Greetings team, we wanted to surface a request around CSAT and configurable attributions
Problem
CSAT is currently attributed to the most recent visible agent who interacted with the customer, regardless of who owned, managed, or resolved the case. This creates inaccurate and potentially unfair attribution, particularly in collaborative or multi-region support models.
A common example: an engineer owns a case end-to-end, performs the troubleshooting, identifies the issue, and provides the resolution. Due to time-zone coverage, a second engineer in another region sends a brief follow-up and closes the conversation. The customer submits a CSAT rating. That rating is attributed to the closing engineer, who had minimal involvement in the actual resolution.
The reverse also occurs: an engineer receives a negative CSAT simply because they sent the final message, even when the customer's dissatisfaction relates to earlier interactions, product behavior, delays, or factors entirely outside their control.
When CSAT is used as an individual performance metric, this attribution model distorts the data and reduces confidence in the scores as a meaningful measure of agent contribution.
Requested Feature
Configurable CSAT attribution, allowing workspace admins to select from the following models:
- Conversation owner: attribute CSAT to whoever owns the conversation at resolution. Best fit for case-ownership support models.
- Last visible agent: current Intercom behavior, retained as an option.
- Team or conversation-level attribution: distributes CSAT across contributors, useful for collaborative support models where multiple agents share a case.
Business Impact
- Individual CSAT scores become more representative of actual case ownership and contribution.
- Reduces attribution errors caused by regional handovers in follow-the-sun support models.
- Improves the reliability of CSAT as an input into agent performance measurement.
- Better supports collaborative support structures where case resolution is shared across teammates.