Allow us to configure logic to skip CSAT surveys if the final agent message was not viewed by the customer | Community
Skip to main content
Submitted

Allow us to configure logic to skip CSAT surveys if the final agent message was not viewed by the customer

Related products:Workflows
  • June 12, 2026
  • 1 reply
  • 9 views
B

We have encountered situations where users overlook the final message sent by our agents before the conversation is marked as closed, often due to inactivity. Since CSAT surveys are automatically triggered once the conversation is closed, some users submit negative feedback because they believe they never received a response.

It would be helpful to have the ability to configure a workflow condition that skips the CSAT survey if the final agent message has not been viewed or opened. This would help ensure CSAT ratings more accurately reflect the customer's full support experience, rather than being skewed by unseen agent responses.

1 reply

I
  • IlijaVelax
  • New Participant
  • June 12, 2026

 

This is a real gap in how Intercom handles CSAT timing. The cleanest workaround right now is to add a delay step in your closing workflow before the conversation is marked resolved. If you use a Workflow that waits 24-48 hours after the last agent reply before closing, you give customers a better chance of seeing the response first. It does not solve the "unread message" problem perfectly, but it reduces the cases where the ticket closes 2 minutes after an agent replies and the customer never had a chance to read it.
 

Another approach some teams use is to set CSAT to trigger only on conversations where there was at least one customer reply after the agent's last message, which you can approximate by filtering on conversation activity in Workflows. It is not a native "was the message read" gate, but it filters out a chunk of the low-signal ratings where the customer went silent before the final reply landed.
 

Worth upvoting this feature request too, because read-receipt-based survey suppression would be genuinely useful and it is the kind of conditional logic that Workflows should support natively.

 

(Disclosure: I work on Velax, an analytics layer on top of Intercom at https://velax.ai - not directly relevant here, but we do track CSAT trends per agent if you want to spot which agents are most affected by this kind of ghost-close rating skew.)

Terms & ConditionsAccessibility statement