Cross-tenant Data Sharing with Power BI: A Comprehensive Guide

In today’s world of digital data, information is everywhere, and sharing it has become more crucial than ever. However, when it comes to sharing datasets between users from different tenants, things can get a bit complex. Each group, or “tenant,” within an organisation operates in its own designated space with unique datasets. When users from one tenant wish to access semantic models (datasets) from another, several challenges arise, requiring multiple steps to ensure smooth data sharing.

I’ve personally faced this challenge in the past, attempting to set it up independently. Unfortunately, I discovered the process was not as straightforward as I hoped, and there was a lack of helpful information available. After investing significant time in troubleshooting, I eventually cracked the code. Since I couldn’t find a comprehensive online tutorial, I decided to create a step-by-step guide myself. With the assistance of this tutorial, the setup shouldn’t be as daunting as it initially seems. ๐Ÿ˜Š


In this scenario, we have two organisations: My own company referred to as โ€œOrganisation Xโ€ and another company referred to as “Organisation Y”. I have an additional account, โ€œben.de.keyser@organisationy.beโ€, equipped with a Power BI license at Organisation Y.


The goal is to share the “Basic Sales Report” semantic model from Organisation X with “ben.de.keyser@organisationy.be.” This enables โ€œben.de.keyser@organisationy.beโ€ to establish a DirectQuery connection not only to his own semantic models within Organisation Y but also to the shared semantic model from Organisation X.

Step 1) Invite the user you want to share your dataset with to your organisation


To initiate the process, we start by navigating to portal.azure.com, where we can see the homepage.

After that, we proceed to the Microsoft Enterprise ID and click on the “Users” tab. Within the “Users” tab, we select the “Invite external user” button.

In the external user tab, we add the โ€œben.de.keyser@organisationy.beโ€ to the organisation as you can see below. After this, you will receive an email with an invitation to join the organisation with an external account.  After you accept this, you will have a guest account in the organisation!

Step 2) Change settings within Power BI service

Once the external user has been added to our organisation, the next step involves checking specific settings within the Power BI Service.

Initially, we must enable guest users to operate within their own tenant with the shared semantic model. Youโ€™ll need a Fabric Administrator license to do this.

If weโ€™ve enabled this, it’s crucial to activate the option that permits DirectQuery connections to Power BI semantic models (datasets).

Step 3) Allow external sharing on the semantic model

Begin by selecting the semantic model (dataset) you wish to share with users from another organisation. Once you’ve made your choice, proceed to enable “External sharing” for the selected semantic model.

Navigate to the settings of the semantic model, and there you’ll find the “External sharing” setting, which needs to be switched on:

Step 4) Enable external semantic model setting in Power BI Desktop

Next, let’s open Power BI Desktop. Initially, we must activate the feature that enables us to connect to external semantic models. As of the current writing, this feature is still in preview. To enable it, navigate to “Options and settings” -> “Options” -> “Preview features” in Power BI Desktop. In the preview features section, you can enable the feature, as illustrated in the screenshot below:

Step 5) Connect to the external dataset within Power BI Desktop

This is where it becomes a bit tricky, so carefully follow the steps outlined below.

Begin by logging out of your current account within Power BI Desktop, allowing you to access the sign-in option located in the upper right corner.

Once logged out, click on the sign-in button and enter the email associated with your external email address. For instance, in my case, this would be โ€œben.de.keyser@organisationy.beโ€.

After that, youโ€™ll need to pick an account in the following screen. Here, you need to choose โ€œUse another accountโ€.

Next, choose โ€œSign-in options

And pick โ€œSign in to an organisationโ€

Subsequently, enter the name of the organisation that owns the semantic model you intend to connect to. In our scenario, this would be “organisationx.be.”

Next, choose your external account and log in:

At this point, you are logged into organisation X using an account from another organisation. Navigate to “Get data” -> “Power BI semantic models,” and you’ll be able to connect to the shared semantic model we’ve set up:

However, as you can see, you cannot view the semantic models from your own organisation (organisation Y). Currently, access is limited to the semantic model from the other organisation through your guest account.

To fix this, click on “Sign out” once again. This action will log your account out, as done in a previous step. Next, log in again using “ben.de.keyser@organisationy.be,” without connecting to the organisation as previously outlined in the steps.

Following that, once more, click on “Get data” -> “Power BI semantic models,” and now you can observe the datasets within your organisation. If you navigate to the “External data” section and wait for 2/3 seconds, the dataset from the other organisation, which was shared with your account, will become visible.

Great! We have now successfully achieved our goal, gaining access to both our organisation’s semantic models and a shared semantic model from another organisation.

Important note

Sometimes, you may come across the following error:

If you come across this error, simply go back to the beginning of step 5. Log in again to the organisation’s domain name, then log out. Afterward, log in once more without connecting to their organisation. Sign in with your own organisation account, and the external data error mentioned above should disappear. The external semantic models shared with you should then reappear.

It seems there might be a minor glitch in the current Microsoft setup. As of now, restarting from step 5 is the suggested solution to resolve this error. Please note that this is still in preview mode, so such situations may occur.

In conclusion, navigating the challenges of external data sharing, particularly across different organisations, demands a systematic approach. By following the steps outlined in this guide, you’ve successfully learned how to invite external users, share semantic models, and overcome potential challenges within Power BI. Although occasional errors may arise, such as the one highlighted in the side note, the suggested solution makes sure things get back on track smoothly.

Now that you have these insights, collaborating and making the most of shared datasets should be straightforward. Stay tuned for any updates, but for now, you’ve got the fundamentals of external data sharing in Microsoft’s setup. Enjoy exploring!

Ben De Keyser

Consultant @ Lytix