This is an example of filtering data using summarized information. This applies to a use case where the goal is to perform some action on a subset of things, because they happens to fall into some category based on stats from all the data. For instance, this might apply to an email voting scenario, where you want to respond to the emails in the “winning” category, based on which option got the most votes.
The approach is to branch the entire set of items straight to a merge, and through a summarization to the other input of a merge, to “augment” the original items, with the corresponding summarization info, using a matching data field. The combined data includes each original object, and the matching summarization fields. The workflow below shows the rest of the details.