DOCUMENTATION
Sign In

Get access to My SAS, trials, communities and more.

User’s Guide
 

Creating Optimization Input Data

Create Optimization Input Data

  1. On the navigation bar, click Optimization Optimizationthen selectInput Data
  2. Click Add.

    By default, only input data campaign libraries are listed. To display all libraries, click All libraries. To filter the displayed list of libraries, enter information in the Search box and click Start Search.

    Note: Only libraries that are not associated with input data are displayed.
  3. Select a library and click Create input data.
  4. On the Properties tab, perform these actions:
    • Enter desired information in the Description field.
    • Analytical scenario, the default selection, indicates that the best solution is determined by running a full optimization of a scenario that uses the input data.
  5. On the Data Setup tab, select the tables and columns that you want to use. For more information, see Setting Up the Input Data.
  6. On the Customer Attributes tab, select the customer attributes that you want to use as filters for the input data.
  7. On the Orchestration tab, specify the schedule for uploading the optimization input data.
  8. Click Save to save your changes.
See Also

Setting Up the Input Data

Assign Tables from a Library

You must assign the appropriate roles to the tables and columns in a library in order for SAS Customer Intelligence 360 to correctly interpret the input data. On the Data Setup tab, select from the menu each table, column, or setting for the campaign, customer, and history libraries. If there is only one table available for a required or optional assignment, the table is preselected.

The fields that are marked with an asterisk are required. All other tables and columns are optional. Select optional tables and columns only if you are using them in your input data.

Note: Data validation code fails if the underlying data sets use SAS name literals such as "prob to respond"n instead of prob_to_respond or “credit score”n instead of credit_score.

Change Content of Physical Input Data

After a library has been selected for use by an input data item, you might want to change the content of the physical input data. Here are some of the changes that you could make to the physical input data:

  • add new tables to or remove tables from the original set of tables
  • add new columns to or remove columns from the original set of tables
  • change the format of columns in the original tables
  • add observations to or remove observations from the original tables

The first three types of physical changes impact library and table metadata and must be propagated into the library. To propagate, it is recommended to use the Update Metadata feature within Library Manager in SAS Management Console. The updated library metadata is reported and made available to an associated input data item within 10 minutes. The associated input data item and any associated scenarios change to an out-of-sync status, which can be seen by clicking Refresh on the opened input data item. When an input data item is out of sync, it must be republished through the Orchestration page. The user is responsible for recognizing that the addition or removal of observations requires a refresh of the input data item from the physical tables by uploading the data item to the Orchestration page.

Republishing or uploading an input data item performs impact analysis on associated scenarios. Changes to a scenario’s measures, contact policies, limits, and objectives are reflected in the related scenario pages after the republish or upload of the input data item has completed. The Messages page contains information about any scenario elements that might have been removed as a result of physical table changes. At this point, you can make changes to a scenario so that the changed input data can be used. If the scenario has an out-of-sync status, you can republish it in order to optimize it. If the scenario is not out of sync, optimize it by using the Optimize Now button on the Orchestration page. It is also possible to update metadata programmatically by using a SAS program. Contact your SAS representative to learn how to programmatically update metadata.

The fourth type of physical change does not impact metadata and therefore does not require metadata updates (although the physical change could change the distinct set of values for character columns in the original tables). If this is the only type of change made to the physical tables, there are no detectable changes to the metadata even if you complete steps to update the library metadata. The associated input data item and any associated scenarios do not change to an out-of-sync status. However, the user is still responsible for recognizing that the addition or removal of observations requires a refresh of the input data item from the physical tables by uploading the data item to the Orchestration page.

Note: If you delete or re-register table metadata, you must complete these tasks:
  • perform the data setup and customer attributes steps again for an associated input data item
  • re-define measures, contact policies, limits, and objectives for associated scenario objects

This discussion assumes that library metadata is updated, and that it has not been deleted and re-registered.

Working with Multiple Libraries

On the Data Setup tab, you can make table assignments for input data from the campaign, customer, and history libraries, which might exist in different physical locations and might use different libraries.

