Digital-POV: Google Assistant is now Bilingual
Google Assistant now features multi language recognition making it even more useful for anyone living in a multilingual household. Users can set up two languages in the Google Home app and when they ask the Assistant a question in either language, it will automatically recognize and respond back in that language.
Details and Implications:
Currently the assistant can understand any pair of languages within English, German, French, Spanish, Italian and Japanese. The key factor behind the growth of voice search is convenience. Customers want to find what they are looking for as fast and efficiently as possible. Businesses have been preparing for this shift over the past few years but as Google Assistant now features multi-language recognition, this creates more challenges for brands.
Multilingual Disintermediation: When using a voice search device, users are given one result: the most relevant based on queries and context. Brands will either be the best, or not appear at all, making competition ruthless. The shift toward multilingual assistants will force businesses to be highly relevant for all languages. While this will create more opportunities to be visible, it adds another level of complexity for optimisation across different platforms and media.
Misspellings Vs Mispronunciation: Whilst machine learning allows devices to understand different written variations and spelling errors, deciphering spoken differences of words is a whole new challenge. Voice search devices, up until now, understand variations of pronunciation with a 90% success rate for a single language. Adding multiple languages to the mix is likely to multiply the number of errors. Advertisers will need to be cautious with the choice of brand keywords – maybe even to the brand itself (who has not had the argument about whether it is pronounced “Nyke” versus “Ny-kee”). The multilingual assistant will open a multitude of possibilities for pronunciation that brands will need to quickly adapt to in their messaging.
Country-Specific Targeting: SEO experts around the world fight to serve the right content, in the right country, in the right context for the right audience. With the use of local SEO tactics and technical implementations such as ‘Hreflang tags’ (which help decide what language version a user is shown) agencies can better guide search engines to understand user’s intent, location and context. The addition of voice will be a game changer. Publishers are currently required to display versions of the site in the official language(s) of each country. Now, they should consider creating optimised versions of these in highly relevant idioms widely spoken in specific locations (Spanish in the US, French in Canada, etc.). This adds more complexity for content creators and SEO specialists who will need to find a way to bypass country-specific boundaries to focus on language specificities, changing the way we currently look at on-site targeting overall.
Google intends on “expanding to more languages in the coming months” and also says it is working on trilingual support too. By allowing users to search in various languages, Google Assistant may also in time allow marketers to have access to a whole new subset of data: correlating language use with behaviours. For example, a French expatriate might use their mother tongue at home relaxing with their family and English at work with their colleagues or out with friends. Therefore, through machine learning, devices can start registering behavioural patterns linked to the use of specific languages to display even better targeted advertising. Adding this additional layer of data on top of traditional demographic and geographic targeting will offer a new lens on customer behavioural patterns.