The quality of a fallback tells you more about a chatbot than the quality of its best answers. Any tool can handle opening hours. What happens at the edge of its knowledge is where trust is either kept or lost, because the alternative to a good fallback is usually a confident invention that a customer then acts on.
A useful fallback does four things. It signals uncertainty plainly rather than burying it in hedging. It gives whatever partial answer it legitimately can, since most unanswerable questions are partly answerable. It offers a concrete next step rather than a vague invitation to get in touch. And it captures enough detail that the person picking the conversation up does not make the visitor start again.
Clerkzo is configured to say it is not sure rather than guess, and human handoff is available on every plan so the route to a person always exists. The conversations that end in a fallback are also worth reading, because a question the widget could not answer twice is usually a gap in your own content that a correction or a new page will close permanently.
Related terms
Browse all 142 terms- Speed to leadSpeed to lead is how quickly a business responds to a new enquiry, and it's one of the strongest predictors of whether that enquiry becomes a customer.
- First response timeFirst response time (FRT) is the elapsed time between a customer's first message and the first meaningful reply from the business. It is a standard support and sales metric because it measures how long someone is left waiting before they know a human or system has engaged with their request.
- Conversation FlowConversation flow is the path a chat takes from the opening message to a resolution, including the order questions are asked, the branches available and the points where the conversation can end or transfer to a human.
- Dialogue ManagementDialogue management is the part of a chatbot that decides what to do at each turn of a conversation, tracking what has been said, what information is still missing and what should happen next.