How Changes to Voice Search Will Impact Your Business

New technology is exhausting. Just a few years ago, voice search radically changed how companies conduct business online. And mom and pop shops are just starting to catch up to trends like content clusters and conversational keyphrase prioritization. Well, the rules to deliver the best result changed again. Google Assistant and others like it transform the voice search game by introducing an advanced AI called the BERT Model. 

But don't be discouraged by new and confusing acronyms. If you are tired of always being a step behind in SEO trends, we have a solution. And it begins with shifting your focus. But before we explain what we mean, let's provide some context by plotting Google's development from personalized search, voice search, and the current iteration of Google Assistant. 



What is Personalized Search?

Google started working on personalized search in 2003, so it is certainly not a new idea. In the broadest definition, personalized search uses search and web history, location, device, and user profile data to deliver customized search results. A core part of achieving Google's mission has always been to better understand the differences and similarities between user behavior. In their most recent patent for personalized search, they state that, "In reality, a user like the random surfer never exists." "Every user has his own preferences when he submits a query to a search engine."

There is a strong desire to understand the personal biases influencing search behavior. And in 2011, Google was providing a wide diversity of user-specific results for the exact search phrase. But after strong criticism and a series of studies, Google decided to dial back much of its personalization. In a 2018 interview with CNBC, they stated,

Google started working on personalized search in 2003, so it is certainly not a new idea. In the broadest definition, personalized search uses search and web history, location, device, and user profile data to deliver customized search results. A core part of achieving Google's mission has always been to better understand the differences and similarities between user behavior. In their most recent patent for personalized search, they state that, "In reality, a user like the random surfer never exists." "Every user has his own preferences when he submits a query to a search engine."

There is a strong desire to understand the personal biases influencing search behavior. And in 2011, Google was providing a wide diversity of user-specific results for the exact search phrase. But after strong criticism and a series of studies, Google decided to dial back much of its personalization. In a 2018 interview with CNBC, they stated,

"A query a user comes with usually has so much context that the opportunity for personalization is just very limited," - Pandu Nayak.

This doesn't mean they aren't providing personalized results at all - and it certainly doesn't mean that they don't want to know the context behind the search. Instead, it means that personalized results are less about the individual and more about their intent. Understanding the evolution from voice search to Google Assistant will better explain this new objective. But before we explain, let's review the most significant factors influencing personalized search in 2021: 

Location

The most significant influence on personalized search results has and continues to be location. This is straightforward but critical for SEO. Sometimes making small changes that optimize location-based searches can make a massive impact on your business. For example, we recently added "Ventura County" to title tags on a garage conversion service page, and as a result, the rank jumped from 70 to 31. Of course, this isn't the only way to optimize for local search, but it is one example of the impact of a small change on a business. 

Google continues to prioritize location in personalized search because it always informs user intent. If someone searches for the "best coffee shop with a view," the city and town they are searching in will tell Google of the searcher's intended meaning behind the word "view."



Search History

Google still uses search history to personalize search results, but it no longer uses 180 days of browsing history. Instead, it only uses the previous search in its calculations. When Google was using that long list of browsing history, they attempted to use a personality profile. But as they stated, this only created confirmation bias and wasn't effective in delivering accurate results.

By narrowing the personalization to the previous search query, Google can make better inferences on the searcher's intent, not their personality. This shift of focus on searcher intent came in 2019 with the BERT model. But to better understand this evolution, we need to explain how voice search revolutionized SEO.  

How Voice Search Changed SEO

Voice search changed SEO by introducing conversational language to search queries. This started with technology like Siri and Google Now, which at the beginning, gave users the ability to access devices using voice commands. As their popularity grew, so did search phrases. This created a challenge and opportunity. The challenge was in shifting the ranking algorithms to cater to this new user experience. Search engines had to learn to answer particular questions instead of matching results to a robotic-sounding series of keywords. This paradigm shift happened somewhere between 2015-2016. 

Google led the charge by shifting its prioritization from keyword matching - to content cluster mapping. Instead of primarily relying on keyword density and page authority to determine the best match, they relied more on analyzing the number of related questions answered within a topic cluster. Our most recent article on content cluster strategy for new businesses explains this in greater detail.  

