Comparative analysis using the filtered measure plugin

[AssafHanina]

Comparative analysis using the filtered measure plugin

By default, filter selections in Sisense are applied at the dashboard or widget level. The Filtered Measure add-on enables the creation of additional selection states, allowing you to display two or more different sets of selections within a single widget. For more details, please refer to the link below.

Here are some common use cases for Filtered Measures:

1. Comparative Analysis

   • Example: Compare the same metric (e.g., Revenue) across two different date ranges (e.g., this month vs. last month) to analyze trends or identify growth patterns.

2. Filter Applied Directly on a Measure

   • Instead of applying a filter to the entire widget or dashboard, apply the filter directly to a specific measure (e.g., applying a filter to "Sales" based on a particular product category, without affecting other parts of the dashboard). This allows for more granular control over the displayed data while preserving other elements of the analysis.

This article highlights a comparative analysis of a single measure, Revenue, across multiple dimensions—Date and Category. It incorporates two filters for each dimension: Primary and Compare.

Background

When it comes to comparative analysis, dashboards are often designed to display measures broken down by the dimensions being compared. This approach can make it challenging to extract specific values and may require adding additional dimensions for grouping.

Example Dashboard

This dashboard currently does not address the following questions:

• How can we filter by Category and compare Bikes to Wheels & Wheelsets across different Regions?
• How can we evaluate the performance of the Category Bikes against Wheels & Wheelsets over time?
• How can we compare a specific group of members within a Category to the overall total of the Category?
• How can we compare specific date ranges, such as Yesterday versus the same day last week? 
• How can we provide flexible date selection options tailored to the user’s preferences?

Filtered Measure - Implementation Steps

1. Enable the Measured Filter Plugin
   Ensure the Measured Filter Plugin is activated in your environment.
2. Modify the Data Model
   - Navigate to the Data Model.
   - Locate the Dimension table that requires a comparison
   - Duplicate the column you want to compare (e.g., the Date column) and assign it a meaningful name, such as Date_compare
   - Build or Publish the data model after applying the changes
3. Prepare the Dashboard
   - Create a new dashboard or update an existing one.
   - Add both filters - Primary and Compare - as Dashboard Filters
   - Maintain a clear naming convention for the filters (e.g., Date for the primary filter and Date_compare for the comparison filter).
4. Update Widgets with Measure Filters
   - Create a new widget or update an existing one
   - Add the relevant measure filters (Primary and Compare) to the metric. For example:
   
   
   
   
5. Rename Filter Measures
   - Rename the filter measures by adding an @ symbol before the name.
   - Update the formula syntax to include the renamed filters. Example
   
   
   
   - Ensure that the filter name in the dashboard matches the Measure Filter name used in the formula.
   
   - Repeat the Same for the Compared measure
   
   
   
   

Key Points of the Filtered Measure Plugin

1. The @ symbol is required in the formula to apply the dashboard filter as a parameter. However, it should not be added to the name of the dashboard filter itself.
2. Grouping by the Compared Dimensions is unnecessary and adds no value.
3. Filtered measures support multiple measures and dimensions
4. In the same formula, it is possible to include measures for both the Primary measure and the Compare measure. For example:
   
   
   
   
5. For Indicator, Blox widgets, and any widget using a single measure applied to either Primary or Compare, disable the corresponding filter. Similarly, turn off the other filter when switching between them.
   
   
   
   

6. If both measures, Primary and Compare, are used, ensure that both filters are enabled.

Here are some common examples of comparative analysis:

Use Case 1 - Compare Category Dimension

This example compares between Category Bikes and Wheels & Wheelsets over 3 groups by dimensions:

1. Displaying the Revenue asan  Indicator for each one of the Categories.
2. Compare the Revenue over time 
   
   
   
   
3. Compare the Revenue by Region
4. Summary Table of Revenue Comparison by Employee 

Use Case 2 - Compare multiple Dimensions: Date and Category

This example compares Dates and Categories.

The implementation of the measured filter is the same as shared above: (sum([Order Revenue]),[@CategoryName],[@Date])

1. Enable dynamic comparison between different time ranges from the viewer's side, rather than being predefined by the dashboard designer.
2. Compare Category and Date over time 

Use Case 3 - Compare Different Date Ranges with Different granularities

Compare Year to Single Quarter Example

Compare Different Custom Date Ranges Example

Use Case 4 - Compare the selected Category with the total, excluding the selected category.
This allows users to view the performance of a specific category in contrast to the remaining categories

Best Regards

Assaf