April 16, 2020
Due to the advancement of speech recognition, voice search optimization SEO has exploded in popularity; in fact, a prevalent stat thrown around in the SEO community is that about 50% of all searches will come from voice by next year.
Now, whether you’re skeptical about that presumptuous prediction or not, as a savvy digital marketer, you cannot ignore these facts:
Many adaptive SEO’s are finally making conscious efforts to implement Voice Engine Optimization strategies (more on this later), but before you learn the best practices of VEO, you should understand how it works.
Voice search or conversational search is just what it sounds like. Instead of typing keywords into a search query to generate results, you are able to speak into a device.
The evolution and advancement in voice search technology is making it easier than ever to search on home assistants.
So what exactly are these home assistants? The major voice search platform includes:
To sum it up even more, here are what the biggest names in voice search rely on:
Backlinko, a reputable source for SEO analyzed 10,000 Google Home search results and discovered the following:
The average voice search result is around 29 words.
36.4% of results for voice search implement schema on its page.
The average voice search result from a Google Home was written at a ninth-grade reading comprehension level.
Google tends to rely on long-form content as the average word count of a page that was used as a voice search result is 2,312 words.
Check this out, 40.7% of the voice search answers coming from Google Home were from featured snippets.
Due to the complexities of human linguistics, producing accurate voice recognition results have been a difficult feat to accomplish.
This is because you can say the same word with a different pitch and cadence.
Compare that to a static input such as an image; images are treated as an array of numerical values which can be directly fed into a computational neural network–a computer can figure out what an image is because each pixel is assigned a numerical value, which represents a certain color on the RGB scale.
The speech recognition software takes the input of the original analog sound wave and assigns a value based on the wave’s height.
After recording the wave heights at equally-distanced points, the software is able to assign each point a numerical value, after which the data is fed through a recurrent neural network; this type of neural network is able to influence future predictions using its memory.
For example, if you say “hel”, the neural network can extrapolate that “lo” will be the next syllable, forming the word “hello”.
Thanks to the Nyquist–Shannon sampling theorem, the input of the soundwave can be immaculately reconstructed into digital form just from measuring the equally distanced recorded soundwave heights.
Once the sound wave is digitized and the automatic speech recognition system processes the data and turns it into text, the now-converted text is used as a search engine query.
One of the more prominent reasons why voice search is growing rapidly is due to the impeccable accuracy of transforming voice to text; Google, the most accurate of them all as of this writing, is able to recognize human voice with up to 95% accuracy–nearly that of a human.
The type of search engine depends on the device you use; for example, if you use Google Voice or Siri, you can expect to receive results based on Google’s algorithm. (Fun fact: Google pays billions of dollars to Apple just so they can be the default search engine for all of Apple’s devices).
If you use Amazon’s Alexa or Microsoft’s Cortana, you can expect to see results from Bing.
A common misconception of voice search optimization is optimizing solely for smart speakers.
According to this Bright Local study, 56% of individuals used voice search to find information about a local company in the last twelve months on a smartphone.
Compare that to only 18% of the individuals surveyed stating that they use a smart speaker for that purpose.
In addition, 46% of voice search users reportedly look for a local business using voice search on a daily basis.
Here’s another pertinent stat:
53% of individuals report that they use voice search when they are driving.
What do you think they’re searching for? Perhaps a local business?
From the same Bright Local Study, here are the typical next steps a user takes after using voice search to find a local business:
Clearly, “voice search” isn’t just a hype word anymore; if you don’t start implementing voice engine optimization strategies, you’ll fall behind in the ever-changing environment of SEO.
There are a few thorough case studies on voice search optimization and its impact, one of the more popular being Backlinko’s behemoth 10,000+ word article.
Below you will find a summary of his most notable findings after analyzing 10,000 Google Home results, along with some actionable advice from the Conduit SEO team.
The average page loads in 4.6 seconds for voice search queries.
Action: Head over to Google’s page speed insights tool to make sure that the page you are trying to rank loads at 4.6 seconds at the most.
If your page doesn’t pass this test, Google’s tool gives you some key opportunities to improve the page’s speed, some of which include eliminating render-blocking resources, minifying CSS, compressing large images, and reducing server response times.
70.4% of Google’s voice search results are secured with an SSL certificate.
Action: Migrate your client’s site to HTTPS. If done incorrectly, it can produce pernicious effects. If you are unsure how to do this, reach out to a qualified digital marketing expert here at Conduit.
