“By 2023, over 50% of customer interactions at Orange will be handled via a chatbot.”
Who hasn’t come across a “chatbot” when making with a complaint or help request? They have become an effective first port of call before an advisor intervenes. And companies have capitalized on this value. They are willing to invest generously in this conversational artificial intelligence (AI) to improve its performance. In doing so, customers get an increasingly comprehensive experience, where humans and tech work hand in hand. Beyond “pop-up” text chatbots, the evolution of bots toward using voice and image is promising.
From text to voice and image
“Orange aims to handle more than 50% of customer interactions via a chatbot,” says Clément Aincy, Managing Consultant at Sofrecom. These chatbots play a key role in simplifying digital experiences and empowering customers. The focus is now firmly on this technology. Thanks to AI, 40% of customer service calls are set to be automated.
Newly mature technologies allow for customized voice experiences with “callbots.” Trained to have an optimal understanding of natural human language and a range of customer requests, these callbots can call customers to perform tasks such as invoice recovery. Finally, with the current boom in video calls, bots that combine computer vision and augmented reality are set to become an additional asset for simplifying the user experience. They will be useful for making it easier to install and troubleshoot Wi-Fi hubs by superimposing an illustration onto a real-time smartphone image of the hardware to show what needs to be connected where.
Achieving a new synergy between advisors and chatbots
In addition to these technological advances, bots are transforming the role of customer advisors. In the current setup, chatbots are the first line of communication in customer relations and take care of simple tasks, freeing up advisors so they can focus on solving more complex issues. Chatbots can route customer requests to the best-qualified agents and detect feelings of dissatisfaction in customers, at which point the chatbot transfers the customer to an advisor, as is the case at Orange Bank with IBM Watson technology. Some solutions, such as LivePerson, even enable the advisor to invite a chatbot into their conversation with a customer. The customer’s experience is smoother, faster and more efficient.
What about digital duos?
Training a chatbot still has to be left to the experts. However, major progress could be made via a two-way collaboration between chatbots and advisors. The latest developments in so-called “generative” chatbot AI are opening up new horizons. This generative AI feeds on unstructured data and can therefore be trained on huge databases. On its own, and without necessarily having been pre-trained, generative AI is capable of handling increasingly complex customer service interactions. In a trial run by Amazon, with the aim of training AI on 5 million responses extracted from 350,000 chats with customers, the advisor could select or modify a response suggested by the chatbot, without detracting from the overall understanding.
According to Clément Aincy, “this suggests that, although AI currently assists advisors, in future, advisors will support chatbots in their customer interactions and will even create their own digital doppelganger based on their behavior and communication styles, like a form of reverse personalization.”