The By Transfer Etiquette metric automatically evaluates how agents handle internal customer transfers during voice interactions, ensuring consistent adherence to transfer standards throughout the handoff.
The metric assesses three aspects of pre-transfer communication:
| Aspect | Description |
|---|
| Customer notification | Verifies that agents inform customers before initiating a transfer |
| Reason explanation | Checks whether agents clearly explain why the transfer is necessary |
| Destination information | Evaluates whether agents communicate where the customer is being transferred |
Key Benefits
- Evaluate 100% of internal transfer interactions automatically.
- Reduce customer frustration from poor handoffs.
- Identify coaching opportunities for better transfer communication.
- Maintain consistent service quality throughout the transfer process.
Create By Transfer Etiquette Metric
-
Navigate to Quality AI > Configure > Evaluation Forms > Evaluation Metrics.
-
Select + New Evaluation Metric.
-
From the Evaluation Metrics Measurement Type dropdown, select By Transfer Etiquette.
-
Enter a descriptive Name (for example, “Pre-Transfer Communication Check”).
-
Select the applicable Language(s). You must select at least one; you can add or remove languages later.
-
Enter an evaluation Question (for example, “Did the agent inform the customer before transferring to another department?”).
-
Configure Pre-Transfer Communication Assessment — verifies whether agents provided required information before transferring.
-
Configure the Detection Window — defines how the system detects agent behavior before the transfer:
| Option | Description |
|---|
| Last Utterance Before Transfer | Evaluates only the final agent utterance before transfer |
| Configurable Time Window | Set a detection window in seconds (1–60, default: 10) |
-
Select an Evaluation Method:
Gen AI-Based
Uses natural language analysis to assess transfer etiquette contextually.
-
Enter a Description of expected behavior (for example, “Agent must inform the customer about the transfer, explain the reason, and provide destination information”).
Deterministic ML (BGE-m3)
Compares embedded similarity scores against sample reference utterances.
-
In Answer, add sample utterances (for example, “I’d like to transfer this call to another department”).
- Add multiple utterances per configured language. Each language has its own set.
- Check/uncheck utterances to include or exclude them.
-
Set the minimum Similarity threshold (0–100%, default 60%):
- Below 60% — orange (does not meet standard)
- 60–100% — green (meets standard)
-
Select similar utterances from the sample list to enable the Create button.
-
Select Create.
The Create button activates only after all validation checks pass. Once saved, the metric becomes active and appears in the Evaluation Metrics dashboard.
Scoring Logic
When Transfers Are Detected
| Context | How scoring works |
|---|
| Evaluation Forms | Each transfer instance is assessed separately. All transfers must pass for the metric to pass. If any transfer fails, the entire metric weight is deducted. |
| Scorecards | Each agent’s portion is evaluated across all transfers. If any instance fails, the specific weight is deducted entirely. |
