Message Extracts
Updated
This extract provides granular visibility at the individual message level. Each record corresponds to a unique message identified by a Universal Message ID and captures key attributes such as message type, intent, and other message-level properties.
Purpose
To analyze message-level interactions across channels using unique message identifiers.
To track message type, intent classification, and other message-specific attributes.
To support detailed audits of customer and brand communication exchanges.
To enable granular reporting for interaction volume and engagement analysis.
Extract Configuration and Availability
The extract configuration and availability details are listed below:
Extract Attribute | Details |
Channel Split | All channels (based on Channel Snapshot at message level) |
Agent Split | Agent-level (based on Agent ID) |
Granularity | One row per unique Universal Message ID. |
Date Filter | Extract depicts data based on "Message Post Date". |
Update Frequency | 1 mins after the "Message Post Time" |
Custom Field Behavior | Custom Fields by default are not added in this extract but feasibility of custom field exists in the particular extract based on below:
Default Behavior: 1. Case Custom Field: Current Value 2. User Custom Field: Current Value 3. Message Custom Field: Current Value |
Data Schema
The schema details are listed below:
Metrics / Dimensions | Definition | Unique Key | Data Type | Is Null |
Date | Date on which event occurred. | Date | N | |
Case Details | Associated case to the event, which includes the Case subject, Case ID and the Description. | Varchar | N | |
Case Number | The ID number associated to the Case Details | Y | Number | N |
Channel (Snapshot) | Provides the name of the channel that is associated to the case at the time of the particular action. Once the channel is changed, any action done on the case post that will be reported against the changed channel name i.e it will indicate the snapshot channel associated to the case. | Varchar | N | |
Account | The name of the social network account. When used with case details, this will display the name of the "Brand Account" through which the first message associated to the case is grabbed. Messages are ordered in a case based on the time they were associated to the case, not necessarily the date and time the messages were created. | Varchar | N | |
Message Type | Indicates the type of message if this is a live chat message, voice transcript or a voice recording. | Y | Varchar | N |
Message Sent By | Denotes the owner of the message if the message is being sent by the fan (customer) or the brand (system and agent). | Varchar | N | |
Agent | The sprinklr user for whom the report is being generated. | Varchar | N/A if message sent by is 'Fan' | |
Agent ID | ID of brand who made the response on the case including bot and system messages. | Y | Number | N/A if message sent by is 'Fan' |
Message Date | The calendar date when the inbound or the outbound message was sent | Y | Timestamp | N |
Message ID | Denotes the unique message ID associated to each message of the case details. | Y | Varchar | N |
Message | The text of the inbound or outbound message associated to the case. | Varchar | N | |
Total Message Count | The number of inbound and outbound messages to an account and the particular case including messages from the customer, or replies by the agent/ bot. | Number | N | |
Has Brand Response? | This denotes If the Inbound message has brand response or not. | Boolean | N |
Data Settings
Reporting Preferences allow users to customize the default formats for specific metric types. These preferences are applied at the user level. If no changes are made, the system will continue to honor the following defaults:
Default Date Format: Day, MM DD, YYYY - Eg: Tue, Jan 15, 2026
Default Timestamp Format: MM DD, YYYY, HH:MM AM/PM - Eg: Jan 15, 2026, 12:36 AM
Default Time Duration Format: HH (h) MM (min) SS (s) - Eg: 3h 02m 35s
For more details, refer to the related article.
When Custom Fields are added in an Extract, they must be treated as Unique Key during calculations to ensure accuracy and consistency.