Data model best practice

[vejas]

Hi, 

    Which way better:

 -  create one big data model with multiple fact tables - purchase, sales, inventory, general ledger and dimensions (all main sql tables used by ERP)  and create Power Bi different reports on top of that "General" data model or

 - create separate dedicated data models with 1 or 2 fact tables  to analize sales, inventory, financial data etc. 

[v-juanli-msft]

Hi @vejas 

For the method 2:

It would make the data size less,

but if you have sales data in one Power BI file, have inventory, financial data in another Power BI file,

then it maybe difficult to analyze "sales, inventory, financial data" together.

For method1, if you create correctly, it may not slow the performance so much.

Here are some references:

Building a Microsoft Power BI Data Model

Power BI performance best practices

Best Regards

Maggie

Community Support Team _ Maggie Li
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[vejas]

so it just matters of perfomance ?

[v-juanli-msft · Accepted solution]

Hi @vejas 

it is not just matters of perfomance.

When you have too large data in a Pbix file, it may not support to publish the report to web side.

Thus you could not share Power BI reports to more people.

Also, with a correct and good model, it is much easier to create calcuated columns/measures so that it is simple to create proper visuals.

Best Regards

Maggie

Community Support Team _ Maggie Li
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.