Best Practices for Report & Visual Performance Optimization

A report that takes 10 seconds to respond every time a user clicks a slicer will be abandoned by business stakeholders. Fast reports (rendering in under 1 to 2 seconds) require disciplined architecture across visual design, DAX formulas, and data modeling.

1. Visual Design Best Practices: The Rule of 15

Every visual on a Power BI canvas generates at least one separate query to the underlying VertiPaq or DirectQuery engine:

• If a page has 30 cards, 5 slicers, 4 bar charts, and 2 tables: clicking a single slicer fires over 40 simultaneous queries!
• Browsers and workstations have limited thread pools, causing query queuing and visual stutter.

2. DAX Formula Best Practices

1. 1
   Avoid IF checks on Blank measures:

• Bad: IF([Sales] = BLANK(), 0, [Sales]) (forces calculation of unneeded rows).
• Good: [Sales] + 0 or rely on native visual blank formatting.

1. 1
   Use Variables (VAR / RETURN) aggressively:

• DAX variables are evaluated once and cached, preventing redundant recalculations:

1. 1
   Prefer DIVIDE() over / operator.
2. 2
   Avoid referencing entire tables inside FILTER() when a single column suffices:

• Bad: FILTER(Fact_Sales, Fact_Sales[Region] = "North")
• Good: KEEPFILTERS(Fact_Sales[Region] = "North")

3. The Performance Analyzer Tool

Power BI Desktop includes an integrated profiling tool called Performance Analyzer:

1. 1
   Go to the View ribbon tab and check Performance Analyzer.
2. 2
   Click Start recording.
3. 3
   Click Refresh visuals.
4. 4
   Power BI will display exact execution times for every visual, broken down into:

• DAX Query: Time spent in the VertiPaq engine calculating numbers.
• Visual Display: Time spent by the browser rendering pixels, axes, and legends.
• Other: Time spent waiting in the thread queue.