Page 17 - CBM28
P. 17

~ Oct8ne ~





            they can make modifications to the chatbot’s structure without   contributes to future conversions because the chatbot has
            requiring technical expertise or external support.   resolved their product-related doubts, making conversion more
                                                              likely.
            What
            Information                “AN EXISTING CUSTOMER HAS MANY                        EVO Banco in
            Was Considered                                                                   Numbers
            for Chatbot           COMMUNICATION CHANNELS, BUT DURING THE                     Over a three-month
            Configuration?       SIGN-UP PROCESS, BEING A CRITICAL PROCESS,                  period, EVO Banco
            EVO Banco                                                                        received 5,407 help
                                  WE ARE PLAYING WITH THE CONVERSION AND
            primarily considered                                                             requests, of which
                                WE BELIEVE THAT THIS TYPE OF FLOW IS WHERE
            two initial lines:                                                               99% were handled
            the product and                                                                  by the chatbot.
                                WE CAN BEST TAKE ADVANTAGE OF THIS TYPE OF
            problem-solving for                                                              This confirms EVO
                                                      SOLUTION.”
            account opening.                                                                 Banco’s objective
            They established                                                                 of providing
            that initial questions                ASIER DE LA TORRE - EVO BANCO              autonomous
            would be related to                                                              responses to
            the product itself                                                               potential new
            (intelligent account): its features, withdrawal options, ATMs,   customers, as only a small percentage was directed to human
            etc. As the user progresses, the dialogue tree shifts towards   agents. Additionally, 100% of the sessions were initiated
            solving operational and technical issues at each step of the   without a trigger, meaning users proactively sought help
            process. To create the content tree, they relied entirely on the   through the chat. ••
            historical data they had regarding customer inquiries.

            User Feedback on Oct8ne Bot Interaction
            “It’s good, and in the event that someone says it wasn’t
            helpful, we redirect them to a human agent,” confirms
            Asier. He explains, “There is a high percentage of those who
            interact with the chatbot to resolve product-related doubts
            who, after completing the dialogue tree, say it was effective.
            Approximately 75% of chatbot openings for product-related
            questions report understanding and finding it helpful.” This











































                                                            17
   12   13   14   15   16   17   18   19   20   21   22