Power BI is a business intelligence platform that allows businesses to share reports and datasets within the Power BI Service, the cloud environment of the tool. One method of creating efficiencies in Power BI is to centralize and share the datasets so they can be reused for more than just one report.
In this blog, we will cover what a shared Power BI Dataset is, why and when datasets should be shared, how to share datasets, as well as how to use a shared dataset.
What Does it Mean to “Share” a Power BI Dataset?
Sharing a Power BI Dataset is a concept whereby developers can centralize a dataset by hosting it in a shared workspace within the Power BI Service. Once a dataset is hosted in a workspace, those with at least Contributor access can leverage the dataset to build additional reports. Those with View access to the Workspace App can be given build permissions on the dataset, enabling them to self-serve their own data either directly in the Power BI Service from their Personal Workspace or using the desktop application.
Why Share Power BI Datasets?
There are many benefits when it comes to sharing a Power BI dataset, as described in the previous section.
First and foremost, sharing datasets reduces redundancy. Instead of storing multiple copies of the same dataset, a single copy can be used to service as many reports as are required.
This not only means reducing the physical consumption within your Power BI tenant but also means less strain on the data source because less Power BI Datasets need to be refreshed.
The other benefits are mostly to do with the democratization of data, which means several things in this context.
Providing Self-Service: giving end-users build access to the dataset(s) within a Power BI Workspace App allows them to connect to the datasets and build their own reporting. This allows for customization and ad hoc analysis while also reducing the workload on the development since the end-users may not need to come to them for one-off data requests (check out this blog on Power BI Workspace strategies for self-service).
Enabling Citizen Data Analysts: if your business’ data environment is built using a set of tables and views in Snowflake, for example, having a centralized Power BI Dataset will prevent analysts from needing to be intimately familiar with the Warehouse Database, Schema, Table, and View details in Snowflake, and allows the same set of tables and views to be available for the immediate creation of reports, without additional configuration.
In summary, sharing (or centralizing) your Power BI Datasets creates efficiencies by reducing data duplication, which means eliminating storage consumption in Power BI as well as reducing the strain on your data source of origin. It also allows for the literal “sharing” of datasets for your end-users.
How to Share Power BI Datasets
If you have ever published a Power BI Report to a workspace, you’ve already completed most of what is required in order to share a Power BI Dataset. Once you have published the report, all that’s left is giving the right people the right access.
Option 1 – for other developers: as described earlier, anyone with at least Contributor access to the workspace where the dataset is hosted can create reports using the dataset. Be careful. However, this method also gives those with access the ability to edit the dataset. Therefore, this option should only be used for members of the development team to leverage the same dataset for multiple reports. To do this, click the Manage Access button at the top of the workspace and assign the user(s) as Admin, Member, or Contributor.
If the companion report published with the dataset is no longer needed, you can feel free to delete it. The dataset will still be maintained in the workspace for users to leverage.
Option 2 – for self-service: if you want to provide your end-users with build access to the dataset (as opposed to edit access) so they can self-serve their own reporting, you can do that in the Workspace App settings. Update or create the Workspace App by clicking the button on the top of the workspace and going to the Audience page. In the Edit Audience panel on the right, open the Advanced settings section and check the box next to Allow people to build content with the datasets in this app audience.
End users will now see the dataset available for their use in the desktop application as well as in the Power BI Service.
How to Use Shared Power BI Datasets
Now that you have done all the legwork to set up your shared dataset in Power BI Service, using the dataset is beyond simple. You can create a new report using the shared dataset either in the Power BI Desktop Application or within a workspace in Power BI Service.
From the Desktop Application:
Go to Get Data (or Onelake Data Hub) > Power BI Datasets.
The OneLake Data Hub will appear and show you all the datasets you can access – based on either of the methods in the previous section.
Search for and select the appropriate dataset, then click Connect
You should now be connected to your dataset and can use the appropriate fields and measures to build out reports. One thing you will notice is that while you can add new measures and calculated columns, you will NOT be able to edit the DAX from the ones in the dataset you’ve just connected to.
From the Power BI Service:
With at least Contributor access to the workspace where the dataset is hosted. Click the + New button on the top left of the workspace and then click Report
With build permissions to a Workspace App’s underlying dataset. Open your personal “My Workspace” and click the + New button on the top left, and then click Report.
On the next screen, click the Pick a Published Dataset button.
Search for and connect to the dataset in the popup window.
Start building the report in the new window that opens.
Tips and Reminders for Working With Shared Power BI Datasets
You will use Power BI’s Live Connection method to query the dataset for each visual change. This is similar to Direct Query, but since the Shared Dataset can be an import Dataset, you can reduce the number of queries to the source data.
Users make impactful decisions based on the data they are presented with. Ensure that your data is modeled so all aggregations of measures are appropriate.
Just because you’re sharing the dataset doesn’t mean you must share all the data. Setting up Row-Level Security will help ensure your data stays in the hands of only the users who need it. Blog: Power BI Row Level Security
Sharing Power BI Datasets elevates efficiencies in several ways. As developers, we get to reduce redundancies and data duplication, thereby reducing unnecessary footprints in Power BI as well as the source systems.
Self-service is also a benefit because it gives users the ability to answer their own questions by providing access to the dataset, not just the report. This direct access can be very beneficial to end-users just beginning their analytics journeys, something those who may be newer to DAX or Power Query will not double be thankful for.
If you need additional help or are curious about leveraging Power BI, reach out to the experts at phData today for help, guidance, and best practices!
Shared Workspaces allow groups of users to share reporting and collaborate on analysis, including shared Power BI Datasets. This is the first step in developing a self-service analytics model.
When a Power BI file is published to Power BI Service, it splits into two segments: the report, which consists of all of the visualizations, and the dataset, which is the model of all of the data sources and any data that may be loaded into these data sources. The Dataset can then be shared to create other Power BI Service or Desktop reports.