Datasets Section

Overview

A dataset in Discover is based on a single Sugar module. The dataset is comprised of all the fields from the module. Data is updated on the current snapshot continuously from Sugar. A new snapshot is created at the start of each day leaving previous snapshots for a historical record. The Datasets section in Discover provides access to edit and configure the dataset's fields and calculations.

Prerequisites

To view the Datasets section, you must have access to Discover and have a Discover role of Admin or User. To make any changes to a dataset, you must have a Discover role of Admin.

Datasets Section Interface

To access the Datasets section, click the Datasets tab in the navigation bar. Any datasets linked to Discover are listed. If the Last Refresh column says "Ingesting", your Sugar data is being continually loaded as changes are made.

To edit the dataset, click the three-dots menu to the right of the dataset name.

The following options are available for a dataset: 

Special Dataset Fields

There are special fields defined in each pre-configured SugarCRM dataset allowing you to build certain types of historical comparison reports or analyze data related to stage transitions within opportunities. Understanding these fields is important to understand so that you avoid creating incorrect or misleading results.

Snapshot Date

Discover's technology continually looks for changes at the database level and refreshes Sugar datasets. A new snapshot of the database is created between 12:00 a.m. and 1:00 a.m. in your instance's time zone. The date and time of the snapshot are stored in a field called "Snapshot Date". This field is useful when creating reports in which you want to visualize and compare the state of records at different points in time.

For example, to show the stages of the opportunities over the past two calendar weeks, you would need to use the "Snapshot Date" for the x-axis and filter the data by the snapshot date. This will show how many opportunities were in each sales stage on the specific snapshot date. 

Current Snapshot

While the Snapshot Date field captures the date and time a snapshot was taken, the Current Snapshot field tracks which snapshot represents the most current data. When creating a new report, a filter with "Current Snapshot = 1" is added to the report by default so that your default state is to view only the current state of the records. To create a report that compares different states (snapshots) of records or requires access to all snapshots of data (i.e., funnel visualizations), you must manually remove the filter by clicking the x next to the filter in the Filtered By panel.

Stage Transition

The Stage Transition field is used to track which snapshots in your data represent a transition from one stage to another. For example, as Opportunities progress through the sales process, the Stage Transition field tracks when the Sales Stage field changes. This field is useful in cases where you are performing calculations specifically between stage transition events. For example, you want to calculate the time an opportunity spends in a given stage by comparing the Snapshot Date when the opportunity transitioned out of the stage from the date in which it transitioned into the stage.

Dataset Field Configuration

To customize a dataset's fields, select "Field Configuration..." from the three-dots menu next to the dataset to open the Field Configuration section.

Field List

The field list displays the following attributes for each field:

To add a new calculated field to the dataset (i.e., a dataset-level calculated field) that can be used for all reports, click the Add Calculated Field button in the top right corner of the screen. For more information on calculated fields, refer to the Reports Section documentation.

Search and Filter Field List

At the top of the Field Configuration screen is a search bar for you to easily search for a field of interest. You can also choose to filter the fields list to narrow your search by clicking on an icon to the right of the search box. Click the icon again or click the Clear button to remove the filter.

Option	Description
	Show all fields with a format of String.
	Show all fields with a format of Number.
	Show all fields that are calculated fields.
	Show all fields with a format of Timestamp.
	Show all hidden fields.
Clear	Clear the filters.

Field Actions

When you hover over a field in the list, there are several field actions to the right of the last column. The actions depend on whether the field is a Sugar field or a calculated field. 

Option	Description	Availability
	The field is currently hidden. Click to make the field visible in the Query Layout panel.	Sugar and calculated fields
	The field is currently visible. Click to make the field hidden in the Query Layout panel.	Sugar and calculated fields
	Preview the data in the Data Preview panel.	Sugar fields
	Edit the calculation for a dataset-level calculated field. For more information on defining calculated fields, refer to the Reports Section documentation.	Calculated fields
	Delete the calculated field. This action cannot be undone once you save your changes. Any reports that contain references to the deleted calculated field will be affected.	Calculated fields

Once you have made changes to your dataset, click an option at the bottom of the screen to save your changes:

Any changes you save cannot be undone.

Dataset Settings

To customize a dataset's settings, select "Dataset Settings" from the three-dots menu next to the dataset to open the Dataset Settings screen. In the Dataset Settings screen, you can configure the fiscal calendar and funnel configuration used by your Discover tenant.

Fiscal Calendar

While a calendar year always begins on January 1 and ends on December 31, a fiscal year can start and end at any point during the year as long as it comprises a full 12 months. A company that starts its fiscal year on January 1 and ends it on December 31 operates on a calendar year basis. Some companies opt to follow a fiscal year instead of a calendar year because their fiscal year better fits their natural business cycles. For example, ending the fiscal year in January gives retailers a more accurate financial picture to report. Seasonal businesses also tend to use fiscal years for accounting purposes.

Discover makes reporting based on a fiscal calendar easy by relying on the dataset to provide the rules for the fiscal calendar year.  Once enabled, users can simply enable this functionality on the report to apply the Fiscal Calendar to the following component areas: Date Filters, Calculated Fields, Visualizations, and Discovery Insights. 

In the Fiscal Calendar section, choose the fiscal calendar year start date using the date picker. While it is typical for the first day of the month to be chosen, any day can be used as the start date. Then choose whether to display the fiscal year based on the start or end of the month. For example, if your fiscal year is chosen to begin on February 1, 2023, your fiscal year spans from February 1, 2023 to January 31, 2024. If you choose to display the fiscal year based on the start month, the first fiscal year will be reported as 2023. If you choose to display the fiscal year based on the end month, the first fiscal year will be reported as 2024. Click "Apply Changes" to save changes without exiting, or "Save Changes and Exit" to save changes and return to the Datasets section.

Funnel Configuration

The funnel visualization contains unique calculations such as Average Cycle Time and Average Days in Stage. These calculations rely on the date that a record is created and closed, as well as the ID. By default, the field called Date Created is used for the created date, the field called Expected Close Date is used for the end or close date, and the field called ID is used for the record ID. If you would like to base calculations on different fields, you will have to edit the funnel configuration settings.

In the Funnel Configuration section, select which field to use for each of the values. If the dropdowns display "Select", it means the default values are being used.