an entity or aspect of that entity based on accolades and awards. We’re not referring to a lotto but rather something like a Grammy.

The next step in ranking on voice search is to isolate which entities will have these metrics and cover them by writing compelling, targeted content.

Cover the core answer and consider all the various entities connected to that answer to reinforce that you’re referring to the same entity and have the authority and information to give the best solution.

Provide context with schema markup

If you’re not acquainted with schema markup, we recommend you learn about this SEO technique.

It would be best to use schema markup on all of your content as it helps search engines understand what your web pages are about by providing context, which helps you rank better in search and for more specific queries made by voice search.

According to a Backlinko study, only 36.4% of voice search results contain schema markup, despite its impact on content for voice search. For example, if I were to ask Google “Find a recipe for Chicken Casserole” and followed that question with “Can I have the ingredients and instructions to make it?”, Google would pull down a list of the information from a site with recipe markup on the page.

Schema markup makes it easier for search engines to align content with user search queries, with approximately 40% of spoken responses being pulled from featured snippets.

How to Optimise for Featured Snippets and Voice Search

• Identify your top 10 question-based, long-tail and filler keywords, this might require some digging, but the research will be instrumental to your success.
• Pin down related search queries using answerthepublic.com, which is a valuable keyword tool that visualises search questions.
• Restructure your content into a series of questions and answers as voice search queries contain many question words such as how, what, when, why, where. 
• Maintain a conversational and natural-sounding tone.
• Optimise your content for rich answers, meaning your content should contain either all or at least one of the following: featured snippet, knowledge box, knowledge panel or a knowledge graph.
• Optimise loading speed as voice searchers and spiders won’t wait around for a slow website to load.
• Ensure your page is mobile-friendly to ensure visibility and increase content exposure.
• Improve content readability by ranking around nine on the Flesch Kincaid Grade level and avoid using technical jargon.

AI and voice search

Although voice search technology is still relatively primitive, artificial intelligence (AI) is helping digital assistants evolve and become capable of interpreting and understanding the nuances of language such as:

Spelling errors and queries

If you ask Alexa, Google Home or Siri, “Where is Buckingham Palace?” and follow that up with “How do I get there?” and “How old is it?”, the search engine can interpret what “it” is based on your initial voice search.

Even if Amazon, Google or Apple cannot immediately answer your search query, the voice assistants will ask you to help them find the answer by providing more context.

Answers based on location

Suppose you’re at Birmingham New Street (Grand Central) train station and you’re trying to find the Town Hall. In that case, you can ask voice search: “How far is the Town Hall?” without specifying that it’s Birmingham Town Hall you’re seeking, and Google will understand that the “Town Hall” you’re referring to is the one in Birmingham based on your location.

To make instant purchases

You can also ask voice assistants to order products or services to make shopping easier. For example, asking Siri to “order me a pizza from Pizza Hut” will bring up several listings of Pizza Hut restaurants ordered by distance. Siri will then ask whether you want him to call the number to book or provide you with the directions to the restaurant.

Many tech giants have already optimised their content for voice search to have the edge over the competition – now it’s your turn! But remember that voiceover is not the dominant means for search, so don’t neglect your other SEO methods.

Not all content needs to be voice-search friendly

While the rise in smart speakers means that creating content geared towards voice search is important, you shouldn’t necessarily build your whole SEO strategy around voice searchability, especially as the new technology hasn’t yet hit critical mass.

Voice search isn’t replacing text search; it’s complementing it.

While marketers and content creators can no longer ignore the impact of voice search, you should think of it as a supplement to text search rather than a replacement. So, instead of optimising all your content for voiceover, create some low-hanging fruit for the early adopters, which will be more effective for SEO as text search is still the dominant form.

For example, suppose you’re a sports company that has optimised a piece of content on a football match purely for voice search; this may not be the most effective way to drive engagement. Remember, Siri and Alexa will update a user on the score but won’t narrate the action in detail, so if you’re trying to reach users interested in sports commentary, you could be failing to reach a key segment of your audience.

Currently, voice search tends to be focused on providing results relative to the user’s location, such as:

• To ask for directions
• Locate a company building
• Look up cinema showtimes

People also tend to use digital assistants to search for quick facts. Answers are usually pulled from top-ranking pages or the featured snippet, so it’s also essential to add schema markup to your web pages to help boost your chances of being a voice search result.

Summary

To put it simply, if you want to optimise and rank well for voice search, it boils down to having these three things:

• A strong domain
• Strong content
• Content divided into logical and easily digested segments