Summary
Add the ability to organize Power Query queries into user-defined groups that can be refreshed independently within Power BI Desktop.
Problem
As Power BI models become larger and more modular, developers often work with hundreds of queries across multiple subject areas such as Sales, Finance, HR, and Inventory. Today, refresh operations are limited to refreshing individual queries or the entire model.
This makes development slower and more resource-intensive, particularly when only one functional area has changed.
Proposed Solution
Introduce Query Groups that allow developers to:
- Create logical groups of queries (e.g., Finance, Inventory, Sales, HR).
- Assign one or more queries to each group.
- Refresh an individual group without refreshing unrelated queries.
- Refresh multiple selected groups.
- Continue to support full model refresh when required.
Benefits
- Faster development by refreshing only the relevant portion of the model.
- Lower memory consumption during development, especially for enterprise-scale models.
- Better organization for complex semantic models with hundreds of queries.
- More efficient debugging and testing by isolating specific data domains.
- Improved productivity when working with modular data architectures.
Example
A model contains:
- Finance (45 queries)
- Inventory (60 queries)
- Sales (55 queries)
- Manufacturing (70 queries)
If only the Inventory data has changed, the developer could refresh only the Inventory group rather than processing the entire model. This would reduce refresh time, lower memory usage, and improve the development experience.
Business Impact
This feature would significantly improve the developer experience for enterprise Power BI models by enabling faster iteration, lower hardware requirements during development, and a more scalable approach to managing large semantic models. It would also encourage more modular model design and make maintaining large datasets substantially easier.
