Social Robots in Retail Banking

Problem

In the context of retail banking, the rise of online banking and omnichannel customer journeys have created significant challenges for customer relationship management. While social robots have been identified as potential tools to improve customer service, customer acceptance of these technologies remains a critical barrier. Most research focuses on consumer reactions after interacting with robots, but less attention has been given to the pre-interaction phase, where initial perceptions based on a robot's appearance and presence can significantly affect usage intention.

Objective

This study aims to investigate how the design of the initial interaction between customers and social robots (i.e., proactive vs. passive design) influences customer comfort, trust, and usage intention. Specifically, the research explores how these factors are interrelated and how the robot's design can be optimized to enhance customer acceptance. The comfort theory is used to explain how customer comfort influences their willingness to engage with robots, with trust serving as a mediating factor.

Methodology

A field experiment was conducted in a retail bank setting, using the Furhat social robot (modeled after bank founder). A total of 128 participants interacted with either:

• Proactive design: The robot initiates interaction when participants are detected.
• Passive design: The robot remains silent until participants initiate interaction.

Participants completed a standardized questionnaire assessing their comfort, trust, and usage intention, along with demographic and control variables (e.g., age, experience with social robots). Observers also evaluated participants’ emotional states (valence, arousal, dominance).

Key Findings

• Comfort: The proactive design led to higher psychological comfort and positive emotional reactions (e.g., smiling, positive comments). While physiological and holistic comfort also favored the proactive design, these differences were not statistically significant.
• Trust: The proactive design was slightly more trustworthy than the passive design.
• Usage Intention: Surprisingly, the passive design resulted in slightly higher usage intention, though this difference was not statistically significant.
• Emotional Reactions: Participants interacting with the proactive design showed higher emotional valence, indicating greater satisfaction and happiness.

Interpretation: Dual Processing Theory

The results can be explained by dual processing theory, which distinguishes between automatic and reflective human thought processes:

• The proactive design triggers automatic processes, relying on immediate positive impressions and generating higher comfort and emotional engagement.
• The passive design, on the other hand, encourages reflective decision-making, allowing users to consciously evaluate the robot, which builds trust and may explain the higher usage intention, despite lower comfort scores.

Managerial & Consumer Implications

• For Managers: A balanced approach that integrates both proactive and passive designs can improve customer acceptance of social robots by optimizing both comfort and trust.
• For Consumers: Proactive robots provide more engaging and intuitive experiences, while passive designs offer users greater autonomy, allowing them to interact at their own pace.
• Societal Impact: These findings suggest increasing societal acceptance of social robots in retail banking, with trust being a critical factor. This reinforces the need for interaction designs that cater to both emotional and reflective engagement.