> ## Documentation Index
> Fetch the complete documentation index at: https://docs.reddy.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Metrics Glossary

> Definitions and calculations for all performance metrics in Reddy dashboards

Reddy displays performance metrics across Training and Live views. This glossary explains how each metric is calculated.

## Training Metrics

Metrics for agent performance in training simulations (Script Trainings, Systems Trainings, Mini Modules, and Objections).

<AccordionGroup>
  <Accordion title="Training Score" icon="chart-line">
    **Also known as:** Training Compliance Score

    **Calculation:** Average compliance score across all completed training simulations within the selected date range.

    **Score Range:** 0-100%

    **Breakdown by Training Type:**

    Training scores are weighted averages across different simulation types:

    | Training Type      | Weight | Description                            |
    | ------------------ | ------ | -------------------------------------- |
    | Script Trainings   | Full   | Traditional call script simulations    |
    | Systems Trainings  | Full   | Software/system navigation simulations |
    | Mini Modules       | 1/5    | Short scenario-based exercises         |
    | Objection Handling | 1/10   | Objection response scenarios           |

    <Tip>
      Click "View breakdown" in the dashboard to see scores split by training type.
    </Tip>
  </Accordion>

  <Accordion title="Training AHT" icon="clock">
    **Also known as:** Training Average Handle Time

    **Calculation:** Average time per completed end-to-end training simulation.

    **Includes:** Script Trainings and Systems Trainings only

    **Excludes:** Mini Modules and Objection training sessions

    <Note>
      Mini Modules and Objections are excluded because they're significantly shorter than full simulations and would skew the average.
    </Note>
  </Accordion>

  <Accordion title="Total Training Time" icon="stopwatch">
    **Calculation:** Sum of all time spent actively inside training simulations from when the agent clicks "Start Simulation" through when they click the button that ends the simulation and takes them to the feedback screen.

    **Includes:** All training types (Scripts, Systems, Mini Modules, Objections)

    **Per Agent View:** Shows cumulative training time for the specific agent

    **Dashboard View:** Shows total across all agents in the selected filter

    <Info>
      This metric counts the actual simulation time, not the time reviewing feedback or viewing results.
    </Info>
  </Accordion>

  <Accordion title="Average Training Time" icon="chart-user">
    **Dashboard View Only**

    **Calculation:** Average total time spent per agent on all training simulations within the selected date range.

    **Formula:** Total training time across all agents ÷ Number of unique agents
  </Accordion>
</AccordionGroup>

## Live Call Metrics

Metrics for agent performance on live customer calls.

<AccordionGroup>
  <Accordion title="Average Compliance" icon="badge-check">
    **Calculation:** Average QA compliance score across all live calls within the selected date range.

    **Score Range:** 0-100%

    **Minimum Duration Filter:** Calls under 60 seconds are excluded by default (configurable per organization)

    <Info>
      Compliance scores are based on your organization's QA scorecard configuration.
    </Info>
  </Accordion>

  <Accordion title="Average Handle Time" icon="clock">
    **Also known as:** Average Call Duration, AHT

    **Calculation:** Mean duration of all live calls within the selected date range.

    **Measurement:** Time from call start to call end (in minutes)

    **Minimum Duration Filter:** Calls under 60 seconds are excluded by default
  </Accordion>

  <Accordion title="Number of Calls" icon="phone">
    **Also known as:** Total Calls, Call Volume

    **Calculation:** Total count of live calls within the selected date range.

    **Per Agent View:** Shows calls for the specific agent

    **Dashboard View:** Shows total across all filtered agents
  </Accordion>

  <Accordion title="Average Daily Active Agents" icon="users">
    **Dashboard View Only**

    **Calculation:** Average number of unique agents who took calls each day within the selected date range.

    **Formula:** Total unique agents per day summed across all days ÷ Number of days in date range

    <Tabs>
      <Tab title="Example Calculation">
        Date range: 7 days

        * Day 1: 15 agents made calls
        * Day 2: 18 agents made calls
        * Day 3: 16 agents made calls
        * Day 4: 14 agents made calls
        * Day 5: 17 agents made calls
        * Day 6: 12 agents made calls
        * Day 7: 16 agents made calls

