Call Logs

Agent Conversations and Campaign Call History provide the details you need to understand individual conversations. Use filters to narrow the list by date, agent, direction, campaign, status, or outcome before opening a record. Campaign views add contact search, outcome tabs (All, Qualified, Callback, No Outcome, Pending), and date-range selection. Choose the agent history for general voice and chat activity, or campaign history for calls connected to a campaign.

​

What you can review

• caller and agent information;
• call type, status, and duration;
• transcript and recording, when enabled;
• summary and outcome;
• errors and usage information.

Campaign records can also show the campaign, contact, agent, call status, business outcome, qualification result, and lead information.

​

Understand the record

The transcript shows what the system recognized and generated. The recording, when available, provides the actual audio context. The summary and outcome help with review, but should be checked against the conversation when a decision or customer issue depends on accuracy. Call status describes the technical lifecycle of the call. Business outcome describes what the conversation achieved. Review both when measuring performance.

​

Investigate a call

1. Open the call record.
2. Compare the transcript with the recording.
3. Check the agent, knowledge, and transfer behavior.
4. Record any change needed to the prompt or knowledge.
5. Check whether the same issue appears in nearby calls.
6. Retest the affected journey after making a correction.

If the issue appears across several calls, compare their agent version, time range, provider, and campaign. A repeated pattern usually points to a shared configuration or content issue.

​

Create a review routine

Review a mixture of completed, failed, transferred, and unusually short or long calls. Sample across different agents, languages, and campaign periods. Tag or record repeated findings so the team can prioritize the change with the greatest customer impact. A sustainable weekly routine: Monday, triage all failures and transfers from the weekend; Wednesday, sample 10 completions for answer quality and update the test script with new phrasing; Friday, compare the week’s outcome distribution with the previous week and file one improvement (knowledge, prompt, or routing) with its evidence call IDs. Log every finding as agent, symptom, example call ID, suspected cause, and fix - without the call ID, findings are opinions.

​

Use call data responsibly

Limit access to people who need it, follow your retention policy, and avoid copying recordings or personal information into unapproved tools. Share a call identifier rather than full customer details when requesting support.