Compare shares across dimensions
If you want to see the share of paying users in a city out of all paying users, you can use a single-event filter to filter only the numerator by City is Beijing and get the result. You can also copy the analysis metric multiple times to compare the shares of all cities at the same time.
If you need to compare many dimensions, you can also use Event Breakdown, which compares the shares of different dimensions with the same logic as Group by. Event breakdown supports only event properties as breakdown properties. For multiple events, the available breakdown properties are the intersection of their event properties.
Unlike Group by, which applies to all events, Event Breakdown lets you specify the scope it applies to. If you want to compare the share of new users across all channels, you can apply it only to the numerator. You can also add another analysis metric to view the number of registered users and the share for each channel at the same time.
If you want to compare dimension data with a hierarchical relationship, such as the share of each city within its province, you can use group-by and event breakdown together. By default, Before group-bys, is selected, which shows the group values of the breakdown property first. Depending on how you prefer to read the data, you can switch to After group-bys, to show the breakdown property last.