When Backlinko released its analysis of 10,000 Google Home voice queries — we were fascinated.
After all, we do spend a good chunk of time around here thinking, talking, and writing about voice search. Heck — I even filled my home with three different smart speakers to see which one worked best. And when it comes to the future of SEO, I believe that voice search will only continue to play a fundamental role in how we not only seek and consume information, but also, how Google ranks and surfaces it.Backlinko set out to answer when it conducted this study. Using what the team believed to be 11 potential ranking factors — things like PageSpeed, Schema markups, and the enabling of SSL — founder Brian Dean reported key takeaways on the most significantly factors of voice search SEO.
Now, our experts have weighed in on what we have found to be the most compelling of those findings.
Our Experts Weigh In: Key Findings on Voice Search Ranking Research
1. Page Speed
“PageSpeed appears to play a major role in voice search SEO. The average voice search result page loads in 4.6 seconds (52% faster than the average page).” – Backlinko study
What Our Experts Say: Victor Pan, Head of SEO
When it came to page speed, our SEO experts actually believe it might not play a major role in voice search optimization. “PageSpeed is probably not a [direct] factor in voice search,” says Pan. “But sites that are well-optimized for user experience tend to rank highly, and voice search results tend to come from top-ranking results.”
That said, it would be interesting to conduct further research to determine if these same faster-loading sites — the ones that receive better voice search rankings — also have higher search engine results page [SERP] rankings. “Drawing conclusions without proper experimental setup is something our industry tends to do,” admits Pan, “though I’m for faster sites, all the time.”
2. HTTPS and SSL
“HTTPS websites dominate Google’s voice search results. In fact, 70.4% of Google Home result pages are secured with HTTPS.” – Backlinko study
What Our Experts Say
Each night, Moz tracks the most significant SERP features and ranking factors in its MozCast Feature Graph — and, as of writing this post, HTTPS results (sites with enabled SSL, or the Secure Sockets Layer technology that encrypts your website connection so hackers can’t intercept your data) accounted for 79.8%.
In other words, Pan explains, Moz tracks the SERP for 10,000 keywords, nightly — and of all the sites ranking for those 10,000 keywords, within the top 10 results, 79.8% of them are HTTPS.
“If we believe in Mozcast’s keyword sampling methodology, 70.4% of Google Home results pages being in HTTPS discovered during Backlinko’s research is actually on the low side,” says Pan, especially when you consider “not all of those queries tested by Moz necessarily lead to voice search results.”
So, when you compare Backlinko’s voice search HTTPS results figure (70.4%) with Moz’s (79.8%), it actually suggests HTTPS is not a ranking signal for voice search, Pan explains.
“If it was,” he says, “the number from Backlinko would be higher than the average reported by Moz.”
3. Answer Length
“Google prefers short, concise answers to voice search queries. The typical voice search result is only 29 words in length.” – Backlinko study
What Our Experts Say: Matthew Howells-Barby, Director of Acquisition
“This makes a lot of sense to me,” says Barby. “Google is likely taking a very similar approach to sourcing voice responses as it does to Featured Snippets.”
That was the same conclusion we drew when it was announced Google would soon be incorporating more than one Featured Snippet in the SERP for more complex queries.
And that display of more Featured Snippets in the SERP could indicate a higher barrier to earning clicks, especially if the snippets answer the query sufficiently enough the user doesn’t need to look — or click — any further.
Marketers, for their part, would benefit from preparing for that outcome. “Content creators should be adding voice-ready — and featured-snippet-ready — snippets into their content,” Barby explains, “to make it as easy as possible for Google to surface the answer they need.”
4. Schema Markups
“Schema may not play a key role in voice search rankings. 36.4% of voice search results come from pages that use Schema (which is only slightly higher than the worldwide average of 31.3%).” – Backlinko study
What Our Experts Say
Here’s another finding that doesn’t come as a shock to Barby.
“Similarly to the length of answers being pulled into voice responses, the lack of reliance on Schema markup doesn’t surprise me,” he says.
He points to a recent HubSpot study that found the same thing to be true for Featured Snippets.
“It seems like Google prefers to rely on its AI to surface the right result,” he says, “rather than the creators of content.”
5. Authoritative Domains
“Authoritative domains tend to produce voice search results significantly more than non-authoritative domains. In fact, the mean Ahrefs Domain Rating of a Google Home result is 76.8.” – Backlinko study
What Our Experts Say
For reference, the Ahrefs Domain Rating measures “the strength of a target website’s total backlink profile (in terms of its size and quality)” on a scale of 0-100.
That scale is what Ahrefs calls “logarithmic,” which means the room for growth between Domain Ratings of 77 and 78 is much larger than one between 27 and 28. So, the higher the Domain Rating, the more difficult it becomes to gain another point or two.
The reason behind this high Domain Rating of Google Home result pages is smart speakers like this one only provide one answer to your question, rather than the SERP’s list of several.
“This means that Google needs to be extremely confident that they’re giving you accurate information,” the Backlinko report explains. “And to help ensure accuracy, the voice search algorithm may rely on domain authority over page authority.”