        Average Daily Active Agents = (15 + 18 + 16 + 14 + 17 + 12 + 16) ÷ 7 = **15.4 agents**
      </Tab>
    </Tabs>
  </Accordion>
</AccordionGroup>

## Additional Metrics

<AccordionGroup>
  <Accordion title="Calls Per Agent" icon="user-check">
    **Dashboard View Only**

    **Calculation:** Average number of calls per agent within the selected date range.

    **Formula:** Total calls ÷ Number of unique agents who took calls
  </Accordion>

  <Accordion title="Sessions Per Agent" icon="graduation-cap">
    **Training Dashboard View Only**

    **Calculation:** Average number of training sessions per agent within the selected date range.

    **Formula:** Total training sessions ÷ Number of unique agents who trained
  </Accordion>
</AccordionGroup>

## Understanding Trend Changes

Many metrics display trend indicators showing performance changes compared to a previous period.

**Trend Calculation:** Current period value minus previous period value

**Previous Period:** Equal length to current date range, immediately preceding it

<Tabs>
  <Tab title="Example">
    Current date range: Last 7 days (Nov 5-11)

    Previous period: 7 days before that (Oct 29 - Nov 4)

    If current compliance is 85% and previous was 80%, the trend shows: **+5%** ↑
  </Tab>
</Tabs>

<Warning>
  Trend changes require sufficient data in both periods. If the previous period has no data, no trend will display.
</Warning>

## Time Measurements

### How Training Duration is Measured

Training session duration starts when the agent clicks "Start Simulation" and ends when they click the button to complete the simulation and view feedback.

**Duration includes:**

* Time spent listening to and speaking with the AI customer
* Time navigating through system screens (Systems Trainings)
* Silence or thinking time during the simulation

**Duration excludes:**

* Time reviewing feedback after simulation ends
* Time viewing the QA scorecard
* Time on the simulation setup screen before clicking "Start"

### How Live Call Duration is Measured

Live call duration is measured from call start to call end as provided by your phone system or API integration.

## FAQ

<AccordionGroup>
  <Accordion title="Why do my metrics show a dash (-)?">
    A dash appears when there is no data for the selected date range and filters. This usually means:

    * No calls or simulations occurred in the selected period
    * Filters are too restrictive (e.g., filtered to a product with no activity)
    * The agent has not completed any qualifying sessions

    Adjust your date range or remove filters to see if data appears.
  </Accordion>

  <Accordion title="Why does Training AHT differ from Total Training Time?">
    Training AHT is the *average time per simulation*, while Total Training Time is the *cumulative sum* of all simulation time.

    If an agent completed 5 simulations of 8 minutes each:

    * Training AHT: 8 minutes
    * Total Training Time: 40 minutes (5 × 8)
  </Accordion>

  <Accordion title="Do incomplete or abandoned simulations count?">
    No. Only completed simulations where the agent reached the end and viewed feedback are included in training metrics. Simulations that were started but abandoned do not count toward metrics.
  </Accordion>

  <Accordion title="Why are short calls excluded from metrics?">
    Calls under 60 seconds (configurable) are typically incomplete, test calls, or disconnects. Excluding them provides more accurate performance measurements. Your organization can adjust this threshold.
  </Accordion>

  <Accordion title="How often do metrics update?">
    * **Training metrics:** Update immediately when a simulation is completed
    * **Live call metrics:** Update in real-time as calls are processed
    * **Dashboard aggregations:** Refresh when you change filters or date ranges

    Manual uploads may take a few minutes to process before appearing in metrics.
  </Accordion>

  <Accordion title="Why do agent-level and dashboard-level metrics differ?">
    Agent-level metrics show performance for a single agent, while dashboard-level metrics aggregate across multiple agents. Some metrics (like Average Daily Active Agents) only appear at the dashboard level because they measure team-wide patterns.
  </Accordion>
</AccordionGroup>

## Want Custom Metrics?

Track business-specific data by adding custom metadata to calls. Filter and analyze using your own metrics alongside standard Reddy metrics.

[Learn how to add custom metadata →](/guides/add-call-metadata)