Skip to content

Standard Registration Metrics Library

Event Definition

Based on registration_event tag Asemic will generate activity table, one row per active user at the time of their first event. Table is implicitly available with the definition equivalent to:

yaml
table_name: analytics.registration
columns:
  user:
    data_type: STRING
    tags: [entity_id_column]
  time:
    data_type: DATETIME
    tags: [event_timestamp_column]
  date:
    data_type: DATE
    tags: [event_date_column]
  platform: {datatype: STRING}
  country: {datatype: STRING}
  application_version: {datatype: STRING}
  build_version: {datatype: STRING}
  # + other columns taged with `registration_property`

Properties

yaml
properties:
  registration_date:
    data_type: DATE
    can_filter: true
    can_group_by: true
    registration_property:
      source_column: date

  registration_platform:
    data_type: STRING
    can_filter: true
    can_group_by: true
    registration_property:
      source_column: platform

  cohort_day:
    data_type: INTEGER
    can_filter: true
    can_group_by: true
    computed_property: 
      select: 'DATE_DIFF({date}, {registration_date}, DAY)'

  registration_country:
    data_type: STRING
    can_filter: true
    can_group_by: true
    registration_property:
      source_column: country

  cohort_size:
    data_type: INTEGER
    can_filter: false
    can_group_by: false
    computed_property:
      select: 1

KPIs

yaml
kpis:
  registrations:
    select: SUM({property.cohort_size})
    where: '{property.cohort_day} = 0'
    x_axis:
      date: {}

These metrics and properties provide a basic framework for analyzing user registrations and cohorts.