Steps: Import Metric Data

Using a Workday-delivered integration, you can:

Import your organization’s historical data directly into Workday for Workday to use in forecasts.
Import predictions from an external system instead of relying solely on Workday's internal forecasting models.

To forecast Labor Optimization metric data, you must first import historical sales data using the Workday-delivered integration. For increased forecasting accuracy, Workday recommends importing 2-3 years of historical data. The delivered integration should be configured with a Document Retrieval step with the desired transport type. Example: SFTP.

The file format must be comma-separated (CSV) with a file size no larger than 1GB. To minimize the file size and overall file count, you can compress the CSV into a ZIP archive.

If you haven’t previously imported historical data (Example: scheduling organizations, start date, end date, metric), you can also remove any rows where the value of the historical sales data is zero. However, to change a non-zero value to a zero, you must send a row with the unique key listed above, and send a zero.

The historical sales data must be aggregated in time intervals equal to the Labor Demand Period Length (15, 30, or 60 minutes).

In each row of data, you must convert the time interval start times and end times to the local timezone of the corresponding scheduling organization.

Example: An interval with a 9:00AM start time formatted for 3 scheduling organizations in 3 different timezones:

British Summer Time (BST): 2019-01-01T09:00:00+01:00
Eastern Daylight Time (EDT): 2019-01-01T09:00:00-04:00
Pacific Daylight Time (PDT): 2019-01-01T09:00:00-07:00

The CSV file must contain these columns:

Column	Description	Notes
scheduling_organization_id	The reference ID of the scheduling organization.	You can't use the WID or any other identifier. Enter the reference ID for the scheduling organization (not the high-level scheduling organization). These reference IDs cannot change between tenants during migration.
subgroup_organization_id	The reference ID of the subgroup organization.	The column header is required but values must be left blank to support an upcoming enhancement.
interval_start_datetime	The start timestamp for the aggregation interval.	The datetime format is: YYYY-MM-DDThh:mm:ssTZD Example: 2022-08-31T09:00:00-07:00 The datetime must be converted to the corresponding scheduling organization’s local timezone.
interval_end_datetime	The end timestamp for the aggregation interval.	The datetime format is: YYYY-MM-DDThh:mm:ssTZD Example: 2022-08-31T09:30:00-07:00 The datetime must be converted to the corresponding scheduling organization’s local timezone. The interval between the interval_start_datetime and interval_end_datetime must be equal to the Labor Demand Period Length.
metric_id	The reference ID of the Labor Optimization metric you plan to forecast.	Labor Optimization metrics are configured in your tenant and assigned default reference IDs on the Maintain Labor Optimization Metrics task. You can update the default reference IDs to represent the Labor Optimization metric name. The reference IDs are case-sensitive and the maximum character limit is 32. These reference IDs cannot change between tenants during migration. See: Maintain Labor Optimization Metrics
metric_value	The aggregated value of the Labor Optimization metric for the time interval between the start/end date/times.	Examples: 10, 100.50. Negative values aren’t supported. Don’t include rows if the metric_value is zero.