When a Power BI visual shows the same number in every row, or a slicer seems to do nothing, the cause is almost always the data model, not the chart. Relationships tell Power BI how tables connect and which way filters travel.
One-to-many: the normal case
A good model has dimension tables (Customers, Products, Date) and fact tables (Sales). Each customer appears once in Customers but many times in Sales. That is a one-to-many relationship, shown in the model view as 1 on the dimension side and * on the fact side.
Power BI needs the “one” side to have unique values. If your Customers table has duplicate IDs, Power BI will refuse to create a one-to-many relationship, which is a useful warning that the data needs cleaning first.
Filter direction
By default, filters flow from the one side to the many side: picking a city in a Customers slicer filters Sales. This is called single direction and it is what you want most of the time.
Both directions (bi-directional) lets filters flow back the other way. It can feel convenient, but it often creates ambiguous paths, slower reports and totals that are hard to explain. Use it only when you understand exactly why you need it.
Common mistakes
- Many-to-many by accident: joining two fact tables directly instead of through a shared dimension.
- Inactive relationships: a dotted line in the model view means the relationship exists but is not used unless a DAX measure calls USERELATIONSHIP.
- Mismatched data types: a text “101” will not match a number 101.
- No date table: time intelligence works best with a proper date table marked as such.
A simple habit
Before building visuals, open the model view and check that every fact table connects to dimensions with one-to-many, single-direction lines. This star shape fixes most strange totals before they happen.
Related reading: star schema explained, why every report needs a date table and DAX CALCULATE.
Power BI modelling is part of our Data Analytics course in Bangalore.
Book a free demo class at our BTM Layout centre, and ask a counsellor for batch details.
