Voice search is when you ask a computer a question using your voice – rather than typing. Think Star Trek’s ship’s computer or possibly HAL9000 from 2001: a space odyssey – but without the menace.
According to Sci-Fi books, films and TV shows we should’ve been jabbering away to robots and ‘AI’ devices for decades. But the reality is, voice controlled computers have taken a long time to become even vaguely useful – compared to manual inputs.
Recent advances like Apple’s Siri, Google Home, Amazon’s Alexa and Microsoft’s Cortana have brought voice recognition into our hands, our homes and even our vehicles.
Better technology and easier access are shifting how we interact, but what’s not clear is how search, as we currently know it, will have to change. Let’s take a look at what it can and can’t do and how you can prepare for the future.
What can voice search do?
One of the main features of voice search is the ability to ask complex questions in natural language (Who, What, When, Where, Why, How) and get an answer, or answers.
When we type a search query we tend to limit our language to 2-3 words – letting search engines fill in the gaps so we can get results quickly. But we’re not as brief when speaking. We initiate the conversation – the equivalent of clearing our throat or saying ‘Excuse me’ – we ask our question and wait for the response.
With voice recognition at the moment this is far from perfect – as the many, hilarious Siri fails show. Until it gets better, we’re not going to trust the answers and will limit our questions accordingly.
There’s also a notable difference in functionality between stand-alone voice devices opposed to voice assistants within visual devices – such as Siri or Cortana. Without a screen it’s much more difficult to provide options. This means search results are hugely limited – usually to single results.
What are the big changes from text search?
There are some text-based search behaviours which will have to be covered differently or discarded completely on voice search:
- Discovery searches
Questions beginning with what and why are far less common on voice searches than who, how, when and where. This shows an understanding that voice searches can only provide a definitive answer, such as ‘Who is Prime Minister of New Zealand?’ or ‘When is Father’s Day in the UK?’
It isn’t being used to discover highly considered purchases or narrow down options, such as ‘What’s the best home insurance to buy?’ – at least not yet.
What’s likely to happen will be recommendations based on what your device already knows about you – which is probably quite a lot.
Despite what tech companies claim, voice keywords are almost certainly being used to serve ads. It isn’t a giant leap to guess that voice search results are going to be equally tailored to your brand preferences, or implied brand preferences at least.
- Local searches
Businesses with physical locations could benefit from longer voice initiated search strings such as “Where’s the nearest sandwich shop?” or “Find a local emergency call-out plumber.”
- Opinion influenced searches
With voice search, we’re less likely to ask questions where we can evaluate answers and ratings. We might type ‘Best restaurant London’, expecting to see a variety of star ratings and reviews.
But we’d have to extend a voice search query with a trusted qualifier to even get us close: ‘What’s the best Italian restaurant in London, according to Time Out?’
Why is voice search important?
Although they’re really clunky now, voice interactions are becoming more and more common. Siri is actively used on more than half a billion devices and Amazon claim that over 20 million Alexa powered devices have been sold.
And, as Forrester shows (below) voice is set to become very sophisticated indeed in the next 3-5 years.
Uptake is set to increase too: ComScore says that by 2020 50% of all searches will be through voice. And the next generation of our workforce are much more comfortable speaking to their devices – 55% of 13-18 year olds in the US use voice every day.
Is it going to destroy your brand?
As all voice search results on stand alone devices are read out by the platform’s automated vocal chords – your carefully crafted logos, typography and brand personality will be translated or lost.
There are risks of your content blending in with everything else or your ‘cute brand language quirks’ sounding really creepy or forced. And forget about ‘show don’t tell’ – it’s tell or nothing.
How can you make sure you stand out?
Like search results pages now, it’s not about the clothes you’re wearing, it’s all about being at the party.
Your content needs to get to the point quickly, be simple and clear – easy to read out loud. Use short sentences and paragraphs and other formatting to break up text. Ordered lists at the top of a page responding to ‘How do I…?’ questions are a great example.
Now is a good time to hire, or work with, content designers (like our content consultants) to organise your content in the best way possible.
As Forrester points out, content will need to be answering the right questions quickly, efficiently and directly. Of course, that involves finding the right questions to answer before you can answer them. And that’s what we’re brilliant at.
There’s also a handy knock-on benefit – the curb-cut effect. This explains the serendipity of changing something for better accessibility and also making it better for many, many others, including GoogleBot.
Right now, voice search is really basic, an amusing gimmick or a way of turning music or the lights on or off. But with more sophisticated AI development and growing use it won’t stay that way for long.
Smart brands will start thinking about voice now, experimenting with how their own content appears and working out which questions they can answer.
Get in touch with our content consultants to help you prepare your brand for voice search