Who can see what in a shared Data Studio report? Choosing between three types of data credentials

KEY POINTS

Key points

• Sharing permissions decide who can open or edit a report; data credentials decide whose identity is used to query the data. The two need checking together.
• Owner’s credentials let viewers read data through the report without holding access to the underlying data themselves.
• Viewer’s credentials query with each viewer’s own access to the underlying data, and do not split data by branch automatically.
• Service account credentials currently support BigQuery only, for cases where the organisation needs to manage a non-personal query identity.
• The owner of a report or data source asset may not be the owner of its data credentials. Both need checking at handover.

Why can some people see a shared report and others cannot?

Being able to open a report and being able to read the data inside it are different layers of permission.

A report is shared with a manager, who opens it and sees no numbers; another colleague with no GA4 access reads it without trouble. Both situations may come down to how data credentials are set.

According to Google’s data credentials documentation, credentials affect who can view the data a data source provides. When checking, it helps to split the problem into three questions:

• Does the person have permission to open this report?
• Whose identity does this data source use to query the underlying system?
• Does that identity still have the access it needs in the underlying system?

So it is not enough to look only at the sharing list, or only to change the credentials. Report permissions, query identity and access to the underlying data all shape what people end up seeing.

Which credential type suits which situation?

First confirm which data the recipients should see, then choose the identity used for queries.

Credentials	Do viewers need access to the underlying data?	Suits	What to check
Owner’s credentials	No; the credential owner provides authorisation.	Authorised managers, clients or colleagues across teams who need to read the data the report provides.	The sharing scope of the report and data source, and whether viewers really should see this content.
Viewer’s credentials	Yes; each viewer queries with their own identity.	Keeping the individual user permissions of the underlying system.	Whether permissions in the underlying system are set correctly; in this mode, hyperlinks and images in dimensions are not displayed.
Service account credentials	No; the service account provides authorisation.	Data in BigQuery, where the organisation wants a centrally managed, non-personal query identity.	BigQuery only for now; requires a Google Workspace or Cloud Identity managed organisation, plus the service account and the necessary permissions.

Which credential types are actually available still depends on the connector and the organisation’s admin settings.

With owner’s credentials, viewers read the data in the report through that credential’s authorisation; they do not gain direct access to the underlying GA4 property or dataset. Google also states that the credentials themselves are never displayed or transmitted to viewers in readable form.

Not having access to the underlying system, however, does not mean sensitive content stays hidden. Before sharing, check the fields, scope and recipients of the report.

If the requirement is that each branch sees only its own data, the separation method needs its own check. Viewer’s credentials follow each person’s underlying permissions, but they do not work out which branch someone belongs to. Row-level permissions in the underlying system, or a design that filters data by the signed-in user’s email, need to be set up and tested separately.

Three risks that are easy to miss

Editing rights, staff changes and the way queries run can all change how data is actually accessed.

1. Editors can change data credentials

People with edit access to a data source can, within what organisation policy allows, change the credential type or make themselves the credential owner.

After a change, the identity the report queries with may change too. For example, switching from owner’s to viewer’s credentials can leave people who read the report fine before, but have no underlying access, facing an error. So edit access to data sources should be managed as well, and after a change, check that the main users can still read the report.

2. Handing over assets is not handing over credentials

Google’s transfer ownership documentation is explicit: when ownership of a data source is transferred, its data credentials stay as they are and are not switched to the new owner’s.

Also, after someone loses edit access to a data source, it may keep querying with their credentials. That does not mean the report will keep working once the person leaves, their account is disabled, or their access to the underlying data is withdrawn.

At handover, list both the assets and the credentials, confirm the successor has access to the underlying data, update and test the credentials as needed, and only then deal with the old access through the staff change process. Passing the files to the next person does not necessarily complete the handover.

3. Credential choice also affects GA4 query quota

Google’s troubleshooting guide suggests considering owner’s credentials when GA4 quota runs short, so viewers share cached data and repeat queries drop.

The condition is that these viewers are already entitled to read the data the report provides. Data that needs individual authorisation should not be switched to a shared credential just to save quota. Check the data scope before the change, and watch whether quota use improves afterwards.

For fuller guidance on collaboration and permissions, see Data Studio tutorial: report permission management (in Chinese); organisations with requirements on encryption and data location can read the new security and compliance features in Data Studio Pro (in Chinese).

How to set and change credentials

Credentials can be set when adding or editing a data source; after a change, check the result under real viewer conditions.

1. When adding a data source. Before adding the data to the report, open “Data credentials”, review the available types and choose one.
2. When adjusting an existing data source. Edit the data source, click “Data credentials” at the top, and choose a new type or use an option such as “Make me the owner”.
3. Check the result. Confirm that the main viewers can read the data they should see, and do not see content they should not.
4. If an option is unavailable. Check the connector’s support and the organisation’s admin policy; service account credentials also need BigQuery and the managed environment set up.

Frequently asked questions

For sharing or handover problems, separate asset permissions, credential identity and underlying data access before changing settings.

I shared a report with a manager, but they see an error. What could cause it?

The data source may use viewer’s credentials while the manager has no access to the underlying data. The credentials may also have expired, the data source may have changed, or the organisation may not have enabled Data Studio. Read the error message first, then decide whether to grant access or, where the data scope allows, switch credentials.

With owner’s credentials, does the other person get GA4 access?

No. They read data through the report and do not gain direct access to the underlying GA4 property. The data the report actually shows still needs to fit the purpose of sharing.

Will a colleague’s reports break when they leave?

It depends less on whether they leave than on whether the accounts, credentials and underlying access the report relies on are still valid. Transferring asset ownership does not update credentials automatically, so check each item at handover and test data access after the successor takes over.

Sharing a report also puts a set of data access arrangements into effect. Recording who it is shared with, the data scope, the query identity and handover responsibility makes it possible later to know why settings are the way they are, and who handles changes. To review the permissions and credentials of existing reports, get in touch.

Further reading

• GA4 numbers don’t add up? Reporting identity concepts
• Why does nobody look at the BI reports you worked so hard on?

References

1. Google Cloud, Data credentials.
2. Google Cloud, Data Studio troubleshooting guide.
3. Google Cloud, Transfer ownership.