It would be require to dive very deeply into your tables to be able to give you concrete advices - therefore it's more generic with the main-goal to reduce the number of fact-tables and if possible to avoid any link-tables and rather to create table-associations.
For this I would concatenate such fact-tables like sales and planning - you could harmonize this by creating a table-source field like 'sales' and 'planning' and the values are within a field value - but even without them and leaving an asymmetrical table with NULL's will work fine. Maybe the coupon-table and the prescription-table could be (after a crosstable-transforming) be concatenated, too.
Other fact- and maybe also dimension-tables would I try to join or to map to these fact-table. I'm not sure if it's possible and also useful to try to get everything in one big fact-table. I mean with them that I have the impression that your datamodel might be in reality two or more datamodels and it could be sufficient to leave them as loosely coupled datamodels (maybe associated with only a date and one further key) to get your required views between them (and be cautious to match not the wrong dimensions within your charts).
- Marcus
