The use of smart speakers and voice search technology has grown rapidly in recent years. According to VoiceLabs and Mediapost, the number of voice-first devices in circulation in 2017 reached 33 million. comScore, meanwhile, projects that 50% of all searches will be made by voice by 2020.
If you have already invested in optimising your website for search engine results, the growth of voice search means that it might be time to revisit that task. Search engine optimisation (SEO) for voice differs from that for typed queries. While typed searches contain the bare minimum in terms of keywords, voice searches tend to be longer and more conversational. In fact, voice searches are 30 times more likely then typed searches to be action queries, according to Google. And they should know.
This means that companies that are serious about their website climbing the rankings (and why wouldn’t you be?) need to pay urgent attention to the importance of voice SEO.
Another thing to bear in mind with voice search is that the results are delivered differently than they are for typed searches. A single answer is provided in response to voice searches, while those typing their searches can choose from a selection of results, making their own mind up as to which looks the most useful. For voice searches, the coveted position zero tends to form the basis of the response that is delivered.
As such, companies that want to engage with voice SEO need to think very carefully about their answer snippets in order to advance to position zero, if at all possible.
Voice SEO comes into play with website translation, just as traditional SEO does. Translating a website, including its keywords and phrases, does not mean it will be optimised for search engines in the target country. Instead, fresh keyword research is required before the translation takes place, to ensure that the appropriate words and phrases are incorporated into the translated copy.
Translation and localization services can help companies tremendously with this process. They can undertake the keyword research in the target language, as well as the translation itself. As such, companies looking to optimise their websites for voice search would do well to find the perfect translation company now!
While companies around the world are seeking to optimise their websites for search engines in multiple languages, those making smart speakers and voice assistant systems are also working on language and translation considerations.
As of the end of August 2018, for example, Google Assistant is able to function as a bilingual system. So far, users can set up any language pairing that includes English, German, French, Spanish, Italian and Japanese. Further languages are due to be added, and there is even talk of trilingual functionality in the not-too-distant future.
Creating a bilingual voice assistant is a surprisingly tricky process, according to Google VP Johan Schalkwyk and Google Speech engineer Lopez Moreno:
“Of course, this process requires a sophisticated architecture that comes with an increased processing cost and the possibility of introducing unnecessary latency.”
As the device has to first figure out which language is being used, as well as what was said, the key for success was to ensure that it could still deliver results almost as fast as a monolingual system. The results so far certainly seem impressive.
Unfortunately, not everyone is convinced by the linguistic abilities of voice assistants. Research from Newcastle, England – home of the famous Geordie accent – has found that Google Home, Alexa, Cortana and Siri all struggle with just a single language, let alone more than one, when regional accents come into play.
Newcastle’s Life Science Centre surveyed 536 people about their struggles with technology. 79% of those with regional accents reported that they had to alter the way they spoke in order to be understood by their voice assistant. Linda Conlon, chief executive of the Life Science Centre, comments:
“The same people who decades ago were frustrated as teens trying to get cinema listings from an automated telephone system are now having the same issues with their smartphones or smart speakers. The technology has moved forward, but the inclusivity to cater for regional accents has not.”
The ability to develop a voice assistant that can deliver superb results, regardless of the language that you speak to it in – and the regional accent that you have when speaking that language – is about more than just convincing people to choose an Amazon Echo device over a Google Home one. The potential of voice assistance is huge. From cars to ovens to central heating systems, pretty much every powered device has the potential to receive – and act upon – voice commands.
This means that those working on voice assistants are in a race to deliver multilingual support, without that support impacting on processing time. At present, Apple is racing ahead of both Amazon and Google in terms of the number of languages that it supports, covering Arabic, Cantonese, English, Hebrew, Japanese, Korean, Mandarin, Malay, Russian, Thai and Turkish. However, with such high stakes, the race is most certainly on, and we can expect to see the big players battling it out for market dominance over the coming months and years. It will be exciting to see who wins!