Just saw this interesting video, a talk by a woman (Nicola Twilley) who created a scratch-and-sniff map of New York. In it, she talks about the ways different demographics respond to scents. For example, Asians like the smell of rose and orange peel more than white people, who prefer eugenol. Everyone likes vanilla. Men like the smell of guaiacyl acetate, a woody-smoky smell, and women like cis-3-hexanol, the smell of cut grass. (So, she suggests, women should wear smoky scents and men should wear grassy ones, assuming they're heterosexual that is.) She also notes that almost everyone is anosmic to at least one thing, meaning they can't smell it all.
Nicola Twilley at Gel 2011 from Gel Conference on Vimeo.
So far so good, and Twilley admits she is not a scientist, but she goes too far in concluding that "we all live in a separate smell universe" -- she says that we all see the same colors and hear the same sounds, but we don't smell the same smells. I think this is sloppy. The research she's referring to doesn't suggest that we experience different smells, it just speaks to different preferences. You'd find preferences for different colors and tones among different demographics and cultures too. If we can assume that everyone experiences a certain wavelength the same way, leaving associations and baggage aside, we can assume the same for smellable molecules. (There are holes in everyone's visual and auditory capabilities too.)
Her two big examples don't help her argument much: She says that some people in the perfume industry describe eugenol as sweet and carnation-like, while others describe it as spicy, like clove. "How can this be?" she wonders. Here's how: eugenol smells both sweet and spicy, as cloves do -- and carnations smell like cloves. They are not different descriptions, they're just both incomplete. Most people need training/experience to both recognize and accurately describe smells out of context. She also says that she has a selective anosmia to Galaxolide, a synthetic musk. In truth most people are anosmic to some types of musk because they're very large molecules (the effective equivalent, I suppose, of very high-pitched tones, which not everyone can hear). But it would be wrong to assume everyone's selective anosmias are totally different, like your neighbor on one side can't smell bacon and the one on the other can't smell garbage. Evidence suggests that most of the smells people are anosmic (or hypersensitive) to were created by humans.