Complications in making a dashboard user-specific

Hello,

I wanted to bring this case up because we spent some time trying to figure it out and contacted Support (because it uses personal info, so I can’t post screenshots here) but ultimately were not able to use out-of-the-box tools and had to create another SQL query / data source to solve this. I’m thinking having something like a User() function might make this and other problems easier to solve within Dundas?

I was working with a data source where each row is a sent card (called a “pat on the back”) with a send date. Each card has a sender and recipient (and their respective supervisors), and they are all Associates in the company I work for.

We were trying to replicate two dashboards from Tableau, one that is a user-specific view of the user’s sent cards, and one that is a user-specific view of that user’s direct reports’ sent card statistics. One of the metric sets in each of those dashboards shows the recipients the user (or their direct report) has never sent a card to to remind users to send those folks a card! Because every row in the data belongs to a sent card, this cannot be done by filtering by sender. (Note: everyone has received a card, so the list of recipients captures everyone possible to send a card to) I accomplished this in Tableau using a field called User Has Sent POB :

{FIXED [Recipient Name + Username] :

MAX([User Is Sender])

}

The User Is Sender field contained there is UPPER(USERNAME())= UPPER([ senderUserName ]) which pulls in the current user, so for each recipient, User Has Sent POB returns True (the max value) if there is a card the user has sent. Because this is a fixed expression, it will give a True/False value for each recipient as to whether the user has ever sent them a card, changing the level of detail from card to recipient. The sheet is filtered to User Has Sent POB = False, so it just pulls in the recipients where the user has never sent them a POB.

We ended up having to create a separate query with all the possible sender + recipient pairs to compare to the pairs in the card data when filtered by user, which takes a lot longer to run than this fixed expression, but it would be lovely to be able to work with the user in the data to create Boolean fields, or have some other way to do things like this!

Hey,

As we discussed generally dashboards can be made user specific by making use of security hierarchy in cube level, which in your case is not useful because the whole dashboard cannot be made user specific. However, I could think of another possibly easier way to do this by passing user/sender names to the cube - a new one, the details of which I sent you via email. Let me know if that is a feasible solution.

1 Like

Thank you, this will help us understand the options for data prep in Dundas better!