You probably know that in today's world, numerous companies are hunting for your information. You only have to talk to a friend over a cup of tea about a new iron, and within a couple of hours your smartphone will start cautiously offering to buy a modern version of the appliance through various applications.
It would not be a big exaggeration to say that technology is literally watching you. Of course, not you as a person – that’s too much. But you as a consumer. Everyone you allow to “see” data about you is trying to analyze it using algorithms and sell it profitably to someone who absolutely needs it. There are no trifles in this process – where exactly you are, what services you use, who you are friends with on social networks, how you travel, how much you talk on the phone – everything can be analyzed and some conclusions can be drawn.
You don't have to go far for examples. For example, did you know that Ukrainian mobile operators offer financial institutions, insurance companies, etc. a ready-made scoring assessment of their subscribers? This is a conditional score that is calculated based on the pattern of your behavior and indicates your creditworthiness and reliability as a borrower. I can't say how popular this service is. But it exists.
Products will tell more about you
But there is one rather serious gap in the “magical” world of big data – your real consumer behavior. Of course, the bank whose card you use can see how much and how exactly you spend money. Data on expenses can be collected in terms of MCC codes, which are assigned to payment terminals of individual points of sale in the economic segment. By the way, it is they that are used to track cashback categories.
The bank knows that the 200 hryvnias you pay by swiping your card at the terminal with MCC 5992 means that you bought flowers. And 3,000 hryvnias with MCC 7538 is a payment for car repairs. But not all MCC codes make it possible to understand what exactly you bought. For example, code 5411 is given to supermarkets. But you could run to it for diapers for the baby or for beer for a party on Friday. Maybe you needed batteries for your car alarm, or maybe hygiene products. The bank cannot know this.
This information can only be seen in detail by the store itself, provided, of course, that you use its loyalty program and have a discount card. Usually, this data is used to encourage you to make repeat purchases. If you regularly buy “baby hygiene”, you may be offered a promotional product from the corresponding category, hoping that along with the “tears-free” shampoo, you will fill up a full basket of other products.
But as researchers from the American University of Notre Dame found out, your purchases at the supermarket can tell a lot about your character and are significantly correlated with your credit rating. Therefore, by understanding what exactly you buy at the store, you can accurately determine your solvency.
Supermarkets can sell information about you
To test their hypothesis, the researchers obtained information on over 30,000 consumers from a large transnational conglomerate. The value lies in the fact that the data provider is both the owner of a supermarket chain and a financial institution that issues credit cards. Thus, the researchers had data from two different economic segments that could be compared in a relevant way.
In subsequent work, they tested a bunch of hypotheses, trying to identify certain patterns. And some "food" habits were very clearly correlated with the diligence of servicing credit debt.
For example, the "healthier" the consumer basket was, the more disciplined a person was in paying off the loan.
“Buying healthier but less convenient foods involves responsible behavior when paying on credit,” the university’s press service quotes associate professor of marketing, Jonghyuk Yang.
Cigarettes or energy drinks on a check are associated with a higher likelihood of credit card defaults or defaults, while milk and other raw foods indicate a high level of borrower reliability.
In addition, the authors of the study found that disciplined payers were “stable” consumers – those who, regardless of their specific grocery basket, come to the supermarket on approximately the same days, spend approximately the same amounts and buy the same brands. Moreover, these findings worked especially well for people for whom there was no credit history.
The latter issue is the most “painful” for lenders around the world. And so further work in this direction could soon create a completely new market: supermarkets could quite realistically calculate their own credit rating for their consumers and sell it to financial institutions.
They will be helped in this by the latest artificial intelligence algorithms, which are quite capable of learning to determine the borrower's solvency without making any hypotheses - simply by comparing consumer behavior and the thoroughness of fulfilling credit obligations.
