Manage Data
Permissions
This page is about who can see and use a specific dataset, not who has which role platform-wide. A role decides what kind of things a person can do in general; dataset permissions decide which specific datasets they can do those things to.
Levels of access
Access to a dataset is layered — each level below implies the ones before it.
Discoverable
The dataset shows up in Catalog search and its name, description, and tags are visible — but its columns and rows are not.
Queryable
A user can run analyses, build charts, and preview rows against the dataset.
Exportable
A user can pull data out of the platform — as a report attachment, a downloaded file, or through the API.
By default, a dataset is queryable by every member of the tenant that owns it. Restrict it from a dataset's detail view — grant access to specific roles, specific users, or leave it open to everyone with a role that can query data at all.
How this relates to Roles
A role such as analyst or viewer grants the general ability to run analyses or read data. Dataset-level permissions narrow that down to specific datasets — a viewer with access restricted to one dataset can only ever act on that one, no matter what their role would otherwise allow. See Permissions under Administration for the platform-wide role model this builds on.
- Restricting a dataset does not hide it retroactively from results someone has already exported or saved — it only governs access going forward.
- A dataset's sensitivity classification (see Metadata) is a label, not an access control — restricting who can query a sensitive dataset still has to be set explicitly here.
- owner and admin can always see and query every dataset in the tenant.