Recommended Actions

  • Release version: Australia
  • Updated March 12, 2026
  • 3 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 Recommended Actions

    Recommended Actions in ServiceNow allows agents to receive context-specific suggestions to enhance case resolution efficiency. This feature helps display relevant actions based on the record context or recommends values for specific fields, streamlining the decision-making process for agents.

    Show full answer Show less

    Key Features

    • Guidance: Provides actionable suggestions displayed as cards in the contextual side panel. Agents can easily access and apply these recommendations to resolve customer issues.
    • Field Recommendations: Offers suggested values for specific fields. Depending on configuration, these values can be auto-filled or presented as messages for existing records, aiding in quicker data entry.
    • Decision Trees: Guides agents through a multi-step process with a series of questions to help troubleshoot issues. This feature requires the Guided Decisions application.
    • Configuration Elements: Involves contexts, context inputs, rules, recommendations, resource generators, and arbitration parameters to tailor the recommendations dynamically based on record context and conditions.

    Key Outcomes

    By utilizing Recommended Actions, agents can expect to resolve customer issues more efficiently, improve case handling accuracy, and enhance overall service delivery. The structured recommendations facilitate quicker decision-making, ensuring agents have the necessary guidance and resources at their fingertips.

    Use Recommended Actions to display relevant actions to agents based on a context of a record or recommend a value for a field.

    Action types in Recommended Actions

    Action type Description Use cases
    Guidance
    A guidance is an action that an agent can take or information that an agent can share.
    • Guidances appear as cards in the contextual side panel in a workspace.
    • Agents can access these cards by selecting the Recommended Actions icon in the side panel.
    • Agents can perform the action by selecting the button on the card.
    • Mary created a case to report an issue with a router. The agent assigned to this case is recommended to apply a resolution from a solved case to resolve the customer issue. The resolution notes and a resolution code from the solved case are copied into the Closure Information section of the current case.
    • Paul, who has an account with StellarVest bank created a case to report a loss of credit card. The agent assigned to this case is recommended to create a work order to dispatch a new credit card for Paul.
    Field recommendation A field recommendation is a value that is recommended for a field.

    Depending on the configuration, the recommended field values are auto-filled or shown as messages underneath the fields for the new records. The recommended field values are shown as messages only underneath the fields for the existing records.

    • You can configure the fields with the text and values that appear in the messages.
    • You can also configure a confidence threshold to auto-fill a recommended value.
    • You can configure a preference to auto-fill a recommended value.
    • An agent wants to reassign the case to a different team who is better equipped to resolve a customer issue. A field recommendation can recommend the assignment group based on the text in the case short description.
    • While interacting with a customer, an agent creates a case on behalf of the customer. Based on the short description the agent enters in the case form, the values for the fields Product, Assignment group, and Assigned to are recommended to the agent.
    Decision tree A decision tree is a guided flow for agents to follow. A decision tree is a multi-step process that asks a series of questions. Based on the answers that an agent enters, a guidance is provided.
    Note:
    The Guided Decisions application (sn_gd_core) is required to create decision trees.
    		 Rina created a case to report that their phone isn’t charging. The agent assigned to the case is recommended to use a decision tree that asks a series of questions to troubleshoot the issue with the phone and provides a guidance in the end.
    Actions of the type Guidance and Guided Decision Tree are displayed as cards within the contextual side panel.
    Figure 1. Recommended actions - guidance card
    Guidance cards recommending the agent to view and attach a knowledge article or attach and add the link in comment as primary action
    Actions of the type Field Recommendation are auto-filled in the fields or appear as messages under the fields on a record form. Agents can follow the recommendations and select these field values.
    Figure 2. Field recommendation messages in the Incident form
    Case form showing highlighted field recommendations under the appropriate fields. For example, based on short description, service desk is the recommended field for assignment group.

    For more information, see Creating guidance and field recommendation in Recommended Actions.

    Elements for configuring recommended actions

    Configuring a recommended action is a multi-step process that involves:
    Context
    A context enables agents to see recommended actions for a record in that table when certain rules are met. For more information, see Contexts in Recommended Actions.
    Context input
    A context input enables you to utilize the entities other than the context table to define rules, recommendations, and resource generators so that recommendations are updated dynamically as the context changes. For more information, see Context inputs in Recommended Actions.
    Rule
    A rule is a set of conditions that applies to a context and determines when a recommended action appears for records in the context table. For more information, see Rules in Recommended Actions.
    Recommendation
    A recommendation is a way to suggest an action to an agent. You can create recommendations with action types of guidance and field recommendation. For more information, see Recommendations in Recommended Actions.
    Resource generators
    Resource generators provide helpful information for guidance and field recommendations. The resource generators use decision table, flow, scripts, Predictive Intelligence framework, or AI search capabilities to generate resources. For example, a resource generator can provide a knowledge article link that can then be used as a recommended action for a case. For more information, see Resource generators in Recommended Actions.
    Arbitration parameters
    Arbitration parameters determine the frequency of issues or the priority order of the recommended actions so that agents get the guidance that they must help resolve customer issues.