Security predicate for sales agreement and account forecast analytics
Key Takeaway
Without a security predicate, anyone who can access a dataset sees all of its records. To restrict rows in a dataset that already exists, edit the dataset itself, because security changes in the dataflow or recipe have no effect after the dataset is created.
When you set up the Analytics for Manufacturing app, you select a security predicate. A security predicate is a filter condition in CRM Analytics that controls which rows (records) of a dataset each user can see. In Analytics for Manufacturing, the predicate decides which sales agreement and account forecast records each user sees. The predicate matters because, without one, anyone who can open a dataset sees every record in it.
What a security predicate does
When a user queries a dataset that has a predicate, CRM Analytics checks the predicate. CRM Analytics then leaves out any record the user can't access.
A predicate applies to one dataset, so you define a predicate on each dataset where you want to restrict access to records.
If no row-level security is applied to a dataset, any user who has access to that dataset can view all of its records. For Analytics for Manufacturing, that means sales agreement and account forecast data is fully open to every user who can reach those datasets. So when every user who can open a dataset sees all of its records, the most likely cause is that no security predicate is applied to that dataset.