The average voice search result is around 29 words.
Action: Don’t try to reinvent the wheel. Be sure to write short, concise answers that properly cover the query.
Although a small correlation, 36.4% of results for voice search implement schema on its page, compared to the 31.3% average for conventional searches.
Action: For competitive niches, anything that can be used to push your client in front of their competitors should be implemented.
Think about how Schema started: in 2011, three of the largest search engines at the time came together to create a semantic vocabulary for search engines to better understand the contents of a web page.
If these rival search engines came together to work on a joint project with a goal to better understand web pages, it’s a good idea for you to implement it.
You don’t need to be a programmer either; there are WordPress plugins such as Project Supremacy that make the creation and implementation of Schema Markup extremely user-friendly.
Domains considered higher in authority have a propensity to show up in voice search results far more frequently than lower authority sites.
According to his study, the mean Domain Rating (Ahref’s metric of measuring how powerful the backlink profile of a site is) of a result is 76.8, which is on a logarithmic scale of 100.
Action: This is one of the points at which the traditional SEO of the last twenty years comes into play. Building backlinks from high authority sites will continue to play a salient role in the foreseeable years.
There are many link building strategies, including guest posting, broken link building, and resource page outreach.
Many link building guides cover each of these in-depth; you will ultimately determine what makes the most sense for your client.
According to Backlinko’s findings, the average voice search result from a Google Home was at a ninth-grade reading comprehension level.
Action: Avoid industry jargon as much as possible, and make your content easily ingestible for your users.
In regard to voice search results, Google seems to be favoring long-form content; in fact, the average word count of a page that was used as a voice search result is 2,312 words.
Action: Cover the topic as in-depth as possible. You shouldn’t force words onto a page for the sake of reaching a certain word count, but you should be fully satisfying the search intent of the user, and write about topics in the article that both the user and Google would expect to see.
40.7% of the voice search answers coming from Google Home in their study were from featured snippets.
Action: You may ask, “how do I show up in featured snippets?” A common way the Conduit SEO team steals featured snippets is from reverse engineering what’s already being used.
For example, how is the content structured in the featured snippet for which you’re trying to optimize?
Are the answers listed in bullets points, are they being pulled from header tags, or something completely different? Follow the structure of what Google is favoring and write a more comprehensive piece of content.
A prevalent stat thrown around in the SEO community is that about 50% of all searches will come from voice by next year.
Yet, as we are witnessing the significant growth and popularity of users searching by voice, marketers are scrambling to figure out how to optimize their website and product pages for search.
So now that you know what voice search is and the ranking factors, let’s see how you can optimize your product pages for voice searches:
When optimizing content for voice search, be sure to focus on the 5W + H questions – who, what, where, why, when and how. Being able to answer a question will grow your webpage drastically.
Organize your product pages with bullet points and tables as this will allow voice search platforms to have a better understanding of your content. This also increases readability and makes it easy for readers to read and engage.
Add schema to your page as this helps search engines better understand your page.
This is incredibly important for voice search as it can allow Google to pick up important information when trying to answer a question.
“Speakable” – which is a schema markup currently in beta on schema.org is a great one to add to your page as this identifies important sections in an article or webpage that are best suited for audio playback.
Be sure to add terms that are related to your product in the product description. Google uses semantic analysis to understand what the query typed or spoken really means. LSIGraph is a great tool to help you find profitable semantically related keywords.
If you are not thinking about voice search optimization as a means of increasing your search visibility, don’t wait until it’s too late.
Voice search is growing and more users are adopting it and as you know, the digital marketing industry is never stagnant.
Take into account the voice search ranking factors and follow the steps above to get started on your voice search strategy.
Change is a ubiquitous component of SEO, so if you want to stay ahead of everyone else, it’s important to keep up on the latest trends and changes in the industry.
Even though many consider voice search to still be in its infant stages, staying ahead of the curve will allow you to stand out, as well as give you the tools and knowledge to implement strategies built to withstand the ups and downs that come with the industry.
In order to stay ahead of the curve and execute the best campaigns possible, our analysts are constantly staying up to date on industry trends by feeding our brains with articles, podcasts, etc.
Working with a white label SEO Partner like Conduit Digital gives you the ability to start optimizing your client’s websites for voice search results.
Voice search optimization or voice search SEO is the process of optimizing your website content and local SEO factors such as location and user generated (third party) reviews in order to increase visibility for voice search results.
Voice search optimization is vital for Local SEO.