Scribd
Upload
56 views6 pages

SOP Power BI Modelling Version 2

This document outlines best practices for modeling in Power BI, emphasizing the importance of understanding data, using parameterized connections, and optimizing transformations. It provides specific guidelines for building reports, such as minimizing unnecessary columns, avoiding bi-directional filters, and achieving a star schema. Additionally, it includes optimization techniques like using the Performance Analyzer tool, simplifying calculations, and scheduling data refreshes during off-peak hours.

Uploaded by

Raj Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
56 views6 pages

SOP Power BI Modelling Version 2

This document outlines best practices for modeling in Power BI, emphasizing the importance of understanding data, using parameterized connections, and optimizing transformations. It provides specific guidelines for building reports, such as minimizing unnecessary columns, avoiding bi-directional filters, and achieving a star schema. Additionally, it includes optimization techniques like using the Performance Analyzer tool, simplifying calculations, and scheduling data refreshes during off-peak hours.

Uploaded by

Raj Das
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

Internal Process & Standard Operating Procedure for Best Power

BI Modelling Practices

Purpose:
This document explains the best modelling practices to be followed while starting or optimising an existing
Power BI report.

Context:
Best Power BI Modelling practices are general guidelines, and you should adapt them based on your specific
requirements and circumstances. Regularly review and optimize your data model as needed to ensure it
continues to meet your evolving business needs.

Best Model Building Practice to be followed while building a


Report:
1. Understand Your Data: Understanding the underlying data will help you to best design data model,
helps you to best represent the data.
2. Use Parameterised source connections while connecting to data source. It will be easy to migrate data
source of report from one environment to another.

Step-1 To create parameters which will be containing the source details.

Step-2 select the parameter while connecting to the source.


3. Maximum transformations while loading the data needs to be done near to the source end. The
hierarchy of transformation prioritisation that needs to be followed is:

 Data source end


 Power query
 Transformation using calculated Columns and Measures

4. Applied steps in Power Query editor while loading and transforming the data must be consolidated,
applying too many steps will slow down the refresh and data load.

5. Unselect unnecessary columns while building a report. Wide tables or tables with too many columns
consume a lot of memory, which affects the processing speed.

6. Split date and time in different columns for a data time column.
7. Unload the tables in power query which will not be used in the development.

8. If the data size is too huge, use incremental refresh while developing the reports in Power BI desktop.

9. While duplicating a table in power query to form any reference or bridge table, use table as reference of
Main table so as any changes done in the main table will be reflected in the reference table.
10. While building a model in Power BI try to attain a star schema, where one fact table is attached to all
the dimension tables.

11. Do not use bi direction filter in modelling, if required use cross filter in the measure while filtering one
side to many sides.

Try to avoid bi-directional filters, instead you want to filter one side from many side use cross filter Dax

CALCULATE([Distinct Count of ProductKey], CROSSFILTER(FactInternetSales[ProductKey],

DimProduct[ProductKey] , Both))

12. Avoid formation of many to many relationships, as this will affect the performance of the report,
instead use a bridge table to link such kind of scenarios.

13. Minimize the use of calculated columns to avoid unnecessary storage and processing overhead.
14. Write efficient DAX expressions to minimize performance impact. Avoid using heavy calculations or
nested functions unless necessary.
Best Optimization Technique to be Followed for a Power BI
Report:
1. Use the Performance Analyzer tool in Power BI Desktop to identify and analyse performance.

2. Identify slow-performing visuals, DAX calculations, or data loading steps and optimize them
accordingly.
3. Optimize your Power Query transformations by avoiding unnecessary steps, reducing the number of
transformations applied.
4. Use filters and transformations in the Power Query Editor to reduce the amount of data loaded into
your report.

5. Report Design: limit number of visuals in a single reporting page, use slicers and filters to reduce
amount of data displayed in a visual.
6. Enable incremental refresh for large datasets to refresh only the updated portion, reducing overall
refresh time.
7. Simplify complex calculations by breaking them down into smaller measures or using variables.
8. Schedule data refresh during off-peak hours to avoid impacting report performance during business
hours.

--------------------------------------------------------------
xxx-------------------------------------------------------------

You might also like