To select the tables from the campaign library, click Campaign Library Settings.

Note: Only libraries that have the “engage_optimize_campaign” keyword are listed.
Note: Selections for the Customer rank measure and Priority column table settings appear only when the input data is using the priority-based scenario optimization method. For more information about optimization methods, see Create Optimization Input Data.

To select the tables from the customer library, click Customer Library Settings.

Note: Only libraries that have the “engage_optimize_customer” keyword are listed. If the customer library uses the same data as the campaign library, any changes to the campaign library affect the customer library.
Note: The Household ID setting is not available if the input data is using the priority-based scenario optimization method. For more information about optimization methods, see Create Optimization Input Data.

By default, the data for the customer library comes from the campaign library. If you want to use the data from a different library, click Select a Library and select an input data library. If you want to use the same data that is used by the campaign library, ensure that Same as campaign library is selected.

To select the tables from the history library, click History Library Settings. By default, the data for the history library comes from the customer library. If you want to use the data from a different library, click Select a Library and select an input data library. If you want to use the same data that is used by the customer library, ensure that Same as customer library is selected.

Note: Only libraries that have the “engage_optimize_history” keyword are listed. If the history library uses the same data as the customer library, any changes to the customer library affect the history library.

Select the Customer Attributes

On the Customer Attributes tab, you select which customer attributes you want to load into SAS Customer Intelligence 360 for use in customer filters, calculated measures, objective measures, and so on. If your Customer table contains additional identifying information that is unique to each customer or household, such as Social Security number or address, do not load these attributes into SAS Customer Intelligence 360. Character attributes that have thousands or millions of distinct values can severely diminish the performance of the application. Performance is affected because each value of the customer attribute must be stored to be displayed in customer filters.

TipFor optimal performance, load only the attributes that you intend to use within the optimization scenarios.

Use the arrow buttons to move attributes between the Available and Selected panes. To move attributes from the Available pane to the Selected pane, select the attributes and click Add. To move all attributes from the Available pane to the Selected pane, select the attributes and click Add All. When you select all attributes, ensure that you are not selecting character attributes with unique values per customer or household.

To move attributes from the Selected pane back to the Available pane, select the attributes and click Remove. To move all attributes from the Selected panel back to the Available panel, select the attributes and click Remove All.

Note: To ensure that input data is valid, a required customer attribute is included in a constraint or contact policy filter even if you do not select the attribute on the Customer Attributes tab.

Schedule and Start Optimization Input Data

Note: To publish input data, you must be able to edit every scenario that uses the input data. Refreshing input data might take longer than a first-time publish of the input data.

When you finish updating your input data, you can publish the input data immediately or set a schedule for the input data.

Publishing the data affects all scenarios that use the input data. Existing results for associated optimization scenarios are deleted and the scenarios must be rerun. Scenarios that use the same input data are listed on the Optimization Optimizationthen selectScenarios page.

On the Orchestration tab, complete these actions:

  • Click No schedule if you plan to publish the data manually.
  • Click Set a schedule to set a time period that the data is active.
    • Select a time zone. For example, you might want to select a time zone that is different from the default if the input data is targeted to a specific locale. By default, the time zone is the same as the time zone that is set for the tenant. If a valid time zone is not set for the tenant, the default time zone is UCT.
    • Click Single occurrence to specify a single time span that the input data is active.
    • Select a start date and an end date for the input data.
      Note: If you select Single occurrence, the start date cannot be edited after the input data is published.
    • Click Recurrence to specify a recurring schedule for publishing the input data. Click Edit.

      Specify these details for a recurring schedule:

      • Start time is the start time for the first occurrence.
      • Frequency specifies the schedule for the recurrence. If you click Weekly, click the day for each occurrence. You can select multiple days in each week. If you click Monthly, select the numbered day or the weekday for the recurrence. For example, you can select the 10th day of each month or the second Tuesday of each month.
        Note: You can edit the weekly or monthly frequency after the task is published and the start date has passed. The recurrence must occur within the start and end dates.
      • Range of recurrence specifies the start and end dates, or that the recurrence continues without an end date.
  • Click Publish to make the input data active.
    Note: When you publish an item, you are prompted to publish any dependent items that are marked Ready.

