How Conversation Transfer to Human Condition Works

Detection: This condition monitors whether a conversation with a chatbot reaches a point where transferring to a human agent is necessary or preferred.

Setting Conditions: Based on the detection of a need for human intervention, conditions are set within the chat flow to either continue with the automated system or transfer the conversation to a human agent. These conditions use a simple binary rule system of “is: yes” or “is: no” to determine the action to be taken.

Rules of Conversation Transfer to Human Condition

• “Is: Yes” Rule: This rule is applied when the condition for transferring to a human agent is met. If the chatbot detects that a transfer is needed—whether through direct user request, inability to resolve an issue, or detection of specific cues indicating that a human intervention would be more appropriate—the conversation is then routed to a human agent for further assistance.
• “Is: No” Rule: Conversely, this rule is used when the chatbot determines that a transfer to a human is not necessary. The conversation continues within the automated system, with the chatbot utilizing its programmed responses and flows to assist the user.

Implementation Example

An example scenario involves a customer support chatbot for an online retailer. When a user expresses dissatisfaction with an order, the chatbot initially attempts to address common issues using automated responses. If the user types “I want to speak with someone” or the chatbot fails to resolve the user’s issue after a few attempts, the chatbot assesses the situation as needing human intervention and applies the “is: yes” condition for transferring the conversation to a customer service representative.

Best Practices

• Clear Path for Escalation: Ensure that your chatbot design includes a clear and straightforward option for users to request speaking with a human at any point in the conversation.
• Smooth Transition: Make the transition from bot to human as seamless as possible, ideally providing the human agent with the context of the conversation up to the point of transfer to avoid requiring the user to repeat information.
• Monitor and Learn: Regularly review conversations that are transferred to human agents to identify patterns or common issues. Use these insights to improve the chatbot’s responses and reduce the need for human intervention over time.
• User Feedback: After the conversation has been transferred and resolved by a human agent, consider asking the user for feedback about their experience. This information can be invaluable in further refining both the automated and human-assisted parts of the service.