Jan 7, 2019

Prepare for the Future of Voice Search

There are plenty of impressive statistics regarding digital assistants and voice search. To cite just a few:

  • Over 20% of mobile searches are by voice (Source: Google)
  • There will be 67 million voice-assisted devices in use in the US by 2019
  • Voice commerce sales will reach $40 billion in the US and UK by 2022

Clearly, this is a growing industry and the physical evidence of the marketing opportunity is all around us.

Devices such as Google Home and the Amazon Echo range are increasingly prominent in living rooms, kitchens and bedrooms in Western countries. Furthermore, anthropomorphic assistants such as Siri, Alexa and Google Assistant are embedded in a wide range of smartphones, cars, and even fridges.  

The Google Assistant interface
The Google Assistant interface

It is a trend that sits at the intersection of increased mobile use, sophisticated machine learning algorithms, and the symbiotic relationship between people and technology. As such, its continued rise seems inexorable.

But what does this really mean for brands? Does voice search require a separate strategy? And if so, what does it entail?

After all, despite the grandiose noises about the seismic shifts voice search will bring, 62% of marketers have no specific plans in place for voice search.

This article will therefore examine what we know about voice search today, before listing some best practices to put any website in the optimal position for future success.

Voice Search: A Natural Extension of Semantic Search

The usual communicative evolution begins with speech and moves on to written language. However, search engines have experienced this in reverse.

The signals that we innately pick up on, such as intonation or memories of past interactions, are exceptionally difficult to incorporate into a search engine.

When I ask a friend, “What would be the best smartphone for me to buy?”, they can tailor the recommendation based on my preferences.

Scaling that level of insight requires very significant natural language processing power, along with the information retrieval technology to sift through billions of results and locate the right one. Smartphones provide more contextual information than desktop computers, but a search engine still needs a reliable way to process and utilize so much data.

The image below highlights the levels of difficulty for a search engine and the technology required in each case:

Prepare for the Future of Voice Search

This matters when we consider voice search strategy. People adapt their behaviors based on the possibilities at their disposal. As marketers, understanding those behaviors is essential if we want to cut through the noise and connect.

Brands create the content that leads a consumer from question to answer. A search engine is the interlocutor that makes the connection.

Google’s Hummingbird algorithm ushered in the age of semantic search, making use of the Google Knowledge Graph to understand the relationship between entities and deliver something approaching conversational search.

Ask Google “Who is the king of Spain?”, and it will respond “King Felipe VI”. Next, ask “Who is his wife?”, and it will respond “Letizia of Spain”.

Google infers that ‘his’ refers to King Felipe. This is a subtle but significant shift that affects how we should create and promote content through search. We can now have conversations through search engine optimization (SEO), rather than one-off exchanges.

Semantic search is changing how people find information and it is heightening their expectations. As the search engine’s capabilities change, so should ours as marketers.

In essence, this development is a natural and vital component of voice search’s rise. Google reports that people are increasingly searching for queries using words such as ‘me’, ‘my’ and ‘I’.

As an indicator of the modern consumer’s requirements from online content, this is very telling. People would only ask these questions if they expect the answer to be personalized and unique to them. These searches are typically carried out by voice, rather than text.

Once more, this points to the difference between voice search and traditional search. Consumers are treating digital assistants as exactly that, a personal helper to get things done quicker and easier than before. We expect the assistant to ‘know’ us.

Another question that is often raised is just how genuine the commercial opportunity is for voice search.

A 2017 study from iProspect revealed that while people predominantly use voice search to get information or enable actions like turning on lights, they are also using them to find stores, research and purchase.

Moreover, the distinction between voice search on mobile and with a smart home device is rather marked. This is perhaps to be expected, given that we carry mobile devices with us and they have screens, whereas home devices tend not to, but it does bring important implications for brands. Mobile phone screens provide a canvas on which to display choice and information, while a home device must deliver one, authoritative answer.

From this, we can start to understand the drivers – both technological and human – that have seen voice search grow so rapidly.

