Category

Prevent Ticket Reopenings with Thank You Detector in BoldDesk

Updated:

BoldDesk Thank You Detector is an AI‑powered capability designed to analyze inbound replies on Closed or Resolved tickets and prevent unnecessary reopenings caused by non‑actionable acknowledgments such as “thanks,” “thank you,” or “much appreciated.”

In typical support workflows, customers often respond after resolution with simple acknowledgments. While these messages do not require further action, they can unintentionally trigger ticket reopening. This leads to unnecessary agent effort, increased queue noise, skewed reporting metrics, and delays in handling tickets that require attention.

To address this, the Thank You Detector intelligently identifies such gratitude‑only replies and suppresses ticket reopening when no further action is needed—helping maintain efficient and focused support operations.

  • Reopen suppression is applied only when the AI confidence score is ≥ 0.90 (90%). If the confidence score falls below this threshold, the ticket will reopen as per the default behavior.
  • The Thank You Detector is exclusively available in the Momentum (Legacy) and Enterprise plans.

Use Case

The Thank You Detector helps prevent unnecessary ticket reopenings caused by non-actionable customer replies, ensuring more efficient and accurate support operations. Below are the key benefits:

  • Keep ticket queues clean: Prevents reopening tickets that need no action.
  • Reduce agent workload: Avoids unnecessary ticket reviews and closures.
  • Improve reporting accuracy: Ensures metrics reflect meaningful interactions.
  • Maintain better customer experience: Lets customers acknowledge support without triggering workflows.

Prerequisites

  • Role: Administrator (to enable/configure the feature)
  • AI Center access: AI Copilot features must be available to the account

Thank You Detector runs only when a ticket is in Closed or Resolved status; it does not evaluate messages in any other status.

How to Enable Thank You Detector

Thank You Detector helps you stop ticket reopens on customer replies like “Thank you”, “Helpful”, “Resolved”, etc. To enable this feature, follow the steps below;

  1. Navigate to AI Copilot in the AI.
  2. Ensure the AI Copilot features are enabled.
  3. Navigate to Customer Support Assistance under configure.
  4. Check the box Thank You Detector.

    Enable Thank You Detector.png

Managing Thank You Detector

Once you enable Thank You Detector as discussed above, you can manage it by following the steps below;

  1. Click on manage.

    Manage Thank You Detector.png

  2. In the pop-up window, select the brand and source then click on update. You can choose multiple brands and sources.

    Detector Brand and Source.png

  • Detector runs only if the reply source matches selected Brand + Source.
  • If the reply comes from an unsupported brand or channel/source, the detector is skipped and the ticket reopens.

Requester and CC Rules

The detector applies in the following scenarios:

  • An end user (contact) replies through the customer portal or a supported channel.
  • A CC user who is a Contact replies.
  • An Agent replies through the agent portal and is also the requester.

The detector does not apply in the following scenarios:

  • A CC user who is an Agent replies.
  • An Agent replies through the agent portal but is not the requester.

The Thank You Detector is recorded in the ticket history. If a thank you note is not detected and the status changes, it will not be recorded in the ticket history.

Status Configuration Check

  • If it’s set to “Do not perform any action on the ticket state” → the system leaves it as it is, so a closed ticket stays closed.
  • If it’s set to “Automatically change ticket status to open” → the ticket will reopen. The ticket remains closed only if the reply is clearly a non‑actionable thank‑you message (AI confidence ≥ 90%), contains no issue, and does not express a negative sentiment.
  • If it’s set to “Create a new follow-up ticket” → a new ticket is created every time; no conditions will stop this.

Decision Rules Examples

Scenario Outcome
“Thank you”, “Thanks a lot” Ticket remains closed
“Thanks, but I have an issue” Ticket reopens
“Can you check this?” Ticket reopens
“Thanks for nothing” (negative tone) Ticket reopens
System cannot evaluate message Ticket reopens

Troubleshooting Common Errors

Ticket reopened even for a “thanks” reply

  • Verify the ticket was Closed/Resolved at time of reply.
  • Verify Brand and Source/Channel are enabled for the detector.
  • Check if the message included a question/issue or negative sentiment.
  • Confirm confidence met the ≥ 0.90 requirement (via history entry).

Detector did not run

  • Check if the ticket status was not Closed/Resolved.
  • Check if the reply came from an excluded actor (for example, CC user who is an Agent).
  • Check if the brand/channel is unsupported or not selected in settings.

Frequently Asked Questions

  1. Does Thank You Detector run on Open or Pending tickets?
    No. It runs only when the ticket status is Closed or Resolved.

  2. Will a ticket stay closed if the customer says “Thanks, but…”?
    No. If the message includes a new request, unresolved issue, or negative sentiment, the ticket reopens.

  3. What confidence is required to suppress reopening?
    AI confidence must be ≥ 0.90.

  4. Does it work for CC replies?
    Yes, only when the CC replier is a Contact. If the CC replier is an Agent, the detector does not apply.

  5. Can I enable it only for specific brands/channels?
    Yes. Admin can configure Thank You Detector for selective Brand and Source/Channel.

  6. Can the detector prevent creation of a follow-up ticket?
    No. If the status automation is configured to Create a new follow-up ticket, the follow-up is created and the detector does not suppress it.

Related Articles

  1. How to Enable AI Copilot Features in BoldDesk
  2. How to Use AI for Sentiment Analysis in BoldDesk
Was this article useful?
Like
Dislike
Help us improve this page
Please provide feedback or comments
Comments (0)
Access denied
Access denied
Access denied
Access denied

No articles or sections found
No articles or sections found