Support Native SQL LIKE / NOT LIKE Wildcard Filtering Across Power BI Reports
Description
Currently, Power BI does not provide a native way for report consumers to perform dynamic SQL-style wildcard filtering using patterns such as
LIKE '____' - Above return all values with exactly four characters
LIKE '__A_' - Return all four-character values where the third character is "A"
Many enterprise reporting scenarios require users to enter dynamic search patterns similar to SQL LIKE and NOT LIKE functionality. While workarounds using DAX, calculated columns, custom visuals, or custom logic may be possible for individual fields, these approaches are not scalable when reports contain hundreds of filterable columns.
Current Limitation
After engagement with Microsoft Support, Advisory, and Architecture teams (Case #2606220040001996), it was confirmed that:
- Power BI currently has no native runtime implementation of SQL LIKE / NOT LIKE filtering.
- Existing alternatives do not adequately support large-scale implementations with hundreds of columns.
- Third-party visuals such as Smart Filter do not fully support SQL-style wildcard matching (for example _ as a single-character wildcard).
- This has been identified as a current product limitation.
Suggested Enhancement
Provide native filtering capabilities that support standard SQL wildcard expressions within Power BI report filters and slicers, including:
- % for multi-character matching
- _ for single-character matching
- LIKE and NOT LIKE style filtering behavior
- User-entered runtime pattern matching
- Consistent functionality across all text columns without requiring custom DAX or calculated columns
Example Scenarios
| User Input | Expected Result |
|---|---|
| ____ | Return all values with exactly 4 characters |
| __A_ | Return all 4-character values where the third character is A |
| %ABC% | Return values containing ABC |
| A% | Return values starting with A |
| %XYZ | Return values ending with XYZ |
Business Impact
This enhancement would provide significant value for organizations with large-scale enterprise reporting environments by:
- Reducing reliance on custom DAX and calculated columns
- Improving report usability for business users familiar with SQL-style searches
- Enabling dynamic filtering across hundreds of columns without additional development effort
- Improving scalability, maintainability, and performance of enterprise Power BI solutions
Supporting Information
Microsoft Support has confirmed that there is currently no supported native Power BI solution for dynamic SQL LIKE wildcard filtering at scale and recommended submitting this requirement through the Microsoft Fabric Ideas portal for product team evaluation.
Recent ideas
Return .png Attachments to Subscriptions
With the latest update to Power BI subscriptions, the separate .png report snapshot attachment has been removed. From Microsoft’s troubleshooting documentation, this appears to be an intentional chan...KDC_Group_PBI2 hours agoNew MemberNew53Views6likes1CommentEnable Table-Level Execution in Fabric Copy Job
a { text-decoration: none; color: #464feb; } tr th, tr td { border: 1px solid #e6e6e6; } tr th { background-color: #f5f5f5; } Fabric Copy Job currently executes the entire copy job when the Run butt...tmihara3 hours agoNew MemberNew5Views0likes0CommentsChange the owner of the content of the created lakehouse
Since there is a problem with operation when the account of the owner of the created lakehouse content is closed, I would like to add the function to change the function of the ownerv-isakaguchi17 hours agoMicrosoft EmployeeCompleted4.1KViews115likes7Comments