Advanced statistics [guide]

In this article, we explain how the statistics reports in our advanced statistics module work.

Team performance

This report shows the average chat handling time for each individual agent and at the team level. The handling time is calculated from the moment an agent receives a new chat (this can be a new contact or an existing chat that is reopened with a new customer message), to the moment the agent closes the conversation. This report therefore helps us to understand how long, on average, each agent takes to manage a chat assigned to him and to compare his performance with that of the other agents within the account.

 

In this report you will notice 2 columns defined as p95 and p99: these values ​​indicate the value of the 95 percentile and the 99 percentile for each agent and at team level: this information helps us to understand how the most extreme values ​​of the series used for the calculation are impacting on the average. In fact, it may happen that an agent must, for various reasons, keep a chat open for a longer time than usual. By analyzing the values ​​of the 95 percentile and the 99 percentile, it helps us to understand if the average has been significantly impacted by the cases that took longer to be closed by the agent.

Online time per agent

These two graphs show the time spent online (available) by each agent within the account. The first graph allows you to view the sum of the minutes in which an agent has been available in a certain period of time (it is possible to filter by date and extract the sum of a specific period of time). The second report, just below, lists all the connection intervals recorded by agent and which, when added together for a certain period of time, are the average shown above. This second graph is used to keep track of all the moments in which an agent has set himself as "not available" within his Callbell account.

Number of closed chats per agent

This report shows the number of times an agent has closed their chats during a given period of time. This report is useful for measuring the activity of agents because, if they close the chats once managed and once the response has been sent to the customer (recommended behavior for a correct use of the platform), the graph will show the number of closures made by a agent during a period of time that can be filtered by hour (hour), day (day), week (week) or month (month).

Queued chat

This report shows in real time how many open chats are being managed by each agent. This helps us understand if, for example, a specific agent is overloaded with chats to manage and helps us understand how, if necessary, to redistribute the chats assigned to agents who find themselves with fewer cases to manage.

Note: the report considers all open chats, regardless of whether they are chats that have a message from the agent as the last message.

Unanswered messages

This report shows in real time the list of all open chats that have a customer message as their last message and are therefore waiting for an agent response (or to be closed) and shows, in the last column, the time waiting for the customer (how much time has passed since the customer sent his last message).

 

This report allows us to analyze in real time how many chats are waiting for an action (response or closure of the chat) by the team agents and allows you to quickly reorder the list of chats in a "ranking" by clicking on elapsed time, title of the last column of the report. This allows us to obtain a list of conversations sorted by longer waiting time and in a few clicks, to open the various conversations within the account to understand what is happening on the chats that are delaying the most time to be managed.

 

Note: the report only considers open chats that have a message from the customer as their last message and excludes chats that are already closed and open chats where, however, the last message was sent by the agent.

Automatic routing

This report, available only in accounts that use automatic chat routing, shows, based on the keywords configured in the automatic routing flow, how chats that interacted with the automatic routing bot were distributed. For example, if in our routing flow, under the section "when the answer is" we have configured 2 options (option 1 and option 2), regardless of the team to which the chat is then assigned *, this graph will show how many contacts during a specific period of time have selected option 1 and how many option 2. This report allows us to quickly identify why customers come into contact with our business, by categorizing the keywords used graphically to choose one of the options offered by our automatic routing menu.

 

* we could in fact have X options which, regardless of the choice of contact, are redirected to the same team; this report analyzes, without distinction to which team the chat is then redirected, the number of times a keyword is recognized by the system.

Number of cases managed / renrouted

This chart shows the number of cases handled per agent, for a filterable period of time. This graph is only available if automatic chat routing has been used in your account: a case is defined as a chat which is assigned to an agent and which is renrranged and sent back to the "bot" once finished. Since a chat is only sent back to the bot once we expect customer support to have actually been completed, this chart allows us to identify how many cases each agent handled in a given period of time.

Average time to return a chat to automatic routing, filterable by keyword and by agent (average, p95 y p99)

These 3 graphs (average, 95 percentile and 99 percentile) show the time it takes on average from when a customer initiates a conversation (first with the bot and then with an agent) to the moment the chat is rerouted by the agent and sent back to the bot. This graph also considers the bot keyword that has been chosen by the customer: this allows for a comparison of the average management time between agents based on the type of case managed. In fact, we know that some types of requests may take longer to be resolved than others and for this reason it is useful to be able to compare the management time based on the type of case (the determining factor is the keyword that identifies which team a chat is assigned once the customer responds to the automatic routing welcome message).