[Interesting article from the consumerist.](Your Credit Card Company Is Building A Psychological Profile Of You) What do you think your purchases would say about you?
I don’t have a credit card, but I find the snow rake interesting as a “safe” purchase. Mostly because I used my debit card to buy one for my dad 
Can I ask a question of our northern posters? What exactly is a “Canadian Tire”? I get the impression that unlike US stores with the word “tire” in them, it’s not a store that mostly sells car tires.
Canadian Tire is the canonical Canadian auto parts and hardware store. It sells car tires. It also sells car parts, car service, lawn seed, kitchenware, power tools, magazines, plumbing parts, batteries, electrical sockets, lumber, bicycles, hockey gear, safety goggles, drain cleaner, small appliances, cheap TV sets, and all sorts of other manly items. It is a Canadian ritual for a father to take a son to Canadian Tire on a Saturday morning.
Is there a link to the article?
Marketplace had a similar story today. Both splurging and going low-budget are seen as red-flags. Splurging is seen as possible stress-relief from financial worries and being frugal as already being in trouble. It seems like such an inexact method and more like an excuse to cut credit lines.
Ah, okay. So more like the Bennys’ stores in southern New England than say, K-Mart.
Weird. I’ve lived in Connecticut for 39 years and I’ve never heard of Bennys.
Sorry for the messed up link here
I figure that they are probably already selling information to insurance companies, so I rarely run bar tabs or buy beer on credit cards.
Huh. In the Netherlands, no-one uses a credit card in every days shopping. But companies want the same kind of info. So they’ve devised two electronic cards, "bonus"and “air miles” that a consumer is supposed to present at each purchase, in order to get some infrequent discount.
So they have all this info on who buys what, and are supposed to tailor advertising to that info, although so far I haven’t noticed anything like that.
When the airmiles card started out, there was a cartoon of one character using a big jar of vaseline to grease up a pair of ice skates, and the other opening and reading a piece of direct mail for gay sex toys.
I am a statistician/data analyst for a credit card company. While there is some truth in this article, much is grossly misleading.
Canadian Tire is an unusual example because the merchant is able to use SKU-level data in its credit and marketing analytics. Other single merchants and analytics services (like Tesco and Dunhumby) can do much the same thing. Dunhumby famously delivered beer offers to males who purchased diapers, hypothesizing that new fathers had little time to go to the pub. This was a magnificent success.
But for the most part, we at the gigantic global credit card companies simply have no idea what people actually buy. Merchants simply do not pass this information to us in general.
They many not pass along details, such as what’s internally available at Canadian Tire or with your grocery store loyalty card, but they can still paint a pretty good picture.
If a person suddenly starts using their credit card at the grocery store, they may be on hard times and don’t have cash on hand to buy food.
If they have been using their card all along at a high-end place like Whole Foods and suddenly stop or switch to using their card at a regular grocery store, they may be in a tight spot.
Now, take a look at their gasoline purchases - going from high-dollar fill-ups to ten or twenty bucks here and there is a classic sign of being short on funds.
And so on…
That’s okay. When I went to the website I found out that there are 28 more of them than I thought. My grandfather used to bring us to the one in Raynham, MA on a regular basis, and I’ve seen just one other one. I think they’re well hidden!
This is actually not at all how it works. But please feel free to continue speculating. I wouldn’t want to ruin your day.
A particularly frightening application of this type of analysis is that done by casinos. Check out the “patron potential” info here: http://www.diamondstream.com/casino-data
What I got from the Marketplace piece is that the credit card companies are using association rule techniques to find potentially “high risk” behavior over lots of data. Of course the story personalized it some. But are you saying credit card companies trying to reduce default risk in their customer pool wouldn’t mine their data for behaviors more associated with later late payment or default?
You know, I read that and laughed at certain dive bars being full of deadbeats, and purchases from liquor stores being an apparent negative indicator as well. And then I grimaced when I supposed creditors might not like me because I happen to like the sauce.
I suppose frequent card use in Las Vegas and at horse tracks would point to high risk. Big bills from Honest Charley’s Speed Shop might not augur well, either. 
I find it ironic that the link to the story also includes another story about how people believe what they read online, second only to recommendations from known people.
So as far as building a profile? I’m glad Maeglin chimed in - it seems to me that all the credit card company knows is what store, and for how much. There is no itemized list (that I’ve ever seen) of items when you sign the paper slip.
To some degree, I want them taking the occasional glance at my purchases. I do what I can to shred personal information that comes through my mailbox, but I cannot control everything. More than once I’ve been called by the credit card company for a charge that looked suspicious (turned out to be legit - but still).
Of course credit card issuers use data mining to reduce the risk of default. It is simply not at all in the manner the Consumerist article or gotpasswords supposes. We also use data mining techniques to manage exposure to potentially insolvent merchants as well. I am not on the credit risk side of my business but due to my profession, but I am acquainted with the techniques. I just don’t have a lot of knowledge about our proprietary techniques. These are unique given the different nature of my company’s business.
What I do not like about the original article is that it suggests that the techniques used by Canadian Tire are practiced more broadly across the industry. This is patently false.
Also patently false are the logic and unempirical narratives that gotpasswords suggests. For example, my credit card company spends a princely sum in marketing to incent people to use their cards in grocery stores. It would be a strange twist if our marketing succeeded wildly but our credit analysis forced us to cut the credit lines of our grocery store marketing responders. On its face this makes no sense, and fortunately for my livelihood, it does not happen.