In the early days of search, the only thing that mattered were keywords. Once black hat SEO tricks like keyword stuffing became the norm, search engines put an end to the deception and today not only filter out such results but also penalize the web properties engaging in such sinister practices.
Today, white hat SEO gurus and experts obsessively research, tweak and optimize their much-prized web possessions. Meticulously keeping up with Google algorithm updates has become a daily chore for many of them. And just when many SEO experts started feeling comfortable about their rankings and discoverability, the world turned upside down and started using voice to search for things. As a result, organizations and individuals suddenly found themselves clueless about how to approach this new world. If you’re one of them, read on for some life-saving tips that will help you avoid forfeiting your previous SEO efforts.
Conversational search facilitated by AI
What exactly brought on this seemingly sudden surge of voice searches? The main drivers of this trend are advances in machine learning and NLP (natural language processing). The second aspect responsible for revolutionizing search is consumer willingness to locate information using conversational interactions. The premise behind this is that voice is a more natural way for us to locate data and gain knowledge, than using a device and a browser to type words into a search bar. How much easier is it for us to simply ask devices for support by speaking, when and where it is most convenient for us?
Voice vs. conversational search
When it comes to voice-activated search, it is important to distinguish between using spoken language and traditional search engines like Google or Bing, and interacting with digital assistants in a conversational manner. These two types of interactions are inherently different and need to be approached as such.
Similar to the way we have been looking for information to-date, simple voice search is still using engines to locate content and produce the desired query results. This mechanism isn’t quite as game-changing because it is still based on keywords, albeit long-tail phrases which are often formulated as questions. Ultimately, spoken queries are converted to text again. In order to master this type of search, companies need to develop content that predicts and answers customers’ most frequently asked questions, in addition to including the words and phrases they would use to ask them in a conversational manner, e.g. ‘how to change a flat tyre.’
Conversational voice queries
Unlike keyword-based voice search, this type of queries is based on natural, spoken language, though not exclusively. In the real world users can switch from using voice to typing, sliding or taping and need to be able to do so seamlessly. Conversational queries are enabled by digital assistants like Google Assistant, Cortana or Siri, which are available on a variety of smart devices, including phones, watches, and speakers, to name a few. This shift in the way people search presents companies with a unique way to create personalized experiences for their customers.
Designing unique user experiences
A lot can be deducted about someone by the way they speak and ask questions. Making sense of these patterns and responding intelligently requires the implementation of sophisticated algorithms, collectively known as ‘machine learning. In addition to being accurate, responses also need to be conversational and emotional enough to feel ‘real’ and not like there’s a robot on the other end. As a result of these interactions, the user should be presented with options to proceed further, e.g. look for a similar product or service, locate the closest retailer offering it, or buy online. For these interactional funnels to work, real-time exchanges need to be designed for humans, similar to how chatbots are built.
Promises of voice search
One of the most notable advantage of voice vs. textual search is its accessibility. It is said that voice-activated search is going to revolutionize the world by putting vital information at the fingertips of the most underprivileged segments of society – the financially disadvantaged, the aging, the uneducated – and in many parts of the world – women. Whereas searching ‘the old fashioned’ way requires some knowledge about how to use hardware and software, the ‘new way’ simply means speaking into a basic device, such as a smart watch or a smart speaker, as you would to a friend or family member. While the former has at times proven to be intimidating or inaccessible to marginalized populations, voice-enabled search promises to provide easier access to the heaps of collective knowledge that is the internet.
Overcoming bias and other obstacles
Digital assistants still famously lag behind in recognizing languages other than English and accents that even slightly deviate from the ‘standard’. They are notorious for being partial to the intonation and accents of white, native-English-language male speakers. This certainly isn’t as much the product of intentional bias as it is the result of having limited types of big data to use for learning. As voice search picks up and is used worldwide by a variety of nationalities and genders, the assistants’ recognition ability is bound to improve and diversify.
Naturally, user experiences need to be translated and localized for voice recognition functionality to work as designed, as has been the case with websites and mobile apps to-date. Although the optimal way for brands to engage with users via voice isn’t yet well defined, the conditions are ripe for experimentation – as with most future-minded tech, early adopters are bound to develop an advantage.
Copywriter: Ina Danova