“Siri, find a good Japanese restaurant in Atlanta.”
“Cortana, how do I incorporate photos into a spreadsheet?”
“Alexa, give me a recipe for pancakes.”
“OK Google, can a giraffe breathe and swallow at the same time?”
For years, companies have focused their online content marketing on text-based searches. What do people type when they’re looking for a product or service like ours, they ask, and how do we capture those phrases in a way search engine algorithms notice?
As voice search tools like Amazon’s Alexa, Apple’s Siri, Microsoft’s Cortana and Google’s Voice Assistant have become more convenient and intuitive, however, customers use them more often — and they’ve changed the way they phrase queries to account for the voice-activated nature of the search tool.
Here, we look at how the move from text to voice search is changing our approach to content, and how to make changes that will capture both voice and text searchers.
How Voice Search is Changing Our Approach to Content
Voice-activated search tools are pervading our homes and workplaces. Studies estimate that by 2022, more than half of US households will have a voice-activated speaker device, according to Sarah Perez at TechCrunch.
But the biggest voice device expected to change the way we search is the smartphone. In mid-2016, Google CEO Sundar Pichai announced that 20 percent of mobile queries were voice-based.
The numbers have grown since, making voice search ever more commonplace and affecting which results rise to the top of those searches.
To complicate matters further for content teams, every device uses a different search engine or combination of search engines from which to draw knowledge, according to Britney Muller at the Content Marketing Institute. For instance, Google Home and Google Voice Assistant both use Google’s data, Siri and Cortana rely on Bing, and Alexa defaults to a combination of Bing and Amazon data.
There’s still no single strategy that will optimize content for every voice device. Focusing content on the way the use of voice changes both the query and the response, however, can help your content rise to the top of voice search results.
Make Keywords Conversational
Text-based searches tend to cut a query down to its most important words. Voice-based searches, by contrast, are structured much more similarly to the way we speak.
“If I were to approach an employee in a pet food store, I wouldn’t typically ask ‘best pet food dog?’” Angus Ewart writes at Small Business BC. Instead, the question would be phrased as “What is the best pet food for my dog?” or “What dog food do you recommend?”
When voice search first debuted, it had a much harder time with recognizing speech and matching it to relevant answers. Voice search’s word error rate was over 20 percent six years ago, but is eight percent or less today, AJ Agarwal writes at Forbes. This means that content marketing teams don’t have to account for misspoken or misinterpreted words anymore — but they do still need to account for the many ways that people might voice a particular question.
Crucial Context That Voice Searches Provide
The good news? Good keyword research has always been about putting a human reader first, says Julia McCoy at the Content Marketing Institute, and good keyword research for voice search optimization is no different.
Optimizing keywords for voice search starts by collecting keywords that surround a target word or phrase. To find the target, consider which questions your customers might ask that would lead them to your company. Don’t skip the question words: Uses of “who,” “what,” “when,” “where,” “why,” and “how” jump dramatically when a query moves from text to speech, says Neil Patel.
Voice search also demands optimization for long-tail keywords, Susannah Noel writes at Clariant Creative. Traditionally, keywords have been short, as the majority of text-based searches are only two to three words long, Purna Virji at Microsoft points out. Voice-based searches, however, are often longer; they’re less likely to be two words and more likely to be longer than four words.
These wordy questions provide valuable context about the searcher’s needs, interests and goals, Ian Kelley writes at Vital. That context is often missing from text-based searches. For that reason, much content aimed at capturing text searchers often skips that context, too.
For instance, Kelley notes, a text search for “best laptops for college students” doesn’t indicate on what basis the searcher wants to determine what’s “best.” Articles and blog posts written to target this phrase may choose their own “bests,” which may not match the customer’s interests.
By contrast, “What are the best-priced laptops for college students?” is a request for comparisons based on cost, which a voice search can more easily provide — and which content marketers can more thoughtfully target.
