Configuring Data Actions

Jen, the planner at ABC Corporation, is working with her team to create a forecast income statement for 2025. But first, she has been asked to calculate projected labor expenses for that year.

A data action is a flexible planning tool for making structured changes to planning data in SAP Analytics Cloud. Modelers design the data actions and planners run them in stories or schedule them in the calendar.

A data action is created based on a planning model and consists of one or more steps that are carried out on a public or private version.

Data actions are used to process mass amounts of data.

Some use cases for data actions include the following:

• Carry out planning tasks automatically
• Copy by model and across models
• Allocations
• Currency translation
• Calculations

Some features of data actions include the following:

• Sequence data action steps
• Embed data actions (just like an include)
• Schedule or ad-hoc execution
• Include prompts for flexibility
• Integrate prompts and story filters

Data actions can be created from the Navigation Bar or from Files.

Recent data actions will also show in the Recent Files section of your home screen.

• Steps in a data action are executed sequentially.
• The result of one step is available for the subsequent steps.
• For each step, the execution scope is defined via filters.

The following types of action steps are available:

• Copy action: copies data within a model based on a set of rules, filters, and aggregation settings. For example, you can use it to copy data from 2018 to 2019 or Actual data to Budget data. With the Copy action, you copy a value from only one dimension member, but you can copy to multiple dimension members.
• Cross-model copy action: allows you to copy data from a different source planning model based on a set of filters and automatic or manual mapping between dimension members.
• Each step of an advanced formulas action allows you to create calculations to apply to the data. For example, you can use scripting to calculate opening and closing balances of headcount based on hires and terminations, looping through the data by time.
• Allocation action: creates a new allocation or use an existing allocation step.
• Embedded data action: runs another data action as part of the one you’re working on. Combining these steps with dynamic parameters lets you reuse data actions with different source or target members.
• Conversion data action: is used for models with measures, conversion steps let you copy between measures while applying currency conversion.

Setting parameters result in prompts when they're run in a story. Prompts are like variables, in that they make the data actions flexible and dynamic.

When a data action is executed from a story, as in the following example, the parameters related to the data action prompt you for source and target data.

Parameter Values are determined by:

• Member selection in the story
• Default values in the data action parameter definition
• A story filter
• An input control

Parameters Type can be based on members or numbers:

• Member: Creates a parameter that represents a dimension member. For example, this type of parameter lets you change the year that you're copying data to or from.

• Number: Creates a parameter that has a numeric value. For example, this type of parameter is added to advanced formulas, and can let you prompt users for values such as growth rate or gross margin percentage.

The Model assignment:

• Can be set to the Default Model. In this case, the model is inherited from the data action header.
• Can be set to a specific model. In a copy step, this can only be the same model as in the data action header. In a cross-model copy step, this could pertain to either the source or target model.

Data actions can be executed in multiple ways:

1. From a story.
2. From an extended story.
3. From the SAP Analytics Cloud calendar.
4. From SAP Analytics Cloud, Add-in for Microsoft Office.

In this section, we’ll focus on using the SAP Analytics Cloud calendar to schedule the data action as a task.

Using the calendar, you can schedule a data action. If a data action takes a few minutes to run, you can schedule it to run at night.

Before the data action runs, its status will be Open.

After it runs with no issues, the status will be Successful.

If something goes wrong, the status will be Failed.

When scheduling a data action, prompts must have assigned parameter values.

Data actions can be used in other data actions that belong to the same model.

For example, you create a data action that calculates subscription revenue based on booking data, with a dynamic parameter for the product type. Using embedded data action steps, you can reuse that data action multiple times in another data action and you can specify a different product type each time.

This lowers the cost of development and maintenance.

Here are some other features of embedded data actions:

• Parameters of outer data actions can be linked against required parameters of inner data actions.
• Parameters of inner data actions aren’t prompted on execution; each must be assigned a value in the designer.
• You can’t use a data action within itself.

When data actions are scheduled from the SAP Analytics Cloud calendar or executed from a story, you can use the Data Action Monitor to check the status.

The Data Action Monitor shows:

• When the data action was triggered
• Who triggered the data action
• Where the data action was triggered
• Which values were used as parameters
• Number of records and duration
• The data action job history will remain based on a retention period of 30 days (default)

Copy actions allow you to copy data from one set of members to another, specifying options for filters and aggregation. For example, you can copy data for a product in one region to another product in other regions.

In the following example, the data action is set up to copy data by version and year at the same time. For example, copy the 2023 actual data to the 2024 plan data.

