Power BI adds visual calculations

Power BI’s visual calculations feature changes where a large class of reporting work gets done. Microsoft says a visual calculation is a DAX calculation defined and executed directly on a visual, not stored in the semantic model, which means authors can build calculations against what is already on the chart or table. (learn.microsoft.com) That matters because many business users do not need a new model-level measure every time they want a running total, a moving average, or a rank. (learn.microsoft.com) Microsoft’s documentation says visual calculations can complete common business calculations such as running sums and moving averages, and that they often lead to simpler DAX, easier maintenance and better performance than alternative approaches. For finance teams, the practical effect is speed. A report author can add a calculation inside the report canvas instead of handing a request back to a model developer for every new KPI view, especially when the need is visual-specific rather than enterprise-wide. (learn.microsoft.com) ### So what actually changed inside Power BI? Microsoft’s Learn documentation now describes visual calculations as a standard calculation option in Power BI Desktop alongside custom columns, calculated columns, calculated tables and measures. In that framework, visual calculations are computed on demand, stored on the visual, and change with user interaction in the report. (learn.microsoft.com) The editing flow is also different from classic measure authoring. Microsoft says authors select a visual, click “New visual calculation” in the ribbon, and then work in an edit mode with a visual preview, a formula bar, and a visual matrix that shows results as calculations are added. (learn.microsoft.com) ### Why are Power BI users talking about running totals and moving averages first? Microsoft’s own examples center on those patterns. The company’s documentation shows a running sum written as `RUNNINGSUM([Sales Amount])`, and its DAX references include dedicated support for functions such as `MOVINGAVERAGE`. The RANK function documentation also notes behavior available in visual calculations, including reset behavior across a visual hierarchy. (learn.microsoft.com) Those are common finance and operations asks. Revenue-to-date, trailing averages, rolling margin views, inventory trend smoothing and ranked customer or product views are usually requested at the report layer, often after the core model is already published. (learn.microsoft.com) Visual calculations fit that pattern because they work against the aggregated data already shown in the visual. ### Does this mean DAX no longer matters? DAX still matters because visual calculations are written in DAX and can use existing DAX functions. Microsoft says the feature removes some of the complexity of the semantic model, but it does not replace model design or enterprise measures. (learn.microsoft.com) The boundary is scope. Microsoft says visual calculations can only refer to what is on the visual, because they are stored on the visual rather than in the model. That makes them useful for report-specific analysis, while shared business logic that must be reused across many reports still belongs in the semantic model. (learn.microsoft.com) ### Where does this help finance teams most? Finance reporting is full of calculations that are repeated, narrow and presentation-specific. A monthly revenue bridge may need a running total on one page, a 3-period moving average on another, and a rank of customers or cost centers on a third. Visual calculations let report authors add those layers without changing the underlying model each time. (learn.microsoft.com) Microsoft also says visual calculations often operate on aggregated data instead of detail-level data, which can bring performance benefits. For dashboard owners, that can reduce the tradeoff between quick iteration and model complexity. (learn.microsoft.com) ### What should teams watch before moving everything into visuals? Microsoft’s own documentation points to a simple constraint: anything a visual calculation references must already be in the visual. That means teams still need discipline about which logic is local to one report page and which logic should be standardized centrally. The next step for most Power BI teams is operational, not conceptual. (learn.microsoft.com) Report authors can start by moving high-frequency requests such as running totals, moving averages and ranking into visual-level calculations, while model owners keep shared definitions for metrics like revenue, gross margin and working capital in the semantic layer. Microsoft’s training module and calculation-options guidance give the clearest map for that split. (learn.microsoft.com)