Asos is building a chatbot stylist that will pivot between sales and e-commerce depending on whether it’s being used on a messenger service or a voice assistant.

The online retailer has extended its Enki chatbot to the virtual Google Assistant 12 months after it debuted the service on Facebook Messenger.

Customers in the U.K. and U.S. can start a conversation with Enki using the phrase “Hey, Google, talk to Asos” from the Google Assistant app on Android or iOS mobile devices as well as the Google Home smart speaker. The chatbot will then direct the person to browse six of the top-selling menswear or womenswear looks on the Asos site depending on what product they want to view. More items could be covered by the chatbox depending on feedback from users.

The service is less than a week old, so Asos would not share any results on how much traffic the chatbot had driven to the site. But Jason Gregory, the senior product manager who led the project, did acknowledge attribution is currently not possible if the user’s voice searches are made via a mobile browser from which they make the purchase, as they might do using Enki.

“The same way we look at how social media channels like Instagram and Facebook drive people to our site will be the same way we judge how successful Enki performs on Google Assistants,” Gregory said. “Lead generation is one of the main metrics we’re looking at now.”

In that example, the sale would be attributed to the mobile device. Gregory, however, believes being able to attribute that sale to a voice assistant will eventually arrive. Experts have said the same.

“All Google Home users require a Google account to set up the device, so this would most likely be the same Google account you search on; therefore, linking the two journeys should be possible. Google already do this for mobile and desktop users,” said John Campbell, head of SEO at Roast.

How much traffic Enki generates for Asos will inform how far it has stretched beyond Google Assistants to other voice platforms. The aim, according to Gregory, is to create a chatbot that exists across all its main social networks and voice assistants in a way that leans more into prioritizing sales or customer service depending on the platform.

As part of its push into voice search, Asos is looking to understand the meaning behind queries rather than specific keywords. It’s a complex journey to map that depends on how the user voice searches.

“We’re trying to learn at the moment which parts of the Asos shopping experience work best on those platforms that exist outside of our own,” said Gregory. “From an online fashion perspective, voice is an opportunity to make shopping to feel more intuitive and hopefully help you find items faster in different ways to how they’d do it in e-commerce.”

Subscribe to the Digiday Retail BriefingA weekly email with news, analysis and research covering the modernization of retail and e-commerce.

  • LinkedIn Icon