Customer support is one of the most important use case of conversational bots. It provides for all-time availability with a significant reduction in operational costs since it reduces the number of human resources required for the support process.
Analytics in support provides an opportunity to identify potential growth areas as well as the efficiency of the process. All the requests that are generated for live chat are tracked here in this segment. This is important to monitor for quality support provided to the bot users. All inbound requests, agent activity, resolution time and queuing time are tracked here for a bot in a given time span. A bot owner can use this analysis to manage the number of agents required and analyze the performance of support agents.
The support dashboard is available to all users who are on the Business plan and above. To access this functionality click on Dashboards on the left panel and click on support.
- Inbound Requests: This is the cumulative number of all live chat requests received for a bot in a given span of time.
- Average Resolution Time: This is the resolution time of all requests averaged over a number of requests over the given span of time.
- Average Queueing Time: This is the cumulative waiting time that each user spent before being attended by an agent, averaged over a total number of users over a given span of time.
- Average Talk Time: This is the cumulative duration of interactions between an agent and a user averaged over all the users.The graph below provides a highlight of the status of inbound requests per day. A user can click on the resolved chat/incoming chat color bubble to select or deselect it from the graph.
The data represented in the graph above is represented in tabular format in the below table.
4. Agent Information
In the agent tab, a bot owner can get information on the performance of the agents.
- Chat Handled: This displays the total number of chats handled by an agent.
A higher number displays the efficiency of an agent but can not be conclusive, average talk time should also be analyzed.
- Chat Transferred: All the chats that an agent transfers to another agent are recorded here.
If the number of chats transferred is too high the bot owner should replan the routing process,
- Talk time: This is the cumulative sum of the talk time of all the chats handled by the agent.
- Average talk time: This is the talk time-averaged over the number of chats handled.
If average talk time is too less it may indicate that the agent is not taking an initiative to answer the queries properly or a very high average talk time may show inefficient handling of customer support requests. In both, the case improvement can be done to provide better support.
5. Category Information
The agent category is defined at the time of assigning the agent role. ALL and Default are the two categories predefined in the system.
The above table shows the category-wise distribution of the inbound requests/chats.
- Incoming Chats: This is the total number of chats received for a given category.
- Resolved Chats: This the number of chats that are resolved from the number of chats that were received for a given category. If this number is low in comparison to incoming chats it shows that resolution is slow and has to be evaluated.
- Avg. Waiting Time: This is the time a chat request was kept on waiting in a particular category. Waiting time is averaged for all chats. This metric can be used to improve the category wise support process and reduce the waiting time to provide better customer support.
- Total Talk Time: This is the total talk time that was spent on all the chats in the given category.
- Avg. Talk Time: This the total talk time-averaged over a number of chats in the given category. If average talk time is too less it may indicate that the agent is not taking an initiative to answer the queries properly or a very high average talk time may show inefficient handling of customer support requests. In both, the case improvement can be done to provide better support.