Revenue Total

Nice one Vvelarde - mystery solved!!

Anonymous - So basically the cross-filter direction of relationships means that selecting items in one table filters out items in the other table in the direction of the filter. So, let's do a simple example:

Table A

ID, Revenue

A,  100

B,  200

Table B

ID, Products Sold

A,  4

B,  2

So, if your filter direction is A->B

And you place "ID" from Table A into a table visual and SUM of Product Sold from Table B, everything will work correctly. The row context of the table visual will filter Table B for the A row to the A row in Table B and same for row B. Everyone is happy.

But, if you place "ID" from Table B and SUM of Revenue from Table A, not so happy. You will get 300 for both rows. This is because nothing is filtering from B to A and thus there is, in effect, no relationship between the two tables from that perspective.

By switching the filter direction to both A<->B then everything works the way you would think they would in both cases above.

Note, cross filter direction is important to know about and more complex models will sometimes require one or the other, there is no 100% this is always how it has to be. So, generally uni-directional relationships work just fine but in more complex models, sometimes you have to switch it to "Both".

Let me know if that explains things sufficiently.