Why Customers Still Prefer Human Agents Over Chatbots for Support

Chatbots Are Growing, But Customers Still Prefer Human Agents

Customers reach out to companies daily to buy products, ask about orders, make payments, or request returns. Traditionally, this happened via phone calls or by chatting with human agents on websites or apps. With the rise of AI, chatbots—software that can hold two-way conversations—have become a popular tool. Companies are routing more customers to chatbots, which has led the global chatbot market to grow from $370 million in 2017 to around $2.2 billion in 2024.

Despite this growth, customers remain hesitant. A recent survey revealed that 71% of customers prefer talking to a human agent over a chatbot. Additionally, 60% say chatbots often don’t understand their issues. This reflects a general lack of trust in AI when it comes to customer support.

Current Practices and Customer Preferences

Most companies use chatbots as the first point of contact. If the chatbot can’t resolve the problem or the customer requests it, the case is passed to a human agent. While this approach is efficient, it doesn’t always match what customers want. Some situations call for a chatbot, but others clearly benefit from a human touch. For instance, many customers are comfortable sharing purchase history for marketing but are far less willing to share sensitive financial information with chatbots.

What New Research Shows

To get a clearer picture, a study analyzed over half a million customer service chats from a major North American retailer. Machine learning was used to examine these interactions in three ways.

• Why customers reach out: Most inquiries fell into six categories—orders, coupons, products, shipping, account issues, and payments. Customers rarely used chatbots for shipping or payment questions, likely because these involve more detailed or sensitive info.
• Language matching: Human agents used language that more closely mirrored customers than chatbots did. This was surprising given the advances in AI. It shows humans can better adapt to the way customers express themselves.
• Engagement levels: When human agents matched the customer’s language style more closely, customers responded faster and more often. Feeling understood increased their engagement.

Practical Recommendations for Customer Support Teams

• Match the channel to the inquiry: Identify why a customer is reaching out before deciding if a chatbot or human agent should handle it. Some issues need human sensitivity, others can be automated.
• Train for language adaptation: Both chatbots and human agents should adjust their communication style to match the customer’s. For humans, this often happens naturally. For chatbots, it requires programming. Customers stay more engaged when they feel understood.
• Demand proof of chatbot effectiveness: Ask technology providers for data comparing chatbots against human agents on efficiency and customer satisfaction. Only adopt chatbots widely if they meet or exceed human performance.

Customers want to feel heard and supported. For now, that usually means talking to a real person. Instead of replacing human agents, chatbots should be part of a hybrid approach that respects customer preferences and assigns the right resource for each task.

For customer support professionals interested in improving chatbot integration or learning more about AI tools, resources like Complete AI Training offer practical courses and insights.