My two-fingered typing has improved over the decades. Of course, it’s had to speed up and adapt to different hardware for the always-on business world we operate in. But back in the 90s when this baby boomer was first divested of his secretary (anyone remember those?) my keyboard skills were so abominable, that Vodafone IT offered to supply me with voice recognition software, so – in theory – I could just sit back and dictate my business missives into a headset, and the copy would miraculously appear on the screen. In fact, they wanted me to be the guinea pig for the company.
Well, I’m sorry to report that the trial lasted about 20 minutes. Even using my best received pronunciation, the output was such gobbledygook, we had to abandon the test and conclude that voice recognition still belonged to the realms of science fiction.
Nowadays however, I’m merrily chatting to my phone to send short emails and texts (complete with dictated punctuation), and voice search has become second nature and so convenient. The rapid advance in accuracy and usability is down to the combination of AI, machine learning and the cloud, combining to continually improve the user experience.
Mobile is the catalyst here, as usage and search volumes in particular overtake desktop, giving us easy access to information on the go. And voice search has hit critical mass with over 20% of total queries, and according to Google, heading for 50% share by the end of the decade. Indeed 75% of consumers say they search more than ever now they can simply use their voice.
But is voice search really easier and faster? Last month at the Google Think Conference in London, we witnessed the ultimate duel between the world’s fastest texter, Melissa Thompson, and the world’s fastest rapper, Ocean Wisdom (using his clearest tones). On the count of three they were asked to search for things like "the best restaurants in Kingston, Jamaica" with both phones displayed for us to see on the screen. Melissa texted at lightning speed whizzing triangles across her smartphone, and relying on auto-correct – which worked. But Ocean was always a second or so ahead as he triumphantly scrolled through his results page first.
Brands must be ‘heard’ to be the single answer the consumer is asking for
Google’s search results are completely geared towards mobile, with relevance, proximity and prominence as the top three factors in the search ranking, which in turn puts the emphasis firmly on location in local search.
But the AI assistant market is red hot with healthy competition from Amazon’s Alexa, Apple’s Siri and Microsoft’s Cortana - all powered by Bing search - competing with Google Now.
Over the last 15 years, brands have been busy adapting their sites and optimising for the best SEO and PPC results based on text queries. Now there’s a clear imperative to turn their attention to voice commands.
But what kind of questions do consumers ask? According to Google data, 20% are researching a product or service, with the next three most popular categories – answering a question, finding out about world events, and researching a location. And what’s unique to voice search – depending on the platform – is there can often only be one answer, not the page of listings we were used to. One reason, perhaps, why voice search on Google Home has yet to be monetised. After all, would we trust one single sponsored answer?
In just a few years, voice search will become the new normal for brands, and structuring data to be ‘heard’ in SEO will be the new battleground. That means AI personal assistants must be able to distinguish between the who, what and – crucially - where, such that your brand ranks top, in real time. And what’s more, the information must be accurate. As voice search becomes second nature to consumers, they will become ever more demanding. Today many might forgive incorrect information from a relatively new service, but very soon consumers will blame the business for inaccuracies, not the search engine.
Brands who are already up to speed and succeeding on mobile and location search are in a good position already. Optimising location, and being consistent across directories, maps and apps, such that search engines prioritise your brand, is key to success in the AI first world.
Larry Page’s perfect search engine
In my time in the digital industry I’ve been privileged to hear first hand from the founders of the biggest digital entities. Five years ago at the exclusive annual Google Zeitgeist conference, Larry Page answered a question with his vision of the perfect search engine: "It would understand whatever your need is. It would understand the world deeply, and give you back exactly what you need." The exacting nature of AI powered voice search is moving us ever closer to Page’s vision.
The question is, will your data infrastructure ensure that your brand gets a good hearing to be exactly what the consumer needs in that moment? Or will I eventually have to learn to type with three fingers?
Guy Phillipson is the chairman of iCrossing UK and the former chief executive of the Internet Advertising Bureau UK.