First, let's correct some dimensionality issues. Instead of doing the following:

[Dim Phone Logins].[Phone Logins].Members - [Dim Phone Logins].[Phone Logins].[All]

Just step one level down:

[Dim Phone Logins].[Phone Logins].[Phone Logins]

The real issue with your query is the calculated member. Imagine it as a pseudo-cursor, executing for every combination of Phone ID and Employee Global ID. While you are attempting to make the calculation aware of where you currently are in the Phone Logins dimension (one level too high in the attribute hierarchy though), you are not doing the same for the Employee dimension.

In my opinion, your calculated member should look something like this:

count

(

nonempty

(

{[Dim Employees].[GlobalID].[GlobalID].currentmember}

* {[Dim Phone Logins].[Phone Logins].[Phone Logins].currentmember}

,[Measures].[Outliers - ACD Calls]

)

)

The above will only work for a query that returns the members of these 2 dimension attributes, otherwise there may not be a "currentmember". If you envision excluding one of these attributes from the query at some point, you may need to step down to the appropriate level by using the descendants() function.

Not knowing exactly what your cube structure looks like, are you not trying to return a count of fact rows here? If so, you could simplify the calculated member as follows (which will work for all combinations of dimensional attributes):

count

(

[Measures].[Outliers - ACD Calls]

)

Hope the above helps...and hope I didn't miss anything syntax-wise. Please post the solution once you have it working.