What is single direction vs both direction filtering?

< 1 min read

Single direction filtering means the filter flows from dimension to fact table only.
For example, selecting a Date filters Sales records but Sales won’t filter Date.
This keeps calculations predictable and improves report performance.
We used this in most star schema models like Customer → Orders.
Both direction filtering means tables filter each other mutually.
Useful when handling many-to-many like Students ↔ Courses via bridge table.
But it can create ambiguity and wrong totals due to circular filter paths.
In one dashboard, product counts increased incorrectly because of this.
So I kept single direction by default and enabled both only for specific visuals.

Power BI Essentials