You won't be able to exclude zeros using the select statement as each row might have a zero for one column but a non zero in another.
So your customer realizes that the N will likely change per question for each store. Nor will they know how may instances per question that it was not answered (=0).
You will haev to make a billion and 1 running totals to get your data with no zeros. If you can talke them into seeing the value in that information it will make the report 100% easier.
If not...
for each question you will need 6 RTs to get your amounts and 5 formula field to get your %.
Maybe there is an easier way but I am not seeing it at this time.
Build it out for Q1 and make sure it works and is what they want before you build it all the way. (I would also consider showing them a version of including the 0s to show them what them an alternative)
RT examples for q1
Name= Q1_all
field to summarize=rsq01
type = count
evaluate=use a formula
table.rsq01 <> 0
reset=on change of group (select store id group)
place in group footer
This gives you your count of all records with no 0 for q1
Next make 5 rts, 1 per possible answer
name = Q1_1
field to summarize=rsq01
type = count
evaluate=use a formula
table.rsq01 = 1
reset=on change of group (select store id group)
place ion group footer
counts only anser of 1
next
name = Q1_2
field to summarize=rsq01
type = count
evaluate=use a formula
table.rsq01 = 2
reset=on change of group (select store id group)
place ion group footer
counts only anser of 2
repeat for all the other values (3-5)
place all in group footer
to get % make 5 formula field
q1 value 1 as {#Q1_1}% {#Q1_all}
q1 value 2 as {#Q1_2}% {#Q1_all}
q1 value 3 as {#Q1_3}% {#Q1_all}
q1 value 4 as {#Q1_4}% {#Q1_all}
q1 value 5 as {#Q1_5}% {#Q1_all}