Notes from the Power BI User Days 2026
At the Power BI Gebruikersdagen 2026 I attended several sessions on modelling, governance, AI and the evolution of the Power BI platform. In this series I share a few notes and observations that stayed with me. This page is about the difference between measures and calculated columns.
A distinction that matters
DAX offers two ways to store a calculation: a measure or a calculated column. On the surface they look similar — both use DAX and both produce a result. But they work in fundamentally different ways, and that difference has direct consequences for the performance of your reports and the reliability of your calculations.
What is a calculated column?
A calculated column is evaluated when the model is loaded or refreshed, and the result is stored in the model, row by row. That sounds useful, but it comes at a cost.
- The column consumes memory in the model
- The calculation cannot see the filter context of a report
- You cannot use it for dynamic calculations
Calculated columns are useful when you need a value at row level that you want to use later as a filter or dimension — for example, a price category based on the selling price per row.
What is a measure?
A measure is evaluated at the moment a visual is rendered — not when the model loads, only when Power BI needs the result, and not for each row individually but for the aggregation the visual requests. That makes measures powerful.
- A measure sees the filter context of the report
- A measure responds to slicers, filters and cross-filtering
- A measure consumes no model memory
The vast majority of calculations in a Power BI report belong in a measure.
What goes wrong?
The most common mistake: someone writes a calculation as a column that should have been a measure. The calculation appears to work in a simple table, but as soon as a slicer or filter is added the result no longer makes sense — the column was already evaluated at load time and knows nothing about what the user selects later. If the same mistake is made in dozens of columns, the model also consumes far more memory than necessary.
Reflection
The distinction between measures and calculated columns is one of the first things I cover in every DAX training. Not because it is difficult, but because it goes wrong so often and the consequences are so directly visible in the quality of a report. Want to work on this properly? See the Power BI training I offer.