Ravindra BagaleCourses & study guides मराठी Track your progress

11. Data Modelling

11.12 Perspectives

Start with a problem

Store managers see six tables and forty measures in the field list. They need only Orders, Date and DarkStore with five measures.

Explain simply

A perspective is a named subset of tables, columns and measures. It hides clutter; it does not secure data.

Show the data

Table In the model In the Store managers perspective
Orders Yes Yes
Date Yes Yes
DarkStore Yes Yes
Product Yes No
Customer Yes No
OrderPartner (bridge) Yes No

Fictional sample data. Practice file: visuals_orders_apr_may_2026.csv, dark_stores_maharashtra.csv (source: src/practice/).

Demonstrate

  1. Desktop › External Tools › Tabular Editor.
  2. Model › Perspectives › New perspective › name Store managers.
  3. Include Orders, Date, DarkStore and the measures Total Sales, Total Orders, AOV, Avg Delivery, Cancellation % › save to the model.
  4. In Power BI, turn on Personalize visuals for the report and choose the perspective (Format pane › Personalize visual › Perspective).

    Your result should look like this:

    Fields offered to viewers
    Orders, Date, DarkStore · 5 measures

Show the result

Your result should look like this (fictional sample):

Measure Value
Total Sales ₹2,100
Total Orders 8
AOV (delivered) ₹300
Avg Delivery (min) 11.0
Cancellation % 12.5%

Explain what happened

The perspective changed only what is offered in the field list. The values are the same, and a user with model access can still query hidden tables.

Explore further

  • Use RLS (rows) and OLS (objects) when data must be secured.

Detailed reference for this lesson:

A perspective (a smaller named view of the model) does not delete anything from the model. It only shows a report author fewer tables.

A perspective is a named subset of tables, columns and measures. It does not secure data. A user who can open the model another way can still see the hidden objects. Row security stays RLS (Module 26). Object hiding for security is OLS (section 26.6), not a perspective.

Power BI Desktop's everyday ribbon does not always show a Perspectives button. Perspectives are created in the model metadata, often with Tabular Editor or TMDL view, and they matter when a large model is used from Excel (Analyze in Excel) so an author is not lost in fifty tables. If your Desktop has no perspective command, do not invent one. Use display folders and hidden technical columns (section 11.7) for a student report.

Worked example. The fictional model has Orders, Product, Customer, DarkStore, Date and a technical OrderPartner bridge. A perspective called Store authors includes DarkStore, Date, Orders and the sales measures, and leaves the bridge out of that view. Ravindra Bagale still has the bridge in the full model for the partner page. Hiding it from one perspective does not stop a DAX query that names the table.

Ravindra Bagale's Tip

In interviews, do not call a perspective security. It controls visibility; it is not a short form of RLS. Many students say this the wrong way round. Remember this.

Fix common mistakes

Wrong result Reason Correction
Thought Customer data was now secured Perspectives only hide Use RLS/OLS for security
No Perspectives button in Desktop Perspectives are edited with an external tool Use Tabular Editor
A new measure is missing for managers Perspective not updated Add it to the perspective

Practise independently

Practice task

Which belong in the Store managers perspective: Cancellation %, OrderPartner, Date[FY]?

Hints
  • Think about what a store manager filters and reads.
Solution

Cancellation % (12.5% overall) and Date[FY]: yes. OrderPartner: no, it is a technical bridge table.

Think like a professional

Say it plainly in interviews: a perspective is convenience, RLS is security.

Ravi Sir's Cards for this lesson

मराठी cards