Sharing in Power BI – how to share Report, Dataset or Dataflow

If your using in Power BI in company should be effective, you don´t have to do one work twice. Sharing instead of duplication brings at least these benefits:

No waste of time
No problems with inconsistencies. It is better to have one table correct then to have many tables “probably almost accurate”.

There are at least three nice ways how to share work in a company:

Report sharing

You can share complete reports. Someone else can look at the charts and, if needed, modify them.

This is a typical use case when “someone creates reports and someone else looks at them”.

Dataset sharing

You can also share Datasets. Which means to share the data model (tables, relations, measures…), when other users can use it and create their own reports from them.

Using of shared datasets is simple, you can just select it as a data source:

When using shared Dataset, you can:

Create visuals
Create measures

You can´t:

Create new relationships between tables
Create new tables from other tables
Modify data using Power Query
See data in tables

In a Workspace it looks like this:

Dataflow sharing (Power Query sharing)

You can also share data source connections and data cleaning procedures (called Power Query in Power BI Desktop and Dataflow in cloud).

One user handles connections, refreshes, authorization etc, other people can use it to build their own data models.

Shared Dataflow can simply be used as a data source:

When working with a shared Dataflow, you can:

Create your own data model – tables, relations, measures…
Create visuals and measures (similarly to shared dataset)
Modify data in Power Query (these modifications are done after modifications in shared Dataflow).

You can´t:

Modify the data source
Control all the data processing. For example, obviously, you can´t add a column which was removed in Dataflow.

The result looks like this:

If your using in Power BI in company should be effective, you don´t have to do one work twice. Sharing instead of duplication brings at least these benefits:

No waste of time
No problems with inconsistencies. It is better to have one table correct then to have many tables "probably almost accurate".

There are at least three nice ways how to share work in a company:

Report sharing

You can share complete reports. Someone else can look at the charts and, if needed, modify them.

This is a typical use case when "someone creates reports and someone else looks at them".

Dataset sharing

You can also share Datasets. Which means to share the data model (tables, relations, measures...), when other users can use it and create their own reports from them.

Using of shared datasets is simple, you can just select it as a data source:

When using shared Dataset, you can:

Create visuals
Create measures

You can´t:

Create new relationships between tables
Create new tables from other tables
Modify data using Power Query
See data in tables

In a Workspace it looks like this:

Dataflow sharing (Power Query sharing)

You can also share data source connections and data cleaning procedures (called Power Query in Power BI Desktop and Dataflow in cloud).

One user handles connections, refreshes, authorization etc, other people can use it to build their own data models.

Shared Dataflow can simply be used as a data source:

When working with a shared Dataflow, you can:

Create your own data model - tables, relations, measures...
Create visuals and measures (similarly to shared dataset)
Modify data in Power Query (these modifications are done after modifications in shared Dataflow).

You can´t:

Modify the data source
Control all the data processing. For example, obviously, you can´t add a column which was removed in Dataflow.

The result looks like this:

Tutorials