Don’t Reinvent the Wheel to Accommodate These Searches
Does this mean you need to write new content for every long-tail variation on every question that might bring voice searchers to your website? Not necessarily. Brian Dean at Backlinko analyzed 10,000 voice queries and responses to discover that very few voice results had the exact words of the query in their title tag. “Therefore, creating individual pages for each voice search query doesn’t appear to be an effective voice search SEO strategy,” Dean says.
Instead, adapting existing content to address some of the most common long-tail questions may be more effective than dashing off a new, unoriginal piece of work for every possible combination of words — especially since the average word count on a voice search results page was 2,312 words, according to Amanda Zantal-Wiener at HubSpot.
Turning existing content into value-packed longform content may do far more for voice search results than creating endless streams of new pages.
Fifteen percent of new Google voice searches every day are for combinations of words that haven’t been Googled before, according to Heather Horton at Ecrubox Digital. That’s because the use of spoken language changes the way we phrase queries — and the contexts in which we make them.
A Google case study revealed that most voice search queries share three common elements:
- They’re more likely than text queries to be about an action-based, on-the-go topic. Think users looking for store hours or directions to a specific location. In fact, one in three mobile searches are local in nature, according to DialogTech’s Christie Huber.
- They’re less likely than text queries to deal with sensitive information.
- They don’t include searches for sites or information that require user interaction, like an online catalog.
In addition to understanding what people ask in voice searches (and keywording accordingly), understanding how those questions are asked — on the go, impersonally, with a quick response in mind — can help companies structure their content to provide the response searchers are looking for, says Elizabeth Bush at the Enveritas Group.
When the “how” aspect of questions is understood, the “how” of answers can be improved. Dean’s Backlinko study found that short and simple was the way to go: On average, voice search results were 29 words long and were written at a 9th grade reading level.
Finally, as Tom Salvat notes at Concured, specificity can help companies rise to the top of voice search results — and the top is where companies need to be, since voice-based search engines typically read only the first result aloud to the user.
Prioritize Featured Snippets
Featured snippets are the parts of a result that Google Home and Google Assistant read aloud in response to a voice search query. In addition to reading the information aloud, Google’s tools also typically cite the name of the website that provided the result, making featured snippets a powerful tool for putting your brand in front of potential customers, Bryson Meunier writes at Search Engine Land.
Unsurprisingly, information on how to create solid featured snippets has gotten a great deal more attention in recent months. Tips for creating content that lands in the featured snippet spot from Julia McCoy, again, at the Content Marketing Institute include:
- Answer the questions readers are asking. “How does,” “how to” and “what is” queries commonly result in featured snippet responses.
- Make your answer short and to the point. Content that doesn’t answer the question won’t get the featured snippet.
- Create high-quality longform content. Featured snippets use the same ranking system as the rest of Google, so quality content is a must. Strive to offer the best answer available.
Finally, stay on top of Google’s news on featured snippets. The company published a guide to featured snippets in early 2018, authored by Danny Sullivan, Google’s public liaison for speech. This guide provides excellent insights from the search engine itself.
In addition to targeting Google’s featured snippet option, optimizing short pieces of content for other platforms, like Bing-based Siri and Alexa, is becoming easier. Schema’s “speakable” extension, set for rollout in late 2018, will allow website owners to indicate snippets of text that are particularly suited to being read aloud by voice assistants, giving content creators the ability to indicate to voice systems what they should read to users.
Why It’s Time to Befriend Voice Search in Content Marketing
Voice search use continues to grow as users discover its accessibility and speed mesh well with their needs. Adapting content to incorporate voice considerations can make it more closely match natural speech, read more engagingly and provide clearer, more concise answers — all of which puts that content, and your company, front and center of search results.
Images by: sifotography/©123RF Stock Photo, dragonimages/©123RF Stock Photo, Kaufdex
Latest posts by Epic Presence Team (see all)
- Navigating the Media Today: Can We Learn to Spot Deepfakes? - January 14, 2020
- From Uncanny Valley to Friendly Neighbor: These 5 AIs Are Surprisingly Lifelike - December 10, 2019
- Digital Truth: How Fact-Checkers Try to Wrangle Lies on Social Media - December 3, 2019