The Explore dataset for Chat is called the chat engagements dataset and many of the metrics and attributes discuss chat engagements. What is a chat engagement and how does it work?
A chat engagement is measured by one interaction between an agent and a customer. Each time a customer chats with your team, they may interact with multiple agents. Each time the chat switches to a new agent, this starts a new engagement. For example, if one chat has been served by 2 agents, there will be 2 engagements for that chat.
In the example above, you can see that the first chat in the table has two engagements, meaning one agent took the chat, then transferred it to another agent who also chatted with the visitor. In the first chat in the table:
- There were 15 total messages in the chat as a whole, 8 messages from an agent to the visitor and 7 from the visitor to the agent.
- The first engagement (first agent interaction) consisted of 7 total chat messages, 4 from the first agent and 3 from the visitor.
- The second engagement consisted of 9 total chat messages, 4 from the second agent and 5 from the visitor.
Take a look at the second chat in the table. You can tell the chat was transferred from one agent to another agent, and the second agent never interacted with the visitor. Why? Because the chat shows two engagements, and the second engagement has 0 agent messages.
For more information about the specific engagement metrics and how they are calculated, see this article: Metrics and attributes for Zendesk Chat