Skip to content

Adherence Heatmap

The Adherence Heatmap feature The Adherence Heatmap enables supervisors to track agent adherence trends across metrics over time. It highlights non-compliant interactions and, on hover, displays the date range, total interactions, adhered interactions, and adherence percentage.

Based on the selected date range, agents, and language filters, supervisors can analyze the color-coded adherence data in the heatmap to identify areas of concern, monitor agent compliance across evaluation forms, metrics, or queues, and identify agents who need additional attention. These insights help supervisors make data-driven decisions and take targeted actions to improve agent performance.

Access Adherence Heatmap

Navigate to Quality AI > ANALYZE > Adherence Heatmap.
Agent Leaderboard

Note

To view this feature, you must enable Auto QA in the Settings.

From the dropdown list, you can select an evaluation form to visualize the adherence metrics in the heatmap.

The adherence heatmap has the following key sections:

Adherence Metrics Visualization

This section explains how the system displays and calculates adherence metrics for a selected form across different date ranges.

  • Applicability of Metrics: The applicability of metrics depends on the queue form selected for the following: :

    • For static metrics (by question), the metric applies to all terminated or failed interactions in the respective queue (form).

    • For dynamic metrics (by question), the system applies the metric only when it detects the relevant trigger. If the trigger is missing, the system doesn't apply the metric.

  • Color Coding & Display: Heatmap tile colors reflect the calculated adherence percentage. The system assigns each tile a color based on its adherence value: red represents 0%, green represents 100%, and intermediate colors progress in 10% increments between red and green. It uses gray when the adherence percentage isn’t applicable, including cases where no evaluated interactions match the selected metric and filters.

  • Failed Interactions and Fatal Errors: When an interaction fails, the system applies the following two conditions::

    • Failed Interactions: Includes all agent interactions that don't meet minimum quality or compliance thresholds. This category encompasses both fatal errors and process failures that cause the interaction to fail. Examples include missed greetings, providing incorrect information, or any compliance breaches.

    • Fatal Errors: The system treats fatal errors as critical issues and immediately flags them, regardless of other correct responses or scores. Examples include major compliance breaches. When a fatal error occurs, the system sets the entire interaction’s adherence score or scorecard to 0% (red) and marks the interaction as failed.

    Supervisors Agents can use flagged fatal errors to self-assess and prepare for coaching. Visual adherence scores (heat maps) and question-level feedback support this process. For more information, refer to AI-Assisted Manual Audit.

  • Hover Tooltip: This displays the adherence percentage and number of interactions for each selected date range. If no interactions exist for a range or filter, it displays nothing.
    No Interaction

    Note

    Interactions flagged as Below Threshold or Duration Unavailable are excluded from the interaction count.

  • Adherence Percentage Calculation: Take the number of evaluated interactions that adhere to the metric and divide it by the number of evaluated interactions where the metric applies.

The system calculates the percentage adherence for each tile (box) in the heatmap based on the evaluation form, agent, and date range.

  • Evaluation Eligibility: Consider only evaluated interactions for metric applicability. Include manually evaluated interactions. Determine metric applicability only from evaluated interactions.

Filters

This enables supervisors to drill down into agent performance based on selected criteria. This outlines the following filter types based on the form, date range, and agent selected.

Date Range Filter View

You can filter the data by selecting a specific date range. The system sets the default date range to the last 7 days based on the selected queue. The system applies the user’s system time zone to determine timezone settings.

You can select from three views (Daily, Weekly, and a Custom range):

Date Range View

You can select from three views (Daily, Weekly, and a Custom range):

  • Manual Evaluation Metrics Indicator: Shows Manual Evaluation metrics with a visual indicator and represents unaudited conversations if no data is available in the Adherence Heatmap. Manual Evaluation Metrics Indicator

  • Daily: When the selected date range is 1–14 days, and each tile shows adherence for a specific day, with a horizontal scroll option if the date range is more than 7 days.

  • Weekly: When the selected range is more than 14 days, and each tile shows adherence for the full week, with similar scrolling and display logic as the daily view.

  • Custom Range: All interaction data from the given date (12:00:00 AM to 11:59:59 PM), in the agent’s time zone, limited to 31 days.

Evaluation Form: Select forms across assigned queues and set a default that persists across sessions (both on the Heatmap page and the QA dashboard). Selecting a new default replaces the previous one and highlights fatal metrics with visual indicators.

Agent: By default, the system displays all agents in the selected queues, including those assigned to the queues linked to the selected form. You can search and multi-select them, and update the heatmap to show adherence for their completed interactions. Only adherence tiles for the selected date range, agent, and language remain active; others grayed out and don’t appear on the map.

  • All Languages: This enables multi-selection of languages from the All Languages drop-down menu on the Adherence Heatmap page, corresponding to the languages selected in the evaluation form.

    Note

    When the language filter is applied, only metrics configured for the selected languages appear under the All Languages drop-down menu.

Clickable Functionality

This describes the functionality of a heatmap interface for tracking agent adherence to specific metrics over a selected date range (selected agent, date range, and adherence status).
Click-Through

The clickable date range tiles (each box is a metric) display the interactions for the selected agents, showing the following metrics for each date:

  • Adherence Display: You can sort the adherence column by percentage, with the default order showing the lowest adherence first.
    Adherence Display

  • Heatmap Interaction: Select a date range tile in the heatmap to open a window, which displays agent-wise adherence for the selected metric, ordered from least to most adherence. This displays the percentage count of adherence interactions.
    Heatmap Interaction

  • View Interactions: Select View Interactions to open the Conversation Mining page and view evaluated failed interactions for the selected agents. The system excludes interactions flagged as Below Threshold or Duration Unavailable unless users evaluate them manually. The system applies unsaved filters for the evaluation metric, queues, metric qualification, and failed interaction count.
    View Interactions

View Interactions Filter Parameters

The following describes the filter parameters for View Interactions:

  • Agents with no applicable interactions for the metric appear at the end of the list.
    No Applicable Interaction

  • The system displays No Interactions for N/A cases. The system shows the Agent Adherence tile in gray when there are no interactions and displays No Failed Interactions for agents with 100% adherence. The system disables the View Interactions option for agents with no applicable interactions or with 100% adherence.
    No Interactions

    Note

    • Click the View Interaction button to view the Conversation Mining page, where you can view all failed interactions and see the tags explaining the filtering criteria based on the metric you clicked.

    • Notification tags are shown only when navigating to the Conversation Mining from Adherence Heatmap, and not when accessing the Conversation Mining page directly.

Action Filters

  • Filter Heatmap by Agent: Selecting this icon takes you to the agent’s heatmap metrics page for the selected date range, as shown in the View Interactions screen.
    Filter Heatmap by Agent

  • View Failed Interactions in Conversation Mining: Selecting the agent interaction icon opens the Interactions page in Conversation Mining to view the agent’s failed interactions.

QA or Supervisor Dashboard: This dashboard displays a simplified Adherence Heatmap with data from the last 7 days, uses a default evaluation form, and doesn't include clickable functionality. You can select and set a default evaluation form from the dropdown. The heatmap includes only evaluated interactions. The system excludes contacts flagged as Below Threshold or Duration Unavailable from all aggregates. For more information, see Supervisor Dashboard.

Note

Interactions flagged as Below Threshold or Duration Unavailable are excluded from all heatmap calculations, interaction totals, and adherence metrics. Manually evaluated interactions are included as normal. Excluded interactions remain accessible in Conversation Mining and Reports.