Schemas

To manage schemas, supporting the data you want to ingest in Experience Platform and use in Mix Modeler:

1. Go to the Mix Modeler interface.

2. Select Schemas, underneath SETUP.

See the Schemas UI overview for more information.

Aggregate or summary data

It is highly recommended to use the XDM Summary Metrics class as the base of the schema underlying any aggregate or summary data you want to ingest in Experience Platform and use in Mix Modeler.

Use the XDM Summary Metrics class for:

• walled garden data, for example data from Facebook or YouTube.

• external factors data, like data from SPX (S&P 500 stock price indices), weather data,

• internal factors data, for example price changes, a holiday calendar.

A schema using the XDM Summary Metrics base class can be simple, as shown in the ExternalFactorSummarySchema below.

This simple schema can be used to ingest datasets containing data for, for example:

• Competitor index data

  timestamp	date_type	factor	value
  2020-11-28T00:00:00.000Z	week	competitor_index	289.8
  2020-12-05T00:00:00.000Z	week	competitor_index	291.2
  2020-12-12T00:00:00.000Z	week	competitor_index	280.07
  …	…	…	…

• Public holiday data

  timestamp	date_type	factor	value
  2020-11-28T00:00:00.000Z	week	all_holidays_flag	0.0
  2020-12-05T00:00:00.000Z	week	all_holidays_flag	0.0
  2020-12-12T00:00:00.000Z	week	all_holidays_flag	0.0
  2020-12-19T00:00:00.000Z	week	all_holidays_flag	0.0
  2020-12-26T00:00:00.000Z	week	all_holidays_flag	1.0
  …	…	…	…

See below for a more comprehensive example of a LumaPaidMarketingSchema using the XDM Summary Metrics as the base class. The schema uses dedicated field groups (annotated with colors) for metrics (AMMMetrics), dimensions (AMMDimensions), and other customer-specific information (CustomerSpecific).

Given the asynchronous nature of profile ingestion, when collecting aggregate or summary data from external sources, it is encouraged to use the External Source System Audit Details field group as part of a schema. This field group defines a set of audit properties for external sources.

Supported data types

Currently, Mix Modeler does support a subset of Experience Platform data types. The following basic data types (fields), mentioned in Basics of schema composition, are supported:

• String
• Integer
• Double
• Boolean
• Long
• Short
• Byte
• Date
• Date-time