But that doesn’t mean small businesses who are still building backlinks and authority should give up on optimizing for voice search, Pan says.
“As someone who worked on SEO for the Fortune 500 companies, these large companies simply had the budget and talent to spend millions in content creation,” he says. “I suspect that sheer quantity of content — which is helpful for users — leads others to link to these big brands they trust.”
That doesn’t mean those brands are authoritative in their domain, per se, says Pan — and that the wider-spread brand recognition might be behind this “advantage in winning voice search,” he explains, which “is what’s causing the average voice search result to have a high Ahrefs Domain Rating.”
6. Social Sharing
“Content with high levels of social engagement tends to perform well in voice search. In fact, the average voice search result has 1,199 Facebook shares and 44 Tweets.” – Backlinko study
What Our Experts Say
“I think that this is a classic case of simple correlation,” says Barby. “Google doesn’t use social shares to rank content. They’ll use social shares to improve crawling — just share a link on Twitter to see that — but not to determine authority.”
“The reality is that most content that gets large volumes of backlinks tend to also get large volumes of social shares,” he continues. “The backlinks will be the element that is moving the needle for ranking is search — not the social shares.”
7. Linguistic Simplicity
“Simple, easy-to-read content may help with voice search SEO. The average Google voice search result is written at a 9th grade level.” – Backlinko study
What Our Experts Say: Padraig O’Connor, Associate SEO Marketing Manager
When we prompt voice assistants to answer a question, we don’t exactly expect (or want) to be read a novel in response. That’s why Google Home, for example, often reads excerpts — or snippets — best summarizing the answer, then leaves a link to the full page in its corresponding app.
It would make sense, then, that Google Home selects the most concise answers that are easiest to read and hear, to allow for optimal listening comprehension. “Simple and easy-to-read content is ideal for voice search,” says O’Connor.
It goes back to the idea of growth within the idea of conversational marketing: the type that uses natural language to engage with users, even if there isn’t an actual human being responding to queries.
“Most people converse in a language that’s easy to understand,” O’Connor adds. “Google wants to do the same with the searcher at the other end of Google Home.”
To address that, he recommends marketers write copy accordingly. “Marketers should ensure that their content is written in a language similar to the way people converse and search.”
8. Title Optimization
“We found that very few voice search results had the exact 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.” – Backlinko study
What Our Experts Say
“At HubSpot, we’ve found that optimizing existing content for new Google SERP features is much more effective than creating a net new piece of content,” O’Connor says. “Our best advice to publishers is to tailor their existing content for voice search, rather than create brand-new articles for voice search alone.”
You might be familiar with HubSpot’s historical optimization project, in which the type of work O’Connor describes is done internally. But he cautions against optimizing for a single type of search. While voice queries might be rising in popularity and volume, many still use other methods of discovering information.
“Ultimately, you want to solve for all types of search,” he says, “be it desktop, mobile, or voice — and not pigeonhole yourself to one particular type.”
9. Parallels Between Desktop and Voice Search Results
“Content that ranks highly in desktop search is also very likely to appear as a voice search answer. In fact, approximately 75% of voice search results rank in the top 3 for that query.” – Backlinko study
What Our Experts Say
Aha! So, O’Connor was right. Even though we expect to see a continued increase in voice queries, desktop still matters.
It goes back to what I noted earlier about the simplicity of any query answer that comes in the form of an excerpt or snippet. That’s one of the things Google has used to determine how to float and rank results: the combination of quality and relevance, even if it means displaying more than one Featured Snippet in the SERP.
As for Barby, these findings don’t come as a surprise.
“Ultimately, Google is still going to be relying on its core ranking factors to determine which of the many relevant results it should surface as the voice result,” he says. “The big difference in voice search is that there’s no longer a ‘page 1 of Google’ — it’s one simple result or nothing.”
10. The Featured Snippet: Full Circle
“Appearing in a Featured Snippet may help you rank in voice search. 40.7% of all voice search answers came from a Featured Snippet.” – Backlinko study
What Our Experts Say: Brittany Chin, Growth Marketer
Backlinko’s discovery of a correlation between the Featured Snippet and voice search responses is somewhat similar to its findings on positive outcomes for brevity and listening comprehension. As I noted before: That’s why Google Home, for example, will often read an excerpt — or snippet — that best summarizes the answer, and then leaves a link to the full page in its corresponding app.
Chin’s research corresponds to this finding. “In a majority of cases, Featured Snippets are there to answer question-based queries in the SERP,” she says. “Google is trying to surface the best and most direct answer to the question right within the results page, which is the same thing it’s trying to do with a question-based voice query — but only get one shot at the answer.”
Chin also notes the goals of the Featured Snippets and voice queries are essentially one in the same: “to answer a question.”
“As for Marketers and SMBs,” she continues, “the strategy for both is fairly similar –continue creating helpful content that answers your audience’s questions in a simple and understandable way.”
Featured image credit: “Google Home device in front of closed blinds” by NDB Photos, used under CC BY / Cropped from original