Advisor Recommendations

This page lists all the recommendations found in Advisor, and for each provides further information as well as links to related topics in the Prophix Help.

Model Manager

This section contains recommendations for the following:

• Formulas on Leaf Member Combinations
• Account Members
• General Members
• Version Members
• Calculated Versions
• Versions with Carry Forward From
• Standard Formula
• Procedural Calculation
• Rule Sets
• Time Perspective
• Number of Dimensions

Formulas on Leaf Member Combinations

Advisor has determined that you have a significant number of formulas tied to leaf member combinations and that this may be affecting performance.

Recommendation

• Convert these formulas either to procedural calculations or to template formulas.

• The fewer the leaf combination formulas, the more efficient your cube will perform.

    Tip:  To help identify the accounts that contain leaf member formulas, run an Export Dimension process for the Accounts dimension and look at the CalculationMethod column for entries that contain (Applied to leaf member combination only).

See also

• Account Calculations
• Export Dimension

Account Members

Advisor has determined that you have a significant number of Account members and that this may be affecting performance.

Recommendation

• Review and delete any inactive accounts that are not required for current reporting purposes.

• If the number of accounts has significantly increased over time due to combining your Chart of Accounts with other dimensions(for example, Company, Department, and so on) you may want to re-think a more sustainable long-term solution.

    Example:  Consider creating a centralized Chart of Accounts and mapping the data in from the source system(s) and/or from feeder Cubes.

• The fewer the Account members, the more efficiently your cube will perform—you might want to consider deleting account members.

General Members

Advisor has determined that you have a significant number of members and this may be affecting performance.

Recommendation

• Identify any members that may be inactive and are candidates for deletion.
• Review and determine whether the current level of granularity is needed for reporting or planning.
• Can detail be shifted to a drill across A data analysis function that obtains the supporting data for a single data cell or chart element from outside the cube (for example, from a relational or other data source). and be summarized at higher levels for reporting and for use in Ad Hoc Analysis?
• The fewer the members, the more efficiently your cube will perform—you might want to consider deleting members.

Version Members

Advisor has determined that you have a significant number of Version members and that this may be affecting performance.

Recommendation

• Delete any Versions that are no longer required for current reporting purposes (for example, historical plans or forecasts).
• If historical data needs to be archived, consider exporting it to a flat file or the staging database for backup.
• The fewer the Version members, the more efficiently your cube will perform—you might want to consider deleting members.

Calculated Versions

Advisor has determined that you have a significant number of Calculated Versions and that this may be affecting performance.

Recommendation

• Identify whether the Calculated Version is needed for analysis in Ad Hoc Analysis or in Templates.
• If a Calculated Version is only needed in templates, consider using template formulas instead and delete the calculated version.
• The fewer the calculated versions, the more efficiently your cube performs.

Versions with Carry Forward From

Advisor has determined that you have a significant number of Versions using the member property Carry forward from and that this may be affecting performance.

Recommendation

• If possible, instead of the Carry forward from property, create a carry forward Version by combining actuals and plan/forecast data using the Infoflex process.
• The fewer the Versions using the carry forward function, the more efficiently your cube performs.

Standard Formula

Advisor has determined that you have a significant number of standard formulas and that this may be affecting performance.

Recommendation

• Consider modeling standard formula accounts using template formulas.
• The fewer the standard formulas, the more efficiently your cube performs.

Procedural Calculations

Advisor has determined that you have a significant number of procedural calculations and that this may be affecting performance.

Recommendation

Consider using template formulas as an alternative to data entry triggers or calculation processes.

Rule Sets

Advisor has determined that you have a significant number of rule sets and that this may be affecting performance.

Recommendation

• Review your rule sets to see whether their underlying reasoning and logic are still needed.
• The fewer the rule sets, the more efficient your cube will perform—remove any unneeded rule sets.

Time Perspective

Advisor has determined that you have a significant number of Time Perspective members and that this may be affecting performance.

Recommendation

• Review your Time Perspective dimension and delete any inactive members.
• Review your calculated Time Perspectives to determine if they can be modeled within the Version dimension instead.

Number of Dimensions

Advisor has determined that you have a significant number of Dimensions in the Cube and that this may be affecting performance.

Recommendation

Redesign your cube to remove redundant dimensions by considering the following:

1. Are all your dimensions required for reporting and/or planning within Prophix?

   Look at this from the end user’s perspective: will all the different page selections ultimately confuse users when they’re trying to navigate to their dataset?