The copy rule doesn't allow the use of version, so there are two key points to remember:

1. Use a parameter for the source version. The source version parameter is then added to the context which will result in a prompt.
2. Every data action will always prompt for the target version regardless of whether the target version prompt is in the filter or not.

Every data action has a system-provided Filter for Version. The default version filter is the target version prompt.

• When choosing members for the copy action settings, you can use a parameter instead of dimension members to make your data action more flexible. Parameters can make it easier to update your data action, and they prompt you for values while running the data action.
• When you define a copy action, the source distribution aggregation will also be copied. You can override this functionality by specifying any dimensions that need to aggregate data differently than the source.
• When a copy action is executed, the default action is to Overwrite the data. When defining the copy action, you have the option to change to Append (add) the data to existing values for the target members.
• Depending on the Filter and Aggregate To settings, you may not need to add a copy rule:
  • If you want to copy the 2025 Forecast to a target version, then you only need to filter the step to the Forecast version and 2025.
  • If you want to copy the 2025 Forecast to 2025 and 2026 in the target version, then a copy rule is required.

In the Aggregate To area of a data action, add any dimension where you want to override the default disaggregation behavior.

For dimensions that aren't included in the Aggregate To area, the copied data will match the distribution of the source data by default. This behavior is similar to copying and pasting a cell with the Paste Details option enabled.

For example, if you want to remove the distribution of data to different customers for privacy reasons, you can aggregate the data to the Unassigned member of the Customer dimension. In this case, all of the copied data is booked to the Unassigned member.

To copy data from one year to another and at the same time copy actual to plan:

• The version filter must be set to the source version, for example, actual or a parameter for the user.
• The copy rule would then contain the source and target time values or parameters.

When Data Action Triggers are added to stories, they appear as push buttons and can be configured using the right panel.

The prompts are based on the data action parameters. In a story, these are the input options:

• Prompt
• Fixed Value

If you choose the Prompt type, there are four Value options:

1. Member Selection
2. Default Value (based on the default value set in the data action configuration)
3. Story Filter
4. Input Control

When a data action is run by selecting the push button in the story, a prompts dialog allows for dynamic member selection.

If the data action is set to always run in the background, or if you select Run in Background  after starting it, you'll return to your story or extended story. You can keep working with your story or extended story but you'll have to wait until the data action finishes to make changes to the same version and you may need to refresh to see the results.

A message appears when it’s complete or you can check the  Notifications list for the results.

Data actions can be set up in stories to publish automatically, so that planners don't have to publish data in a separate step. Alternatively, if a data action is set up in a story not to publish automatically, then the data can be reverted.

There are three publishing options:

1. Do not publish.
2. Publish and fail if there are warnings.
3. Publish and ignore warnings.
   Note

   If there is an issue with some records but not all, this option will process the good records.

Data is often needed from several sources to perform the planning process, for example, prices are in one model but the income statement data is in another. Depending on the circumstances, the data may need to be physically copied while in other cases, a look-up will suffice.

Before we get into the specifics of copying data between models, let's consider the options when working with data in multiple models:

1. Cross Model Copy: Data Duplication:
   • The data needs to be physically copied from one planning model to another.
   • The data will be modified in the target model for example.
   • A snap-shot of the data is needed.
   • Example: Copy actual labor hours from HR as the basis for forecast hours.
2. Combine Data in the Import Process: Data Duplication:
   • The data needs to be physically merged.
   • Do this in data wangling.
   • Example: combine transaction data with exchange rates for immediate conversion.
3. No Data Duplication:
   • Blend data in stories. Example: Add attributes from another source.
   • Use advanced formula look-ups via the LINK command.
   • Example: Look up labor rates from HR for the labor calculation.

In the Mapping area, set how the data will be mapped from the source model to the dimension members of the target model.

For each of the target dimensions listed in this area, you need to either map a source dimension or select a default value. In the following image, you can see that the currency is set to Unassigned. In this example, labor hours are being copied from an HR into an expense model. In the HR model, labor hours aren’t related to currency.

You don't need to map source dimensions that aren't relevant for the target model. For example, you might choose not to map gender or office location from the Headcount model to the Finance model.

The source and target models can have different dimensions. For example, HR has personal area but Expense doesn’t. In the data action, a constant can be used for personal area.

DATA([d/Measures]="Quantity") = LINK([AccountModel], [d/Version]="public.Forecast", [d/Account]="Quantity")