It turns out that there is not an easy way to do this. Check out Slicing by Aggregate | VizPainter for some different options. None are perfect, but will get you close to what you want.
I'd be happy to answer any questions you might have about the different approaches.
From what I could see the source of your problem was the % conversion calculation which you had set to work at an aggregate level instead of at a row level for each Vendor. By changing the % Conversion formula to simply [Converted Costumers]/ [Possible Costumers] you can now create a view counting the number of vendors by group.
Note: Groups will only display when there is a Vendor to be counted in the group. Currently the only way around this I know of would be to create a variable for each group.
Good catch! That works really well as long as the vendor can be split into different groups across months. Based on the way the view was filtered, that may be a safe assumption. But If the % Conversion needs to be done at an aggregate level across months, then the original calculation is correct.
As far as showing each Group -- this can be accomplished using the Analysis --> Table Layout --> Show Missing Rows option. That will cause Tableau to determine the full domain of the groups dimension and show a row for each group (within the entire data source or context, if using context filters), even if there are no values in the current view. This only works for fields used as dimensions, so assuming the calculation can be done at a row level, it is a elegant solution.