I'm not sure I understand what you mean with this statement: "the biggest bubble will correspond to the part made most by these companies"? Your data set doesn't seem to have details about the quantities produced for every product, rather it only shows the quantities produced for every company (there are not enough details). So, I'm taking a wild guess here in saying that you want to size the bubble by the number of companies producing certain product. Can you confirm that's the case?
The solution will likely involve re-shaping the data either way. You'd need to add details by product (more rows) or to have a column for each product (that you can then pivot and count the number of companies producing a product). An example for the second option is attached.
How to Pivot:
Let me know if you have any questions.
Split&Pivot.twbx 151.3 KB