This is a bit technical, but...
Start by creating an aggregate function. You can prototype this by starting with a table filtered to only one Month. Next you would group by both Store and Service and then add a aggregate of sum of Revenue. From here you can do a merge of the table with itself using the Service name as the join key. This will add an inner join column. By clicking on the right corner of the join column, you can choose to aggregate by sum of Revenue. This gives the total revenue per service. Getting the percentage is simply adding a column with dividing the revenue of a store by the total revenue of the service. To add rank, sort the table by percentile descending followed by adding an index column (note in case of tie, an arbitary decision is made as to what is higher).
Now that you have a prototype, enter the advanced editor and add to the start of the query:
(monthTable) =>
Delete the steps where you filter the table to a single month, and replace the following's step's reference to the step you deleted with monthTable. This will give you a function that applies the transform to any month. Change the name of this query to ApplyMonthlyStats.
Finally, in a new query, take your table, perform a group by and choose the count aggregate on revenue column. Then in the formula bar, replace "each List([revenue])" with "ApplyMonthlyStats". Now expand the resulting table column, and you are done.
