EngagementSignalMetric

Represents a measurable quantity that’s derived from an engagement signal, such as the sum of revenue or a count of clicks. Use this object to track user engagement for A/B tests, machine learning model training, and attribution configurations.

Introduced in API version 62.0.

create(), delete(), describeLayout(), describeSObjects(), getDeleted(), getUpdated(), query(), retrieve(), undelete(), update(), upsert()

  • Type: picklist

  • Properties: Defaulted on create, Filter, Group, Nillable, Restricted picklist, Sort

  • Description:

    Defines the type of calculation used on the metric field.

    The default value is Count.

  • Possible values:

    • Avg — Avg
    • Count — Count (default)
    • Distinct — Distinct
    • Select — Select
    • Sum — Sum
  • Default value: Count

  • Type: reference

  • Properties: Create, Filter, Group, Sort

  • Description:

    Required. Represents the ID of the engagement signal that’s associated with the metric.

    This field is a relationship field.

  • Relationship name: EngagementSignal

  • Relationship type: masterdetail

  • Refers to: EngagementSignal

  • Type: reference

  • Properties: Create, Filter, Group, Sort

  • Relationship name: EngagementSignalMetricDef

  • Relationship type: relationship

  • Refers to: EngagementSignalMetricDef

  • Type: reference

  • Properties: Filter, Group, Nillable, Sort

  • Relationship name: Field

  • Relationship type: relationship

  • Refers to: MktDataModelField

  • Type: boolean

  • Properties: Defaulted on create, Filter, Group, Sort

  • Introduced: API version 65.0

  • Description:

    Indicates if the engagement signal metric object is owned by a different org in Data 360.

    The default value is false.

  • Default value: false

  • Type: string
  • Properties: Filter, Group, idLookup, Sort
  • Description: Required. Text label that identifies the engagement signal metric.

These derived metrics serve as the core unit of measurement across the personalization platform. Use them to train machine learning models, measure performance in A/B tests, track outcomes in attribution models, and define custom objectives or compound metrics.