He does focus groups. It’s all about getting more detail, specifics, and answers to open ended questions to (in theory) get deeper insights from a smaller group of people instead of more generic metrics from a larger group. Qualitative data rather than quantitative data.

I wouldn’t be surprised if he got those comparisons from participants. But (without actually watching the interview) I would have to assume that he’s using those statements to represent mildly negative sentiments about Harris from Republican men, and contrasting it with much more severe negativity among Republican women when it comes to Trump.