Voice Search Best Practices

Technical SEO

  • Focus on speed and mobile-friendliness. A study of 10,000 voice search results by Backlinko shows that the time to first byte for a voice search result is significantly shorter than for the average webpage. With Google’s “speed update” rolling out for all users now, this should be the first port of call for any mobile or voice search strategy.
Search speed: The time to first byte for a voice search result versus the average webpage. Credit: https://backlinko.com/voice-search-seo-study
Search speed: The time to first byte for a voice search result versus the average webpage. Credit: https://backlinko.com/voice-search-seo-study
  • Use structured data on all landing pages. One sizeable challenge for digital assistants is that they must comb through trillions of pages to identify the elements that will answer a user’s query. Structured data, taken from the Schema.org standard, helps a search engine to navigate code and understand its contents.
  • Experiment with new data formats. Google has just announced support for the Speakable structured data element. In Google’s words:

“In order to fulfil news queries with results people can count on, we collaborated on a new schema.org structured data specification called Speakable, for eligible publishers to mark up sections of a news article that are most relevant to be read aloud by the Google Assistant.”

The application of this format is limited for now, but it’s not difficult to imagine a future where digital assistants read content directly from all landing pages. Early movers will seize the advantage in this field.


Content Marketing

  • Create conversational content. Voice search lends itself naturally to dialogue, which should be factored into content strategy. Identify common questions or pain points in your industry and, quite simply, answer them better than anyone else does.
  • Write for intent states, not keywords. Voice search queries tend to be much more varied than their typed counterparts. As such, trying to target individual queries within content is a challenging and unnecessary approach. Search engines want to satisfy user intent. Aim to understand and respond to these states, helping people to achieve their task quickly and effectively. This will be more profitable than creating landing pages to target individual queries.
  • Develop a consistent brand voice. The future of voice search will involve brands speaking to their audience. This could be in the form of audio clips embedded in content or the search engine reading out text from the page. Either way, brands should be thinking of how they want their company to sound, rather than just look.

Furthermore, Google offers the following areas for assessment when it comes to this kind of voice search:

  • Information satisfaction: The content of the answer should meet the information needs of the user.
  • Length: When a displayed answer is too long, users can quickly scan it visually and locate the relevant information. For voice answers, that is not possible. It is much more important to ensure that we provide a helpful amount of information, hopefully not too much or too little.
  • Formulation: It is much easier to understand a badly formulated written answer than an ungrammatical spoken answer, so more care has to be placed in ensuring grammatical correctness.
  • Elocution: Spoken answers must have proper pronunciation and prosody. Improvements in text-to-speech generation, such as WaveNet and Tacotron 2, are quickly reducing the gap with human performance.


Local SEO

  • Ensure that names, addresses and phone numbers are accurate across all locations. The recently-announced partnership between Amazon and Yext points to one very clear voice search opportunity for brands, but it is dependent on brands supplying accurate and consistent location information.
  • Consider using a specialist platform to manage local listings and analyze your local search performance. There is a growing range of mobile SEO tools that can help with these tasks.
  • Make it easy for consumers to act on their intentions. This means adding in clear calls to action and directions to further information. Attention spans are at a premium for voice search, so make the most of what little time you do get.


SEO Strategy

  • Think beyond the website. Chatbots, apps and social media are all used to surface information for voice search queries. Therefore, you should optimize your presence across all of these in a consistent brand voice.
  • Use voice queries to plan future products and services. Within an app, it is possible to track and store all voice queries. This can be an invaluable resource when it comes to planning new services, as any unanswered queries will provide ideas with proven demand.
  • One real challenge with voice search is that it is not yet possible to segment queries within Search Query Reports or Search Console to see which were typed versus spoken. That will certainly come, but for now marketers should aim to extract maximum value from the limited data at our disposal.

Clark Boyd
Clark Boyd

Clark Boyd is a digital strategy consultant, author, and trainer. Over the last 12 years, he has devised and implemented international marketing strategies for brands including American Express, Adidas, and General Motors.

Today, Clark works with business schools at the University of Cambridge, Imperial College London, and Columbia University to design and deliver their executive-education courses on data analytics and digital marketing. 

Clark is a certified Google trainer and runs Google workshops across Europe and the Middle East. This year, he has delivered keynote speeches at leadership events in Latin America, Europe, and the US. You can find him on Twitter, LinkedIn, and Slideshare. He writes regularly on Medium and you can subscribe to his email newsletter, hi,tech.

Upgrade to Power Membership to continue your access to thousands of articles, toolkits, podcasts, lessons and much much more.
Become a Power Member

CPD points available

This content is eligible for CPD points. Please sign in if you wish to track this in your account.