There are a few ways of calculating percent change, but it's really down to what type of database you're using. Is this a dimensional (e.g. TM1, SSAS, DMR) or a relational (e.g. SQL Server, Oracle, DB2) source? 

If it's a dimensional you'll want to tuple on the previous member of the year. tuple([Measure],prevMember(currentMember([Cube].[Date Dim].[Date Hier])))

Relational is a little more complicated. I would use moving-total([Measure],2) - [Measure]
Moving-total takes the total of the current and n-1 previous rows. So moving-total([Measure],2)-[Measure] will return the previous rows value. 


It gets complicated if you have another attribute in your output though:

Obviously the prev year for 2011,Camping equipment is wrong. So we have to limit the moving-total to window on Year:
moving-total([Revenue],2 for [Product line]) - [Revenue]

Essentially it's grouping each value in Product line into it's own table, and then doing the moving-total like before. Additional data items can be included by adding the item to the FOR clause:
moving-total([Measure],2 for [Product line],[Region]) 

Now, once you have the previous year value, the next step is to make ANOTHER measure. This one does the actual variance calculation, and is the same for both dimensional and relational sources:
([Measure] - [PrevYear Measure]) / [PrevYear Measure]



That will work great in graphs and crosstabs:

Hope that helps!

There are a few ways of calculating percent change, but it's really down to what type of database you're using. Is this a dimensional (e.g. TM1, SSAS, DMR) or a relational (e.g. SQL Server, Oracle, DB2) source? 

If it's a dimensional you'll want to tuple on the previous member of the year. tuple([Measure],prevMember(currentMember([Cube].[Date Dim].[Date Hier])))

Relational is a little more complicated. I would use moving-total([Measure],2) - [Measure]
Moving-total takes the total of the current and n-1 previous rows. So moving-total([Measure],2)-[Measure] will return the previous rows value. 


It gets complicated if you have another attribute in your output though:

Obviously the prev year for 2011,Camping equipment is wrong. So we have to limit the moving-total to window on Year:
moving-total([Revenue],2 for [Product line]) - [Revenue]

Essentially it's grouping each value in Product line into it's own table, and then doing the moving-total like before. Additional data items can be included by adding the item to the FOR clause:
moving-total([Measure],2 for [Product line],[Region]) 

Now, once you have the previous year value, the next step is to make ANOTHER measure. This one does the actual variance calculation, and is the same for both dimensional and relational sources:
([Measure] - [PrevYear Measure]) / [PrevYear Measure]



That will work great in graphs and crosstabs:

Hope that helps!