Message channel hub

Overview

The message channel hub primarily analyzes and consolidates conversation data. A conversation is a unit used to track and manage dialogues between a customer and an agent or system (bot).

Conversations can exist in various states, which indicate different stages or conditions of the conversation. Some common conversation states may include "In Progress," "Dispatching," "Completed," etc. Using these states helps you track and organize conversations and lets your team clearly understand the current status.

And each status can be further understood from the "Conversation Status" article to gain deeper insight into the meaning of each status and the corresponding actions.

Based on the image example, you can see that the same customer may have multiple conversations, each representing an independent dialogue sequence. After one conversation is completed, the next message will be considered a new conversation.

These conversations are presented in reports according to their different states and show the situation at the time the report was run. Please note that what the report shows is based on the data at the time of report execution and may not reflect the most current conversation statuses in real time.

Fields

Name	Description	Additional notes
Status	The current status of the conversation, such as "In Progress," "Dispatching," "Completed," etc.
Creation date	"Creation date" usually refers to the start date of the conversation, i.e., the time of the first message when the customer or agent proactively initiated contact.	Messages proactively sent by system segmentation. The times of these messages are all included within the range of the creation date.
Queue entry date	"Queue entry time" refers to the time when a conversation enters the dispatch queue and begins looking for a human agent.
Acceptance date	"Acceptance date" refers to the date when a conversation's status enters dispatching and an agent picks it up and begins providing service.
Agent	Indicates the agent currently assigned to the conversation or the agent who started providing service.
Source channel	The source channel of the conversation, such as LINE, Facebook Messenger, etc.

How to interpret the report?

The way you interpret the report will vary depending on the bot flows or business logic you have designed. Here are some common interpretation examples:

1. Channel comparison: Compare the number of conversations and services across different channels. Observe whether a particular channel has a significantly higher number of conversations, which may indicate that the channel needs more attention and improvement, or that a special situation requires handling.

2. Status analysis: Observe the distribution of each status, for example, Dispatching, In Conversation, Resolved, etc. If a particular status has a notably high count, it may indicate a need to focus on conversation handling efficiency for that status or improve bot flows.

3. Time analysis: For example, a significant difference between the number of queue entries and acceptance times is an important observation point. If there is a clear gap between these two counts, further investigation may be needed to understand the reasons, such as whether staffing is sufficient or whether agents can accept in a timely manner.

In addition to the examples above, interpreting reports should be done according to your business needs and specific context.

Frequently Asked Questions

There is a queue entry date but the status is "Conversation Script" or "Bot"?

If a conversation's status is "Conversation Script" or "Bot" and there is also a recorded queue entry date, this indicates that the conversation once entered the dispatch queue but was subsequently returned to bot service by triggering keywords or similar mechanisms.

For example, a customer requested a human agent via a script on 05-14 at 18:23, so the queue entry date would be recorded as 05-14. Later, the customer might enter a keyword like "leave," which triggers the system to return the conversation to the bot's self-service.

In such cases, the queue entry date records the time the customer initially entered the dispatch queue, and a subsequent keyword trigger caused the conversation to return to bot service.

Is the number of conversations equal to the number of service records?

The number of conversations does not equal the actual number of service records.

Sometimes a customer may obtain the correct answer through the bot flow, self-service, or other means and does not need to request human agent intervention. In such cases, the conversation count increases by 1, but the actual service count remains 0. This means the customer's issue was resolved during the conversation without further human intervention.