I like to narrow by engaged shoppers and online shopping together, it doesn’t do much but it does weed out maybe 10-20% of your audience size. In my personal opinion when selling online I’d rather someone have online shopping as an interest than not having it. Engaged shoppers only is too specific for me and only limits the audience to 7 days of them clicking shop now on an ad. You could effectively do online shopping only as that would prob include all engaged shoppers anyway.
I haven’t done any explicit testing to see which works better and I would assume the results would vary case by case. I start with the narrowed adset and once I’ve validated that I’ll generally do a copy adset without the narrow.
One of the ways I sort of validate an audience is by comparing the size change between no narrow, narrowed by engaged shoppers only and narrowed by online shopping only. If I see the two narrowed versions are somewhat close in size and don’t drastically limit the original audience I look at that audience as having higher potential.
Again, nothing has been formally tested which I should do and will do one day. It’s more so a habit now. I might look back at all my past data and see if I can calculate something.