What is Voice Engine Optimisation?
Voice Engine Optimisation (VEO) are efforts to organically boost the position of content in voice-based search engine results.
Apple paved the way for voice search way back in 2011 with Siri – the first voice-enabled virtual assistant. Siri did three things:
Although voice-search technology has had its fair share of teething problems, today, Apple‘s Siri has 41.4 million monthly unique users in the U.S alone. Google estimates that by 2020, more than half of searches will be prompted by voice search. This increase can be attributed to the conveniences that voice search offers, as well as the speech recognition accuracy of voice-based search engines such as Google – currently at 95%.
The stigma for talking to our phones is also dissolving, encouraging voice search to be more commonplace – especially in the home. The more commonplace it becomes, the more brands and retailers will be keen to capitalise on it.
Since Siri’s debut, Google Voice Search, Amazon’s Echo, Windows Cortana and Baidu have joined the race. These giants are developing this technology to maturity. For instance, Baidu recently released its third version of DeepVoice – a system built on deep learning to convert text-to-speech. This version can learn 2,500 voices with just half an hour of data. Baidu also quote that their voice recognition error rate is at just 2%. If true, this means it’s better at detecting speech than a human, who has an error rate of 5%.
The availability of such sophisticated AI has already seen an increase of partnerships between voice-search based and emerging technologies. And as the prevalence of interfaces powered by voice and gestures continues to grow, so will the number of brands looking to get in on the action.
Humans were never designed to read or write. And believe or it not, we were never designed to sit behind a screen. Even though writing has become part and parcel with the everyday, it is still greatly overshadowed by speech.
We can type 40 words in a minute but speak up to 150. We speak with more than 50 main “accent types” in the UK but write with only a handful. We embellish our speech with colloquialisms, adjectives and profanities, but turn to shorthand for written.
Voice-search systems recognise this and are progressively engineering a more human interaction with technology. This is as much to do with changing our behaviour towards technology so it is more speech-led, as it is the personality brands choose for their AI systems.
Amazon is one company leading the way. Its Speech Synthesis Markup Language has recently equipped Alexa with a new set of speaking skills meaning ‘she’ is now capable of whispering, taking a breath and adjusting the rate, pitch and volume of her speech. With the ability to mimic human voice patterns in this way, she is sounding increasingly more human. This is a huge step forward for voice-enabled technology, especially considering how much we rely on voice for communication, emotion and knowledge – and how sceptical some remain towards the twitchy and robotic qualities of less-sophisticated AI.
With 40% of adults reported to be using voice search once per day, the future looks bright for manufacturers. And since Amazon and Google have been leading this field from the go, voice-commerce and voice-search continue to underpin the systems currently available to Joe Public. However, there is scope for voice-enabled technologies to be adopted much further afield.
For brands, the onus is on the retailer community to develop voice as a channel. Although convenience underpins voice-enabled technologies previous successes, focusing solely on this won’t be enough going forward.
To create those breakthrough use cases – to the magnitude of Pokémon Go for augmented reality – brands need to get more creative. Yes, the opportunity to take advantage of the simple things is there, but investing in AI technology that offers value where humans cannot is what will really drive this channel forward. Voice-enabled technology is the impetus needed to create and deliver functional, fun and personable experiences. But it isn’t simply a case of ‘build it and they will come’.
For consumers to engage with a particular voice-skill or voice-app, they have to know it exists. Brands looking to rise above the noise need to market and promote their voice-enabled technologies. Once established, it is the personality and intelligence added to an otherwise mundane process that will give consumers a reason to return.
Now we’ve established that voice search should be on your radar, here are a few items to consider when optimising your website for voice search:
Whereas normal search is done by typing keywords into a search engine, voice search is a chattier thing altogether. This means your content needs to be more conversational too.
Because of this change of tone, search engines are also changing how search queries are processed. Whereas Google used to source websites optimised around keyword(s), since 2013 attention has shifted to the context of the query instead. This means that when you’re searching for “the closest place to buy a cupcake?”, the results understand you’re after a local cupcake store, rather than a generic online retailer who’s nifty with SEO.
Tools such as Quora and Answer The Public (example below) can give insights into trends and identify variants of potential questions, whilst Google’s Search Console can help you see how people find your site and any signs of voice in their searches.
The window of opportunity for voice is far narrower than it is for desktop search. Although brands might be happy to feature on the first few pages of search results, people are unlikely to listen to more than the first search results when delivered by Alexa or Google Home.
Google have introduced featured snippets in response to this. Featured snippets appear at the top of Google’s search results. They are programmatically extracted from sites considered to contain a likely answer to the user’s question, giving brands the opportunity to leapfrog the competition to position zero – even if you organically rank lower on search results. Plus, the more snippets, the more traffic to your site!
Since most voice-searches are locally driven, make sure your business is listed and verified in top tier local directories. Make sure you’re also referring to local landmarks on your site as these have been shown to feature far more frequently in voice search queries. If people ask for a ‘coffee shop near x’ and you happen to be right next door to ‘x’, be sure Google knows it.
Since most voice searches will be performed using mobile devices, mobile optimisation is crucial. Enter your URL into Google Search Console to see if your site hits the mark. Luckily, ours passed first time!
Whilst you’re here, did you catch our podcast on the rise of voice technology? It’s well worth a listen.