I recently tried using an airline’s AI-powered chat assistant to book a flight ticket, primarily enticed by the 10% discount they offered. However, I left the process midway. Why? Every single piece of information had to be manually typed out in a long conversation with the smart chatbot. Simply providing details such as the source, destination, date of travel, and the number of passengers took me three times longer than just choosing these values from a drop-down menu. Also, the artificial dialogues generated by the bot felt less interactive and led to a poor user experience. I booked my flight directly on the airline’s website a few days later.

To give you another example, I attempted to modify a hotel booking using an AI-powered Interactive Voice Response (IVR). The IVR prompted me to verbally state my booking reference number. Ignoring the issue of information privacy (as someone nearby could overhear details about my whereabouts on a specific day), the experience was bad as the IVR would instruct me to press 1 if I said ‘1M23’ or press 2 if I said ‘1N33’. A simple task of typing in a booking reference number was complicated using an AI.

Crafting Experiences for a Wide Spectrum of Users

While I am sure there are quite a few examples where user experience can be made a lot better, the point is most businesses seem to overlook user experience while deploying Gen AI solutions. Even as chat interface makes things very simple for the end user, thought must be given to understand the user and exact scenario where the customer might be using the app. You won’t really want an irate customer whose flight just got cancelled to talk to an impersonal AI who can’t offer anything except information, leave aside empathetic listening.

In many new age user friendly Gen AI solutions, I have seen businesses implement emotional intelligence strategies to better plan for user emotions. Wherever possible, businesses should implement personalized responses tailored to individual users and provide users with an option to seamlessly transition to human support. Personalization is key for staying competitive in any industry and many users still prefer to talk to live human support for any query.

At Konverge AI, we usually give these 3 simple suggestions to our clients in order to build quality chatbots designed for success:

1. Empathy First: We champion “empathy studies,” a design thinking technique to deeply understand user situations. Imagine the frustrated flyer again. An empathetic chatbot wouldn’t just recite cancellation details, it might offer rebooking options, express understanding of their situation, or even connect them to a live agent for further assistance.
2. Click Minimization: Nobody enjoys endless clicking. Design your chatbot flow to minimize user inputs. Aim for clear, concise questions that get straight to the point.
3. Enhanced Voice Recognition: Voice recognition is fantastic, but only if it works! Focus on improving voice recognition, especially for capturing complex information like long sequences of numbers or providing non-voice-based input methods.

By prioritizing these strategies, you can surely build better chatbots that truly elevate user experience. Remember, happy users lead to successful AI implementation!