YouGov has released a new website that collates data to give a demographic portrait of people who associate themselves with certain brands.
The site, YouGov Profiles, brings in figures from "the UK’s largest connected data set" to try to understand the consumer behaviour (media, retail, food, politics etc) of certain types of people.
While it's great for marketing companies who can "mine its huge seam of information in real-time and understand more about their audiences than ever before", it also allows everyone else to find out what see what differentiates fans of [insert name of company] to everyone else - even giving a little cartoon character for each one.
So while readers of The Independent are especially likely to be twenty-something males who eats Italian antipasti and likes cycling, Daily Mail readers are more likely to be a women in their 60s who eat cheese and tomato sandwiches and like golf.
Fans of Premier League champions Manchester City are also especially likely to be women over 60 who shop at Lidl and listen to Oasis, fans of cross-town rivals Manchester United are also likely to be women, but are more likely to shop at Asda and like Simply Red.
YouGov finds that a supporter of Ukip is especially likely to be a man in his 60s or above, who thinks "the country is going to the dogs" and likes the music of Cliff Richard.
Meanwhile, a Liberal Democrat voter is especially likely to be a man in his 20s, who eats aubergine parmigiana and potage bonne femme (posh name for vegetable soup), say "the internet is my main source of information" and like the music of Pentangle.
Speaking of Lib Dem voters there are obviously some concerns about small data samples (there were only 782 Lib Dem voters available), and trying to find something "typical" about a person is an inherently difficult task owing to the complexity of human nature.
Which could help explain why Taylor Swift fans are especially likely to be 18-24-year-old men from Scotland who vote for Ukip all at once...
However, YouGov are keen to point out that the profiles do not show the "typical person" but rather what is "particularly true" about a group.
"For example, if something is only true of 1% of the overall population, but is true of 6% of our target group, it might score very highly (and shows you something interesting and true about that group)," the website explains.
Try the profiler for yourself here.