Tracking deflections in ITSM Virtual Agent

  • Release version: Yokohama
  • Updated January 30, 2025
  • 2 minutes to read
  • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Tracking deflections in ITSM Virtual Agent

    ITSM Virtual Agent includes pre-built deflection topics and topic blocks that help track incidents resolved through Virtual Agent interactions. A deflection occurs when the Virtual Agent resolves an issue—either by creating an incident in the correct assignment group or by directly addressing the user's problem—thus potentially eliminating the need for a live agent or manual incident creation. Deflection tracking is enabled by default, using configurations available in the base system that can be customized as needed.

    Show full answer Show less

    Key Features

    • Pre-built Deflection Patterns: The system provides various deflection patterns that classify Virtual Agent outcomes, including confirmed deflections (e.g., self-resolving tasks, incident triage, knowledge base resolutions), potential deflections (e.g., search results served, query responses), and no deflection scenarios (e.g., inability to assist the user or incomplete tasks).
    • Deflection Metric Storage: All deflection metrics are recorded in the Deflection Metric [ssadeflectionmetric] table, enabling detailed analysis of Virtual Agent performance against each pattern.
    • Analytics and Reporting: Deflection metrics can be reviewed via the Deflection Metrics module in the Self-Service Analytics framework or through the Overview tab of the Conversational Analytics Dashboard. These tools require activation of the Self-Service Analytics Core plugin for access.

    Practical Benefits

    • Track and measure how effectively the Virtual Agent resolves user issues without requiring live agent intervention.
    • Identify which types of interactions lead to confirmed or potential deflections, helping prioritize improvements in Virtual Agent conversations.
    • Leverage analytics dashboards to monitor productivity gains and incident resolution efficiency driven by Virtual Agent usage.
    • Customize deflection configurations and patterns to align with organizational workflows and reporting needs.

    You can use the pre-built deflection topics and topic blocks in Virtual Agent conversations to track the incidents that ITSM Virtual Agent helped to resolve.

    A deflection happens when ITSM Virtual Agent resolves an issue for you. By asking you relevant questions, ITSM Virtual Agent can create an incident in the correct assignment group. As a result, you might not need to create an incident, or need the assistance of a live agent.

    Deflection tracking uses the topics and topic blocks created in Virtual Agent Designer and is enabled by default. For more information, see Tracking deflections using deflection topic blocks.

    By default, ITSM Virtual Agent includes pre-built deflection configurations in the base system. You can use the deflection configurations in the base system or customize them, as needed. For more information, see Create deflection configurations and patterns.

    Base system deflection configuration and patterns.

    The pre-built deflection patterns in the base system are described in the following table.

    Deflection pattern name Description Outcome
    ITSM VA-Self-Resolving Virtual Agent completes a task with or without filing a task record for ITSM. For example, Virtual Agent adds a user to a distribution list. Confirmed deflection
    ITSM VA-Triage & Created Virtual Agent filed and created a task record based on end-user input. Creating this task record shortens the time to resolution. Confirmed deflection
    ITSM VA-KB-Resolve User resolved the issue by using the Knowledge Base (KB) articles displayed. Confirmed deflection
    ITSM VA-Intercept & Resolved Virtual Agent intercepted an end user-created incident, and helped to resolve and close it. For example, Virtual Agent resolves and closes the incident with Issue Auto-Resolution (IAR). Confirmed deflection
    ITSM VA-Search-served Virtual Agent shows the search results based on the user's IT service needs. Potential deflection
    ITSM VA-AI-Search Virtual Agent uses the AI Search results to provide support. Confirmed deflection
    ITSM VA-Query-response Virtual Agent responds to a customer query. It's unclear whether the response was helpful. Potential deflection
    ITSM VA-No-deflection Virtual Agent is unable to assist the user. No deflection
    ITSM VA-User Productivity-Complete User completes a task using Virtual Agent. This pattern tracks those completed tasks that saved time and improved productivity. No deflection
    ITSM VA-User Productivity-Info Provided User was provided information that might be helpful to them. No deflection
    ITSM VA-User Productivity-Incomplete User wasn't able to complete a task using Virtual Agent. No deflection
    ITSM VA-Agent Productivity-Complete Agent (fulfiller) was able to complete a task using Virtual Agent as a service channel. This pattern tracks those completed tasks that saved time and improved productivity. No deflection
    ITSM VA-Agent Productivity-Info Provided Agent (fulfiller) was provided information that might be helpful to them. No deflection
    ITSM VA-Agent Productivity-Incomplete Agent (fulfiller) wasn't able to complete a task using Virtual Agent as a service channel. No deflection
    The deflection metrics are generated and stored in the Deflection Metric [ssa_deflection_metric] table. To view the deflection metrics for each deflection pattern, you can choose one of the following methods:
    • Navigate to the Deflection Metrics module in the Self-Service Analytics framework.
      Note:
      The Self-Service Analytics framework requires the Self-Service Analytics Core plugin (com.snc.self_service_analytics_core).

      For more information, see Activate Self-Service Analytics.

    • Use the Overview tab of the Conversational Analytics Dashboard.

      For more information about this dashboard, see Conversational Analytics Dashboard.