Running Smart Discovery

You can run Smart Discovery to gain insights into how underlying variables influence your data in a specific context.

You can launch Smart Discovery when you start a new story or open a story containing the dataset you want to explore.

There are two sections in the Smart Discovery side panel:

Topic: you can configure the following settings:
- Target
- Entity
- Measure
- Filters
Advanced Settings: you can select a Version or configure Included Columns (measures, accounts, and dimensions) in your Smart Discovery.

How to Launch a Smart Discovery

You can launch a Smart Discovery from the Stories start page or from the Files page.

From the side navigation, choose (Stories). To create a new story, select From a Smart Discovery. To use an existing story, go to the Recent Files section, where you can see the last 25 files you worked on.
From the side navigation, choose Files. On the Files page, you can search for the story in a public, private, or workspace folder.

Once the story is loaded, change the mode to Edit. When the toolbar is collapsed, select Start of the navigation path Tools Next navigation step Smart Discovery End of the navigation path . When the toolbar is expanded, select Tools (Smart Discovery).

Smart Discovery Topic Settings

In the Topic section of the Smart Discovery panel, you can configure different settings. Target, Entity, and Measure are required, but Filters are optional.

Target

The Target is the data that you want to explore.

Depending on your model type, you can do one of the following:

New model type with an account dimension: select the Target dimension or account from the + Select a dimension or an account list under the Topic area.
Classic account model or a new model type without an account dimension: select the Target dimension or measure from the + Select a dimension or a measure list under the Topic area.

Note

When you select a dimension as your target, the Select Target Group Members for <selected dimension> dialog is displayed with all the dimension members listed under Available Members.

Select the members you want to use as part of the target group definition: they're automatically added to the Selected Members column.

It's not possible to add all members to the Selected Members column, as that leaves nothing for comparison during the analysis.

You must have two groups - one representing the dimension members that belong to your selected target group, and one representing the dimension members that don't belong to your target group.
Select OK to continue.

Tip

Your selection is then represented in the Smart Discovery panel as your selected target group members. Smart Discovery analyzes the characteristics of your data selection, that differentiate the dimension members that belong to your selected target group from the dimension members that don't belong to your target group.

Entity

The Entity defines the dimensions you would like to explore in relation to your Target.

To add an Entity dimension, choose + Select a dimension.

Measure

To add a Measure, choose + Select a measure.

Note

The Measure settings appear only when your underlying model is a new model type with an account dimension.

Filters

You can refine your Topic by adding Filters.

Smart Discovery can handle a finite data size. You can use filters to reduce the data size.

Under Topic, you can select Filters from the + Add filters list.

Note

Currently, you can't run a Smart Discovery on more than one million cells. By filtering the data first, you can reduce the number of cells included in your dataset.

The number of cells is calculated by multiplying the number of measures by the number of records.

Smart Discovery Advanced Settings

You can configure additional settings through the Advanced Settings section of the Smart Discovery panel.

Version

You can use the Version dropdown menu to choose a version.

Included Columns

Use the Included Columns section to specify which measures (for classic account models, or new model types without account dimensions), accounts (for new model types with account dimensions), and dimensions to include in the discovery. By default, all measures, accounts, and dimensions in the dataset are included in the discovery scope.

Measures: displays the Set Measures to Include in Discovery dialog.

Note

The option to select Measures appears only when your underlying model is a classic account model, or a new model type without account dimensions. You can run Smart Discovery for measures associated with the "average", "minimum", and "maximum" exception aggregation types. These measures are included in the Smart Discovery analysis and can be selected as Key Influencers.
Accounts: displays the Set Accounts to Include in Discovery dialog.

Note

The option to select Accounts appears only when your underlying model is a new model type with account dimensions.
Dimensions: display the Set Dimensions to Include in Discovery dialog.

Select OK to submit your selections.

The number of selected measures, accounts and dimensions is indicated as well as a listing of the selected items under Measures, Accounts, and Dimensions.

How to Preview and Run Your Smart Discovery

Preview Your Smart Discovery

To get a brief summary of the data you've selected to explore, choose Preview in the Smart Discovery panel. This will open the Smart Discovery Data Preview dialog.

Run Your Smart Discovery

When you're happy with the data you've selected, select Run to launch Smart Discovery.

You can select Run straight away after defining your target, entity, measure, filters, and advanced settings. If you decide to select Preview first, you can choose Run from the Smart Discovery Data Preview dialog. Once Smart Discovery completes the data analysis, the available Smart Discovery results are displayed under separate pages:

Overview
Key Influencers
Unexpected Values
Simulation

After you have finished analyzing the results from your discovery, you can:

Save the Smart Discovery results as a part of the new or existing story.
Share the story with other users in your organization.