RESEARCH INSIGHTS: From expectations to satisfaction: A comparative study of customer interactions with AI chatbots vs human agents in e-Commerce settings

Kylie Decelis graduated with a Master of Science in Strategic Management and Digital Marketing from the Faculty of Economics, Management, and Accountancy at the University of Malta. She is now working as a Social Media Executive at Steves&Co.

In the fast-paced world of e-commerce, businesses are increasingly turning to AI chatbots to streamline customer service. While chatbots offer efficiency and cost savings, not all customers welcome automated interactions. A recent Forbes article highlighted that 86% of customers prefer human interaction over AI chatbots, with 40% experiencing frustration when chatbots misunderstand their queries. This raises a crucial question for businesses: When do customers prefer AI chatbots, and when do they seek human agents?

My research delves into this dynamic by exploring customer expectations and satisfaction in interactions with AI chatbots and human live-chat agents. Specifically, my study examines when customers prefer automated interactions and when they seek human support, providing valuable insights for optimising customer service strategies.

Customers expect speed from AI and empathy from humans

One of the most important things I found was that customers have distinct expectations based on the type of service interaction. For example, when engaging with AI chatbots, they prioritise speed, efficiency and accuracy. Chatbots excel at handling routine tasks such as order tracking and FAQs. However, when queries require personalisation or deeper understanding, customers quickly become frustrated with chatbots’ limitations.

Conversely, customers interacting with human agents expect empathy, adaptability and problem-solving skills. Human agents are particularly valued for their ability to manage emotionally-charged or complex inquiries. While response times may be slower, customers are generally more satisfied when their issues require nuanced understanding.

Customer satisfaction hinges on expectation confirmation

Satisfaction arguably occurs when expectations are met or exceeded. Customers who expected quick, accurate responses from chatbots and received them reported high satisfaction. However, negative experiences arose when chatbots failed to provide relevant answers or misunderstood requests, leading to frustration and a switch to human agents.

For human agents, satisfaction was found to be generally higher, as these interactions often exceeded customer expectations. When agents demonstrated empathy and flexibility, customers felt valued, increasing trust and loyalty. While occasional slow response times were noted, customers were more forgiving when human interactions resulted in effective solutions.

Implement a hybrid approach

What does this mean for businesses? First, given that AI chatbots and human agents each have strengths and limitations, businesses should adopt a hybrid model. Chatbots can efficiently handle repetitive, low-complexity tasks, allowing human agents to focus on high-value, emotionally-nuanced interactions. This division optimises resources and enhances overall customer satisfaction.

Improve chatbot intelligence and hand-off strategies

Secondly, to reduce frustration, chatbots must be designed with better contextual understanding and natural language processing capabilities. Additionally, seamless escalation to human agents should be built into chatbot systems. If a chatbot cannot resolve a query within a few interactions, a smooth hand-off to a human should occur to prevent customer dissatisfaction.

Personalisation and empathy matter

Thirdly, while AI continues to advance, human agents remain essential for delivering personalised and emotionally intelligent customer experiences. Businesses should invest in training human agents to enhance their problem-solving skills and empathy, ensuring they exceed customer expectations in complex interactions.

AI chatbots continue to revolutionise customer service by offering quick and efficient solutions for routine inquiries. However, as my research highlights, customers still seek human interaction for complex or emotional concerns. By implementing a balanced approach that leverages both AI and human expertise, businesses can enhance customer satisfaction, build loyalty and improve overall service efficiency. The key to success lies in understanding customer expectations and ensuring that both AI chatbots and human agents meet and exceed them in their respective roles.

This article is a summary of the student’s dissertation submitted in partial fulfilment of a Master of Science degree in Strategic Management & Digital Marketing. The article is not officially endorsed by the University of Malta. The opinions expressed therein are solely those of the respective alumni and may not reflect those of the University of Malta.