Client: Santander

Chatbot – user-friendly interface

Summary

During UX research we have measured users’ emotions while they interacted with our client’s chatbot. The research resulted in a list of recommendations for designing the content generated by the chatbot. The client received feedback on what part of conversation evoked the greatest emotions among users. We also defined which avatar was the best for the service and explained why we recommend it.

Sector

Banking

Our role

Qualitative Testing
User Testing
Chatbot Design
Usability Testing
GSR Testing

Challenge

How to program

a user-friendly

chatbot?

The main challenge was to explore users’ attitude towards the chatbot, and the emotions that appear in a conversation with the chatbot. To achieve that we have acquired feedback from users in the following areas: communication, chatbot interface usability, and the users’ attitudes towards the chatbot avatars.

Result

A set guidelines for redesigning chatbot generated content

We have provided the client a detailed report with a set of guidelines about redesigning chatbot generated content. We verified the interface usability, examined the users’ attitude towards the chatbot and checked the reception of messages. Furthermore, we used GSR (Galvanic Skin Response) that allowed us to designate the moment of conversation that stirred the most emotions among users. We have indicated which avatar is the most appropriate one and why.