Getting Same Store functionality to work across prior periods, not just current

Question

I have a basic Sales table, with a Store Key, Date Key, and dollar amount.  I also have a SameStore table, which for every combination of StoreKey and DateKey, says whether or not that store should be considered "Same Store" (was operational
last year for the same time period) on a month to date, qtr to date, and YTD basis.  The MTD Y/N, QTD Y/N and YTD Y/N are available via the same store dimension.  Via my date dimension, I have all the usual date calcs, such as prior month, prior
year, QTD, etc...
It is very easy to query the totals for all of my stores as of Sept 2014 and get the current sales amout, as well as get the prior month, prior year, and so on.  Now, if I only want to see the totals for stores that should be considered "same store"
as of Sept 2014, I can throw the MTD=Y up into my filter and then the current month's total will adjust to only show stores where their same store flag for Sept 2014 is set to a YES.  All of the NO's for Sept 2014 get removed... however, if one of those
NO's was considered Same Store in August, their $$$ number is still being incldued in the Prior Month date calc.  I don't want it there, becuase as of September (the focus of my report) they are not considered same store anymore.  
I am getting this:
As of September 2014, Same Store=Yes
Store Current$
Last Month$ Last QTR$
1 (closed 8/15) $0.00
$100.00 $125.00
2 $50.00
$75.00 $60.00
3 $60.00
$65.00 $85.00
But I really want this:
As of September 2014, Same Store=Yes
Store Current$
Last Month$ Last QTR$
1 (closed 8/15) $0.00
$0.00 $0.00
(or this whole row can just be gone completely)
2 $50.00
$75.00 $60.00
3 $60.00
$65.00 $85.00
Just because Store 1 was considered to be a "Same Store" in August doesn't matter since I am running my report as of September.  It is not "Same Store" as of September, so it should vanish from all my calculations.
Currently, dimStore, dimDates, and dimSameStoreStatus all hang off of the main Sales fact table.  I know this is part of the problem.  We are going to try and connect the dimSameStore to the dimStore and dimDates tables rather than the fact table,
but I not sure if my overall design is right for this or not in general.
Has anybody had to do something similar?  If I get a solution from somebody (or a great hint that helps me solve it), I'll send you a Starbucks gift card. 
Thanks in advance!
Mgr, Data Warehouse Svcs

Answer 1

Hi D,
The way I typically deal with this is to use a table, like you have, to mark whether the store was comparable for that Period.  Then, when you run a query, you are comparing two or more periods.  So, your query should restrict the scope to only
include stores that were marked as comparable in all periods.  
For example, if you are comparing this month with same month last year, you should only include stores that were comparable this month and same month last year.  If you have a query that is showing month on month growth for all the months this FY, you
should only include stores that were comparable for all months so far this year.  This is more common for a rolling n period report.
I find it quite interesting that most retailers use some variant of "Same Store Sales" with the express purpose of being able to compare performance with other retailers.  However every retailer has their own definition of what a comparable
store is.
Hope that helps,
Richard