Power BI Best Practices

Using too many visuals slows report performance as each visual requires data processing and rendering. Limit widget visuals to eight per report page, grids to one per page, and tiles to 10 per dashboard. Consider Power BI's new card visual, which displays multiple cards within a single container — consolidating information into a single query.

Personal mode imports data into Power BI, which can cause resource limitations and performance issues with large databases. Standard mode keeps data in its original location, minimizing duplication, reducing memory usage, and avoiding bottlenecks.

Sharing one gateway for DirectQuery live connections and scheduled refresh can overload it during refreshes and slow live connections. Separate gateways let each workload run efficiently without affecting the other.

Calculated measures evaluated at query time are more efficient than calculated columns for complex aggregations. Pair them with filters to keep query performance predictable.

Offloading processing to the source optimizes query execution. Calculations evaluated closer to the data are generally faster than running them inside Power BI.

Star schemas are simpler and more denormalized than snowflake schemas. They reduce required joins, accelerate query execution, simplify report development, and ease maintenance.

Each slicer fires two queries (data and selection details). Too many slicers drastically slow performance. Use the Filter pane to evaluate and remove unnecessary slicers.

Hosting the Power BI tenant and data source in the same region minimizes network latency and speeds up data transfer, query execution, and report rendering.

Partitioning splits large tables into smaller subsets based on specific criteria. Power BI uses columnar indexes, so longer, leaner tables perform better. Importing only the necessary fields and tables improves loading efficiency and reduces resource consumption.

Calculation groups let you define calculations (like time intelligence) once and apply them to multiple measures, reducing measure count and simplifying maintenance.

Templates enable standardized, branded reports faster. They save custom color palettes and themes, auto-connect to commonly used data sources, and ship with common DAX measures.

Add an Apply button to slicers and the filter pane so changes only fire queries when the user clicks Apply. This reduces the number of queries triggered, especially in complex reports.

These relationships significantly impact performance due to increased data processing and query complexity. Avoiding them on high-cardinality columns reduces resource consumption and speeds up rendering.

Float and double can introduce round-off errors and require more processing power. Avoiding them improves accuracy and report efficiency.

Auto-generated date tables create a separate table for each date column, bloating the model. A single custom date table (with date and time split for better compression) streamlines the model and enables more efficient time intelligence.

Disabling attribute hierarchy for non-attribute columns (like measures) streamlines the data model, reducing size and load time.

Use bookmarks, drillthrough pages, and tooltips to reduce data loaded at once. This improves page load times — especially on landing pages where quick access matters most.

Background images use fewer resources than rendering multiple individual visuals for the same effect.

Choose Import, DirectQuery, or Dual based on data aggregation needs and visualization requirements. The storage mode directly affects responsiveness and efficiency.

In DirectQuery sources with enforced primary keys, verify the Assume Referential Integrity setting. If Off, Power BI defaults to slower outer joins instead of inner joins.

RLS limits data access based on user roles, so Power BI only imports data the user is authorized to view. Keep RLS logic simple and push complex calculations to the source to reduce the computational load on Power BI.

Certified visuals have passed rigorous quality testing and are the only custom visuals that work in Export to PowerPoint mode and email subscriptions. Microsoft verifies their performance and robustness.

Sensitivity labels (High, Medium, Low Business Impact) raise user awareness about data security. High-impact data requires a policy exception to be shared externally, helping enforce appropriate handling of sensitive information.

Power BI's default cache update frequency is one hour. Synchronizing it with the source refresh frequency ensures data accuracy and avoids unnecessary processing.

Light backgrounds make reports printer-friendly and easier to share both online and offline.

Limit displayed numerals to four digits and decimals to two places. This improves readability and consistency, especially when scaling to thousands or millions.

Tooltips add context in a compact form. Curate the visuals inside them so they help without overwhelming the user.

Use clear, business-friendly names for columns and measures, and hide unused columns to reduce user confusion.

Turn on the Personalize visuals option when sharing reports. Combined with personal bookmarks, it empowers users and facilitates self-service BI.

Excessive scrolling disrupts analysis. Non-standard sizes display inconsistently across devices, producing poor responsiveness and a disjointed visual presentation.

Drillthrough buttons with context-driven conditional formatting offer a more intuitive way for users to explore data than right-click drillthrough.

Evaluate cross-filtering and cross-highlighting interactions. Remove the ones not relevant to your audience. For DirectQuery reports, use the Optimize Ribbon to tune interactivity.

Field parameters let users switch the measure or dimension being analyzed inside a visual, reducing the need for duplicate visuals or bookmarks.