I have inherited a cube and star schema at my company and need some MDX help
for attrition and churn reporting. The cube is a daily snapshot of a
production system as follows - it has a load date dimension with no <all>
level. Each load date will give you current active customers and cumlative
disconnected customers. So for a given date I can say we have this many
active customers and all total have lost this many. The cube also has a
"end date" dimension that will tell you when a customer left. So if they
are active it is blank and if they disconnected I can tell which day (ie -
Nov 2004 10,000 disconnected). They want to start doing attrition and churn
reporting out of it. I can do it writing SQL fairly easy, however it of
couse is not very dynamic.

For attrition or churn I can grab the number of actives 30 days ago, however
how can I get the number of disconnected since that day 30 days in past? I
would need the 30 days of disconnected to change based on the load date.
Thanks.

Kevin Brooks