Interested in voice search? Be sure to check out our recent white paper on how voice search is evolving to get the perspective of Wpromote VP of Earned & Owned Media Rachel Bucey.
Voice search burst into the mainstream in 2011 with the launch of Apple’s Siri on the iPhone 4. In 2015, the Amazon Echo became the first mainstream smart home device. From there, a number of competitors have entered the voice search market (more on them in a bit).
By now, you probably know that the growth of the audio market has translated into opportunities for your brand to connect with users, particularly for ecommerce, as 40% of smart speaker users shop on their device at least once a month.
But before we move onto why businesses should care about voice search, it’s important we all have a baseline understanding of how the technology even works.
What Is Voice Search?
Voice search is the action of using speech to ask questions and give commands to compatible devices like smartphones or smart speakers. The searcher speaks to the device and the device fulfills the command or answers the question.
For the purposes of this article, we’re going to focus on question-based search since that will be of most interest and value to businesses (as opposed to actions like setting timers or playing music).
How Does Voice Search Work?
Without getting too deep into the technical details of voice search, digital voice assistants work by:
- Processing and transcribing human speech into text
- Analyzing the text to detect questions and commands
- Connecting to external data sources such as search engines to find the relevant information
- Translating that information into a digestible format to fulfill the user’s intent
How voice search works is obviously more complex than this, and there is a lot of machine-learning going on so that these programs better understand natural language and true searcher intent. These programs are incredibly sophisticated and are improving tremendously.
Who Are The Major Players In The Voice Search Market?
The major players in the US smart speaker market are as follows (in order of market share):
- Amazon (63.2%)
- Google (31%)
- Other (5.8%)
However, it’s important to note that the majority of voice searches are actually occurring on mobile devices, not smart speakers. For smartphone assistants, the 2 major competitors are Apple (Siri) and Google.
Where Do Voice Assistants Get Their Information?
Voice assistants and smart speakers pull their information from multiple data sources including search engines, Wikipedia, and other major websites. For a business to optimize their content for voice search, it’s imperative to understand where the different voice assistants source their information.
Alexa sources its information from a number of different third-party sources including IMDB, AccuWeather, Yelp, Answers.com, and Wikipedia. If a user invokes a specific Alexa Skill, that data will be sourced from the database linked to that Skill.
Google, as you might expect, pulls in information from Google Search and their index of trillions of pages. Therefore, if your content is discoverable and crawlable on Google Search, you’re in good shape. If you are worried your content isn’t crawlable, tools like Google Search Console can help with diagnosing potential issues. Our experiments have shown that the majority of the time, the answer returned by Google Assistant matches the featured snippet a user would see if they typed in their question…but that isn’t always the case.
Siri also pulls its data from Google Search. It used to source from Bing but switched to Google in late 2017. However, for local businesses, it relies on Apple Maps data, which in turn uses Yelp for review information.
How Can You Optimize Your Website for Voice Search?
Our internal research has shown that successfully optimizing content for voice search comes down to 3 things:
- Ensure your content is indexed and present on all relevant major data sources (Google Search, Google Maps, Apple Maps, Wikipedia, etc.)
- Use traditional SEO best practices to rank your webpages and/or listings high in search engines. Your content can be perfectly structured for voice search, but if it’s not on Page 1, it’s unlikely to be pulled in as the answer.
- Have your content succinctly and simply answer the question. Our data has shown aiming for 30-40 word answers at a 6-7th-grade reading level works best.
We conducted this research by developing technology at Wpromote called Bixby, Fetch! that automates the asking of questions to voice search devices, transcribes the responses, and then analyzes the webpages from which the answer was supplied. This enables us to gather data on thousands of queries without needing to lock someone in a room for a week with a Google Home speaker. All we need is an initial question list to feed to Bixby, Fetch! and it does the rest.
In addition to the above, there are two types of structured data that were introduced in 2019 that are automatically voice-search friendly on Google devices. HowTo and FAQ schema automatically create an Action for Google Assistant (Actions are equivalent to Alexa Skills). You can find the full documentation and eligibility requirements for these schema types here and here. If you have this type of content on your site, we strongly recommend taking advantage of this markup.
Voice Search in 2020
Since 2015, experts have claimed that next year would be the year of voice search. While we’re not quite ready to say 2020 is the true year of voice search, there’s no doubt that the technology is increasing its utility and proliferation. For example, in the US, it’s expected that nearly one-third of the population will use a voice assistant at least once a month.
In our view, voice search must clear a few hurdles on its way to becoming even more of an acquisition and sales platform for businesses:
- A lack of detailed data around traffic and sales generated from voice assistants. Even with technology like Bixby, Fetch! we’re unable to definitively say how much traffic a voice query selected as the answer actually drives.
- The non-commercial intent of smart assistant users. A PWC study from 2018 showed that only 50% of smart assistant users had utilized the technology to buy or order products. However, 88% of users do use smart assistants to answer questions, so there is an opportunity for brands to increase their exposure by ranking well for voice queries.
Once these hurdles are cleared, the ROI of voice search optimization will become much clearer and stronger. In the meantime, forward-thinking brands can set themselves up for success by optimizing their content now and ranking for queries relevant to their business.
For more, see our recent white paper on how voice search is evolving to get the perspective of Wpromote VP of Earned & Owned Media Rachel Bucey.