Originally Posted by
l'etoile
Clearly they developed a model about who lives in a specific community, but I don't think it's rocket science. On our own each of us could probably pick a community and come up with some spot-on observations. People in high-wealth communities read news, vote, participate in the arts, value education, travel more, buy european cars, etc. Package it pretty and - viola!
I guess I'm saying it fits, but they only know me in a broad sense.
I do think that there's more to it than just zip-code though. Yes, that's a big part of it, but experian has access to much more fine-grained data. Experian claims "Experian Marketing Services’ Mosaic® USA is a household-based consumer lifestyle segmentation system that classifies all U.S. households and neighborhoods into 71 unique segments and 19 overarching groups" and claims to identify 98 percent of households.
People seem to jump from "My address is listed as my parents' and I get their classification, so the system must be zip-code based" which seems like an enormous leap of logic to me vs. at least considering the alternative of there being at least some household-based component. Experian explicitly claims that household-level data is available - I'm not sure there's compelling evidence to doubt them on this.
If you have rich parents and want Delta to think you're American Royalty, set your address to theirs...
My personal experience: as an under-35 couple my wife and I were tagged with a code representing 90% under 35 in a neighborhood that is roughly 90% 45+ and mostly 50+. The code they tagged is a reasonable fit for the few people of our age in the area, but there are all of 2 or maybe 3 households nearby (including ours) that fit that mold. Either Experian did an epically poor job of stereotyping our neighborhood or zip code, or the results are much more fine-grained than that.