Concept: Planning Agent: Data Exploration

This feature is only available to Planning Agent customers in the U.S., Canada, Europe, and Singapore. For more information about the Planning Agent, contact your Account Executive.

Overview

The Planning Agent: Data Exploration skill enables intuitive, self-driven analysis that transforms manual reporting. This feature uses both Generative and Conversational AI to provide contextual data analysis and insights.

Data Exploration enables you to:

Generate summaries, detect anomalies, identify key trends, analyze variances, and more using open-ended questions or auto-populated suggestions.

Export the generated summaries to slides as a starting point for your presentations.

Review and update the formats of charts generated based on your questions.

Transform a time-consuming manual process into an efficient, guided analysis.

Instantly identify, drill into, and understand the key business drivers behind your data.

Only explore the data that you have access to based on your current permissions and access rules.

Exploring Report Data Using Variance Analysis

The Variance Analysis capability of the Data Exploration:

Transforms a manual, time-consuming process into an automated, narrative-driven experience.

Enables you to not only identify variances but also understand why they occurred. You can instantly pinpoint the most significant drivers across your entire model.

To initiate a comprehensive variance investigation within a matrix report, you can use these options:

Right-click any cell intersection within a

Variance

column and select

Analyze variance

from the context menu.

Open the Ask Workday sidebar and select the

Identify top variances

prompt. By default, Data Exploration finds and highlights the top five variance cells in your report for further analysis.

In Ask Workday:

Use the

Summarize this report

tile to gets an overview of the data.

Use the

Detect anomalies

tile to gets details on unexpected data.

Type free-form text in the

Enter your prompt here

field to ask about specific data. Example: "What’s the total for the 4100 Product Revenue account in FY 2024." See Reference: Data Exploration: Prompt Examples .

Data Exploration investigates the variance across all relevant dimensions in your model and provides:

A relevance ranking for every dimension and attribute based on how much of the underlying data is tagged with that specific dimension or attribute values:

When analyzing data from a parent node in the hierarchy, core dimensions such as Time, Account, and Level are always considered 100% relevant.

All other dimensions are assigned a relevance score between 0 and 1 based on their specific contribution to the data set.

A focused response that includes:

The top five dimensions that have the greatest impact on the variance.

For each of the five dimensions, the top three contributors and their corresponding variance amounts.

A pie chart illustrating the top five dimensions, the top three contributors, for each dimension and a grouping of all remaining values into an Other category.

A generative AI-powered summary explaining the findings. Analysts can can use this summary as a strategic baseline and starting point for their analysis.

Further exploration of variance drivers displaying the full list of relevant dimensions and attributes ranked in order of their impact. You can drill down on these drivers.

To focus your analysis on material fluctuations, you can define specific criteria using natural language prompts. Example: "Set variance threshold to 100,000 and 10%".

Data Exploration:

Understands thresholds expressed as absolute dollar amounts, percentages, or both.

Uses the "and" operator to combine criteria and doesn't currently support "or".

Performs a "greater than" comparison using the absolute value of the variance.

Limitations

Currently the variance analysis capability:

Ignores these elements even if they're present on the report:

Custom and Subtotal calculation elements and all Display-As elements

Data in the split rows for split accounts.

Requires the matrix report structure to at least:

Include one Difference element that compares two individual versions (not a version folder).

Display one valid variance column when the

Suppress columns if all zeros or blank

option is enabled.

Supports:

Both level and account attributes in addition to levels and accounts.

Only single elements for report filters.

Only single values for report parameters that are connected to filters.

Visualizing Data Exploration Response

Use the icons at the top of a response to interact with the visualizations that were provided. This option is only available when the response includes tables or charts.

Option	Description
Change Chart Type	Select different chart formats to display the data. If the charts are rendered big, you might need to scroll within the chart to see all the data. You can export the chart to slides.

You can hover over the points or data in a chart for more details on the displayed data. Click the legend items to filter out specific elements and focus on the data you want to see.

Downloading Data Exploration Response

From the Ask Workday toolbar, click

More Actions

(three vertical dots) to download summaries, charts, and tables to presentations.

Only responses that include charts or tables provide this option.

Option	Description
Export as PowerPoint	Downloads to PowerPoint on your computer. Your computer might prompt you to save the presentation or it might automatically save the presentation. You can change the title, the data, and charts.