Availability: Transcript Analytics data is available exclusively for Genesys customers on the CX3 plan, or CX2 plan with the Speech and Text Analytics add-on. This data requires your Admin to run diagnostics and update permissions. If you are unsure whether your account includes this feature or need help updating permissions, contact us at support@brightmetrics.com.
Transcript Analytics fields let you report on conversation-level behavior like sentiment, speech time, silence, and overtalk. These fields are available on Agent, Queue, Evaluation, and Session reports. Use them to spot patterns by topic, workflow, and time period, and to measure the impact of coaching or process changes over time.
- Transcript Analytics reflect entire conversations, not individual sessions.
- When conversations involve multiple queues or agents, metrics like Sentiment Score and Speech Times can combine all participants and do not indicate individual agent performance.
- If you need agent-specific insight, use Agent Activity data and apply a filter to Number of Agents in Conversation to confirm the conversation was not transferred or handled by multiple agents.
1. Where you will see Transcript Analytics fields
In the report builder, Transcript Analytics metrics appear in the Values list alongside your other available metrics. Add them to a report by dragging them into the Values area.
Common layout:
| Rows | Agent Name, Topics (or Queue, Team, Wrap-up Code) |
| Values | Sentiment, Overtalk, Speech time, Silence |
| Filters | Date range (and optionally Agent Name, Queue, Topic) |
2. Add these fields to a report
-
1
Open the report you want to edit (or create a new report). -
2
In Layout, set your Rows (for example: Agent Name and Topics). -
3
In Values, drag in the Transcript Analytics fields you want to track. -
4
In Report Filters, set your Date filter and apply any additional filters you need. -
5
Refresh the report to load the updated results.
Start with a short date range and a single queue while you validate results, then widen your filters after the report looks right.
3. Available fields and what they are good for
Transcript coverage
| Field | What it measures |
|---|---|
| Count - Has Speech and Text Analytics | The count of conversations in the grouping that include transcript analytics. |
Use it for: Validating your dataset before interpreting sentiment, overtalk, speech time, or silence.
Sentiment
| Field | What it measures |
|---|---|
| Sentiment Score | A sentiment score for the selected grouping. |
| Avg Sentiment Score | The average sentiment score across the conversations included in the grouping. |
Use it for: Comparing trends across time, topics, agents, or queues. For comparisons, averages are typically the most stable view.
Conversation time distribution
| Field | What it measures |
|---|---|
| Time - Agent Speech | The total amount of time agents are speaking across the conversations in the grouping. |
| Time - Agent Speech Avg | The average amount of time agents are speaking across the conversations in the grouping. |
| Time - Customer Speech | The total amount of time customers are speaking across the conversations in the grouping. |
| Time - Customer Speech Avg | The average amount of time customers are speaking across the conversations in the grouping. |
| Time - Silence Avg | The average amount of silence time across the conversations in the grouping. |
| Time - Silence Total | The total amount of silence time across the conversations in the grouping. |
Use it for: Understanding how conversation time is allocated between agents, customers, and silence. Useful for identifying imbalances, long pauses, or workflow steps where conversations slow down — especially when broken out by topic, queue, or time period.
Overtalk
| Field | What it measures |
|---|---|
| Count - Overtalk Instances | The count of overtalk instances in a conversation. |
| Count - Overtalk Instances Avg | The average number of overtalk instances across the conversations in the grouping. |
| Time - Overtalk Speech | The total amount of time overtalk occurs. |
| Time - Overtalk Speech Avg | The average amount of overtalk time across the conversations in the grouping. |
Use it for: Spotting friction and interruptions during specific topics or workflow moments.
4. Useful examples
Spot which topics drive the most friction
| Rows | Queue → Topics |
| Values | Count - Overtalk Instances Avg, Avg Sentiment Score |
Group conversations by Topic within a Queue and compare average overtalk instances against average sentiment score. Topics with high overtalk and low sentiment are friction hotspots — often billing disputes, escalations, or complex troubleshooting. This tells you exactly where to focus agent coaching or update self-service options.
Add Count - Has Speech and Text Analytics to confirm you have enough conversations in each topic grouping before drawing conclusions.
Identify agents who dominate — or go quiet — at key moments
| Rows | Agent Name → Topics |
| Values | Time - Agent Speech Avg, Time - Customer Speech Avg, Time - Silence Avg, Count - Handled |
| Filter | Number of Agents in Conversation = 1 |
Break speech time down by agent and topic. An agent with unusually high Agent Speech Avg on a topic like "Dissatisfaction" may be over-explaining rather than listening. An agent with high Silence Avg may be struggling to navigate a workflow step. These patterns are hard to see in handle time alone, but stand out clearly when speech time is broken out by topic.
You must apply the filter to the Number of Agents in Conversation field in order to accurately identify individual performance. See the note at the top of this article.
Measure the impact of a coaching initiative over time
| Rows | Date Group - Week → Agent Name |
| Values | Count - Handled, Avg Sentiment Score, Count - Overtalk Instances Avg, Time - Silence Avg |
| Filter | Number of Agents in Conversation = 1 |
After rolling out coaching on a specific behavior — such as reducing interruptions or improving how agents handle dissatisfied customers — trend these metrics week-over-week for the targeted agents by extending the date range of your report. A declining overtalk average paired with a rising sentiment score signals the coaching is working. This gives leaders a concrete, data-backed way to close the loop on performance improvement efforts rather than relying on anecdotal observation.
You must apply the filter to the Number of Agents in Conversation field in order to accurately identify individual performance. See the note at the top of this article.
Next steps
Once you have a working report, consider saving it as a favorite or scheduling it to run automatically so your team has a consistent view of these trends over time. As you build familiarity with the data, you can layer in additional dimensions — like Wrap-up Code or Team — to refine your analysis further.
Have questions or need help getting started? Email support@brightmetrics.com.