The added data presented by voice search created an opportunity. By asking questions in long-form sentences, Google discovered that it could use that data to better understand the query's intent. But it had to adapt its machine learning model to capitalize on the data. This opportunity motivated Google to develop the BERT model and incorporate it into an advanced interactive AI assistant.

How Google Assistant With the Help of the BERT Model is Changing the Future of SEO

So if you just now caught up to the content clusterization approach to SEO strategy, prepare for a new paradigm shift. In 2018, Google developed the BERT model (an open-sourced system) to understand and analyze search words in relation to each other. So instead of interpreting a sentence in a linear sequence, this system can intuit the context and deeper intent of the entire query. 

For example, Google's algorithm could not use prepositions like "for" or "to" to correctly interpret a query's intent before this change. If someone searched, "2019 brazil traveler to the USA need a visa," the algorithm would not be able to understand if the searcher was an American requesting a visa to Brazil or the other way around. Today, with the BERT model, search engines can intuit this meaning. 

This breakthrough created a massive opportunity for Google Assistant. Instead of just serving as a voice command function, users now have entire conversations with their AI-powered assistants. And keep in mind, the BERT model isn't a static system. The more users engage with follow-up questions to their original query, the better the machine learning algorithm gets at intuiting the intent and context of a question. This has vast implications for the future of SEO. If Google is on a trajectory to better intuit the purpose of a question, it also means a potential future where it understands the meaning of the answer it delivers. 

What Do Changes to Google Assistant Mean for SERPs?

Well, these Google Assistant changes mean that more people are only engaging with the first three results of a query. And, the smarter Google gets at understanding the context of our questions, the more complex and diverse our Google Assistant questions will become. 

We know that early on, students were leading the charge in pushing voice search to its limits. A Google study found that 37% of teens used Google voice search for help with homework. This was before the upgrade to the BERT Model but is Google's most recent public data. The teens in 2014 were already using Google to ask complex questions about historical dates and scientific theories. And as a result, these now young adults are very likely taking advantage of Google's more advanced conversational AI.

There are two essential SEO practices to implement in response to these trends. The first is using structured data. Google has adopted a universal coding language called Schema Microdata that helps crawlers better understand the intent of a page's content. It is used to tell search engines basic information like who the author of a web page is or the location of a business. The data used in these schemas are often used in feature snippets that appear at the top of a search results page. 

Secondly, content marketers should anticipate the follow-up questions many searchers will ask. Like with clustering content, including follow-up questions on a single post will demonstrate that you are a subject expert. But more importantly, it will help Google Assistant provide a better user experience by delivering immediate answers to follow-up questions.

Moving Forward: A Shift in Focus

It may appear that Google has been continuing to shift its goal post over the past couple of decades, but a deeper analysis reveals a consistent throughline. Google has wanted to better understand humanity to answer all of our big and small questions from the beginning. Its retraction of personalized search results was not a reversal of this objective. Instead, it was an insightful and hopeful revelation. The key to understanding humanity doesn't lie in what makes us different. Instead, in a rather beautiful fashion, Google shifted its focus to what unites us: Language. 

However, this revelation certainly doesn't mean that Google isn't collecting and analyzing our personal data. Instead, they are most likely using that to better understand universal trends and shifts in language and context.  So, what does that mean for SEO?

Well, since it appears that Google has shifted its focus to what unites us, we argue that contributors should do the same. We English speakers have a tendency to carry our hyper-individualism into content marketing. For example, our starting point tends to be, "what is the individual asking?" We then try to extrapolate that into, "what are the most common questions my target audience is asking?".

Don't get me wrong, these are both fundamental questions for a content marketer. But, starting there produces a far too common result: one that delivers an answer instead of the best solution. Instead, the starting point should always place emphasis on truth. We should be asking what the most accurate answer to my audience's question is. And as Google reaches the capacity to analyze the meaning of its results, your content and approach will not need to change. You will continue to lead the pack, and Google will reward you for it.

But of course, truth-telling is our passion. If you want a content marketing team that delivers compelling, fun, and insightful truth, we can help. Contact us today to get a marketing plan for your own passion project. 


Are you interested in learning more about voice search for your business? Let's explore the possibilities.

Locations

Dallas, TX | Beverly Hills, CA

Email Us

hello@rtbconsultants.com

Work with RTB


Copyright © 2020 | RTB Consultants