As society develops a greater level of trust with home assistants, the rise in popularity of voice search is inevitable.
This introduction of machine learning into our homes has certainly made life easier for us, allowing us to bake a cake with ease, get the news on demand or play our favourite song in an instant. But what other implications will arise from this change in our digital landscape? Before we dive into the nitty gritty of how voice search affects our work and lives as digital marketers, let’s discuss how we got here in the first place.
Google Home Mini
A Changing Interface
It all began back in 2001, thanks to the introduction of search Image Results and Google Ads (formerly Adwords). The classic 10 blue links that we knew all too well began to disappear. They were replaced with much more sophisticated and comprehensible methods of search. Since then, Google has been updating and optimising their products to create the search empire that exists today.
Left: Google SERP 2001 | Right: Google SERP 2019
Through the many evolutions of the Google search interface, we as digital marketers have played by Google’s rules. From algorithm updates, to new SERP features and platforms, we’ve lived through it all.
And for the most part, rank-devoted SEOs and digital marketers were happy to coexist with Google’s products. But now it seems the Google Knowledge Graph is showing up more and more, bulldozing our organic search results. As digital marketers and SEOs we were already competing with the rise in paid ads.
Fast forward to 2019, and it seems ranking organically is becoming increasingly difficult as the SERP real estate continues to diminish. So with the inevitable adoption of home assistants in the consumer market, and competing aggressors in SERP real estate, we need to adapt our content and SEO strategies to give our brands the best chance during this search renaissance. But first let’s grasp a greater understanding of voice search.
From Text To Talk
Paralleled to the evolution of the Google interface, was the evolution of voice search. Thanks to cloud-based computing, in 2001 Google launched their application “Google Voice Search” for mobile. This app utilised the best in cloud-computing and technology of the time, to match user queries with human speech.
But back in the early 2000s, to perform queries we had adapted to the search engine’s robotic manner of speech by using truncated requests. Instead of typing “where abouts in the world is the Eiffel Tower?” we would type “eiffel tower location”. So this clash in natural speech and search engine understanding led to obvious growing pains and obstacles to voice search adoption for the general public. Despite the lack of adoption, this launch was able to establish a foundation for machine learning, bringing us the sophisticated voice search experience we have today.
Google Voice Assistant
The 2013 Hummingbird update played a major role in helping Google to understand the intent behind search queries and the natural language patterns of humans. This has slowly enabled Google to move from keyword matching to a more conversational manner when returning search results.
As a result of this evolution in natural language processing and machine learning, in 2019, our search engines can return voice search results for much more complicated queries. This is obvious when observing children, who didn’t grow up with Google’s primitive truncated search queries, that our search engines are becoming increasingly conversational.
Google’s current intelligent voice search model and adapting interface is in correspondence with their desire to further understand consumer intent as opposed to just processing the search query typed into the search box. Voice search is a major player in this strategy. So how do we as digital marketers adapt?
Structured Data & Schema Markup:
At this stage, it is difficult to predict what practices digital marketers should be implementing to continue serving their results in the wake of voice search queries and home assistants. But we do know that structured data and schema markup is a great way to help Google understand the content in your website and better serve their customer intent driven strategy.
Structured markup data, if you don’t already know, is essentially a piece of HTML code which helps Google to provide rich snippets in their search results. Rich snippets are proven to increase CTR and drive traffic, therefore giving you a greater opportunity to show for voice search. You can browse what structured data and schema markup may be appropriate for your website in Google’s Search Gallery.
Google Search Gallery
Featured snippets are now appearing in over 30% of all Google search queries. This is important to note because this is generally where voice assistants locate their answers for performed searches. There are five basic kinds of featured snippets – paragraph, list, tables, youtube and multifaceted featured snippets – and it is important to understand each of these. The best way to optimise your content for featured snippets is by providing high quality and useful data for searchers, and adapt the format to your desired featured snippet.
Serve the right content:
Create content that users want to see. For example, the word “how” appears in more searches than any other word, and is closely followed by “what”, “why” and “does”. These words are essential in composing questions, therefore we can infer that our content needs to be answering these questions that our users are asking. Another way to increase your opportunity to rank is by targeting long tail keywords which are akin to natural language. Being the answer or being the action for user intent will be one of the most important factors in these formative years of voice assistants.
It’s All About Intent
Digital marketing is truly on the cusp of some large-scale revolutionary developments, and voice search is one of the major players. Remember Sergi Brin’s idea about the future of Google? That one day there will be no queries, Google will just know what you want and provide it to you there and then. A single interface for the web with all of the content coming into one place to serve you the data that you need without queries.
It seems this intent-driven strategy is serving this exact prediction, and that voice search coupled with machine learning is an active player in making this a reality.