    Items with a Ready status can be changed back to a Designing status by clicking Return to Designing. Changing this status to Designing prevents the item from being published.

    Note: If your item has a Ready status and a start date, the item is not automatically published when the start date is reached. For example, if you are creating input data, your item must be published in order for the start date to activate the input data.

Changes to a scheduled or active input data are not automatically propagated to the published version of the input data. You must republish the input data after you save it to resolve the out-of-sync status. Otherwise, the input data stays out of sync with the published version and the updates are not delivered.

Note: Adding, changing, or removing a schedule for an active input data item does not result in the object going out of sync if the scheduling edit was the only edit of the object. Schedule-only changes do not alter the inherent definition of an input data item in any relevant way, so no republishing is necessary.

On the Upload page for published input data, click Upload Now to immediately upload the input data.

Note: Manual upload of an input data item through the Upload Now button is enabled when an object is in Scheduled status. If a manual upload operation is requested when the input data item is in Scheduled status, the object returns to Scheduled status at the conclusion of the manual operation.

If the schedule triggers and attempts to execute a scheduled operation while manual upload of an input data item is in process, the scheduled operation is blocked and does not execute. If the schedule is non-recurring, the object does not return to Scheduled status at the conclusion of the manual operation. Instead, the object transitions back to Designing or Active status depending on its status when it was scheduled. If the schedule is recurring, the input data item remains as Scheduled because the most recent attempt could not be executed. Lastly, if the schedule triggers and attempts to execute a scheduled operation while the input data item is locked (for example, by another user), the scheduled operation is blocked and does not execute.

Occasionally, input data might remain in the Publishing status for an indefinite amount of time. The input data could appear unresponsive because the upload is taking a long time—for example, if the underlying input data is very large. It is also possible that the input data appears unresponsive because the agent is down, an external event terminated the upload process, or communication with the agent is disrupted. The upload process might have stopped running on the optimize agent. If you suspect that this is the case, click Activate the Non-Responsive Item to move the status of the item from Publishing or Optimizing to Active, with an error. When you click Activate the Non-Responsive Item, the following action occurs:

  • If the optimization process is currently running on the optimize agent, the process finishes running.
  • The input data is returned to the Active status, with an error message. The error is cleared when the item is published again.
  • If the upload process was still running and finishes, the input data returns to the Active status.

View Input Data Usage Information

On the Orchestrationthen selectUsage page, you can view which items use the input data and the items that the input data uses.

The ability to view dependencies between items enables you to see the items that input data depends on to work successfully. Usage information helps you identify how changes that you make to one item might affect another item.

Items That Use the Input Data

The first section shows the items that use the input data. For example, if the input data is used in a scenario, the scenario is included in the list.

The Status column displays the status of the items that use the segment.

Items That the Input Data Uses

The second section shows items that the input data uses.

  • Click Directly Related to list the items that the input data uses.
  • Click All Related to view directly related items as well as any items that those directly related items use. For a segment, the list of directly related and all related items is the same.

    The list displays items with a Designing, Ready, or Active status. Click Show Designing Status only to display only items with a Designing status.

To open an item, click the item. To view item details and to open the item, click View Item Details in the item’s row.

Validate the Optimization Input Data

To quickly validate optimization input data to identify any problems that might occur when you upload the data:

  1. On the navigation bar, click Optimization Optimizationthen selectInput Data.
  2. Open the input data.
  3. Click Save to save any changes that you made to the input data.
  4. Click Validate Input Data.

    SAS Customer Intelligence 360 validates the input data tables against each other to ensure that the data conforms to the required structure. If the input data is valid, a validation confirmation window appears. If the input data is invalid, a window appears that displays a list of error messages.

    Note: The validation messages are not sent to the data store and the statuses of any scenarios that use the input data are not changed.

Rename Optimization Input Data

Select the input data in the table on the Optimization Optimizationthen selectInput Data page, click Actions, and select Rename.

Last updated: August 11, 2026