One thing that's confused me for sometime about how demographics are used in TV is that, well, "the sort of person who would watch X" seems like it would be a useful category that could be targeted. That is to say, you could optimize for "boys will watch this show, and boys will buy these things"; but what about the more straightforward "people who will watch this show will buy these things"? Instead of trying to go for this 2-step correlation sort of thing.
I can't help but think of the phenomenon of network decay -- say of the Sci-Fi network, for instance. Sure, you could notice that your science fiction channel is watched by 18-31 year old males, and try to more heavily target that group; or you could notice that it's watched by science fiction fans and target that. I would naïvely expect the latter to be a more sensible way of breaking things down in this context and thus more effective, but the former seems to be what occurred. Am I missing something?
There are likely a lot more advertisers with money to spend marketing products to sell to 18-31 year old males than to very specifically sci-fi fans.
I can imagine beer, clothing, cars, dating and related (shaving, hair and body care products) would fit 18-31 male audience well. Some of those are likely more valuable markets than straight sci-fi demographic products, I'm imagining.
It may well be that a loose 18-31 male demographic fit is more valuable than a tight sci-fi fit.
This is also often given as the reason radio stations are so homogeneous inside of their genre- if you don't fit the mould, you can't sell advertising slots. It leads to weird and perverse outcomes, but it certainly seems like people who refuse to bend go out of business.
I can't help but think of the phenomenon of network decay -- say of the Sci-Fi network, for instance. Sure, you could notice that your science fiction channel is watched by 18-31 year old males, and try to more heavily target that group; or you could notice that it's watched by science fiction fans and target that. I would naïvely expect the latter to be a more sensible way of breaking things down in this context and thus more effective, but the former seems to be what occurred. Am I missing something?