2. Can any of the detail data be shifted to a drill across A data analysis function that obtains the supporting data for a single data cell or chart element from outside the cube (for example, from a relational or other data source).?

3. Can any natural relationships be established in order to combine existing dimensions?

     Example:  

   You have a revenue cube with Product and Customer dimensions. However, if a specific product(s) only belongs to a single customer, then consider doing one of the following:

   • Building a single-customer dimension with a Product > Customer hierarchy
   • Using alternate hierarchies by tagging each Product with a Customer member property

4. Does the number of invalid data combinations far exceed the number of valid ones?

   This is another reason to combine dimensions.

     Example:  

   You have a revenue cube with Product and Customer dimensions.

   If a specific product(s) usually belongs to a single customer (with some exceptions where products span multiple customers) then you should still consider combining these dimensions with a Product > Customer hierarchy.

   You can ensure these products keys are unique by combining them with the customer key.

Detailed Planning Manager

This section contains recommendations for the following:

• Scenarios
• Objects
• Calculations
• Allocations
• Attributes

Scenarios

Advisor has determined that you have a significant number of scenarios in your model and that this may be affecting performance.

Recommendation

• Review data in each completed scenario and verify that the data is still needed, particularly if it has already been posted to the cube.
• Delete any completed-and-posted historical scenarios.
• If there are historical objects with data in the cube that you need to retain, then also turn on Keep Legacy Data.
• The fewer scenarios that you have, the more efficiently Detailed Planning Manager performs.

Objects

Advisor has determined that you have a significant number of objects in your model and that this may be affecting performance.

Recommendation

• Review the detailed planning objects listing to ensure that all the objects are needed and relevant for the scenarios.
• Review whether the current level of granularity is needed for planning: Can detail be summarized at higher levels for planning and reporting? (For example, planning by Employee versus Job Position.)
• The fewer objects, the more efficiently Detailed Planning Manager performs.

Calculations

Advisor has determined that you have a significant number of calculations in your model and that this may be affecting performance.

Recommendation

• Review your detailed planning calculations to determine whether they can be optimized to reduce the total number of calculations.

    Example:  We generally recommend applying calculations to All scenarios and using the Comment field to document any historical assumptions. We also recommend referencing lookup tables when using tax type calculations, rather than duplicating and creating excessive calculations.

• The more optimized your calculations are, the more efficiently Detailed Planning Manager performs.

Allocations

Advisor has determined that you have a significant number of allocations in your model and that this may be affecting performance.

Recommendation

• Examine the current allocation logic and consider whether allocations can be performed in the cube instead.
• Doing allocations in the cube allows you to import your allocation rates and run a process (for example, Infoflex) to perform the allocation.

See also

• Use case: Basic allocations with InfoFlex
• Use case: Complex allocations with InfoFlex

Attributes

Advisor has determined that you have a significant number of attributes in your model and that this may be affecting performance.

Recommendation

• Determine whether all attributes present are necessary and delete any unnecessary ones.
• The fewer attributes present in the model, the more efficiently Detailed Planning Manager performs.

Fact Table Analysis

This section contains recommendations for the following:

• Records
• Zero-Value Records

Records

Advisor has determined that you have a significant number of records in your model and that this may be affecting performance.

Recommendation

• Do the following:

  1. In your data source file, make sure to summarize your records following the example below prior to running the Import process (this reduces the size of the fact table).

       Example:  

     Unsummarized data:

     Account	Time	Time Perspective	Version	Value
     100	2022M01	BASE	ACT	100
     100	2022M01	BASE	ACT	200
     100	2022M01	BASE	ACT	300
     100	2022M01	BASE	ACT	400

     The same data summarized:

     Account	Time	Time Perspective	Version	Value
     100	2022M01	BASE	ACT	1000

  2. Set up an Optimize Fact Table process that creates an Index, compresses the fact table, and removes orphaned records.
  3. Schedule this Optimize Fact Table process to run on a nightly basis.

  4. Review processes you use such as Infoflex or Procedural Calculations that might be running and storing excessive data in your fact table.

       Example:   You might have these processes storing rates across all dimension combinations for the purposes of calculation; however, there are much more efficient ways of modeling these requirements

• The fewer the records in the fact table, the more efficiently your cube performs.

Zero-Value Records

Advisor has determined that you have a significant number of zero-value records in your model and that this may be affecting performance.

Recommendation

• After running your import or other processes that might be storing zero values, to reduce the size of the Fact Table, run a Remove Data process.
• The fewer zero-value records that you have, the more efficiently your cube performs.