Stay with this term: Big Data . Big Data, translated into Spanish as “big data”, refers to grosso modo , to the set of structured and unstructured data that all users are scattering throughout the network that, in isolation, mean nothing, but that, common and processed positions together with the data of more users, can help find patterns and improve decision making by companies and institutions. Very complicated, right? Let’s see it with an example.
Imagine that you go to a bank to Ask for a loan to buy a house. The bank looks for your profile and analyzes your data. You know that you have been looking for microloan companies on the Internet for a few months, that you have not paid your last electricity bill, that everything you buy is financed and that, in addition, you use a mid-range Android phone . It turns out that, by sharing your data with the data of thousands of users, people with your profile tend to have problems of economic solvency and, therefore, may end up not returning the loan, so the bank reject your request
It sounds terrifying, right? What if we told you that this is completely possible ? What if we told you that the cell phone that you carry in your pocket or from which you read these lines could be a decisive factor when a bank lends you money? Keep reading, because this interests you.
Although the choice of Android or iOS is completely personal, acarried out by the National Bureau of Economic Research American suggests that is a very accurate indicator to predict consumer behavior . As accurate as the credit scores (FICO score) of a lifetime, without going any further. To reach this conclusion, the researchers analyzed 270,000 purchases made between October 2015 and December 2016 in a German website that allows users to buy furniture and finance them (that is, pay them in installments or later).
This store used the fingerprint and tax information about its visitors to qualify if they can or can not access a loan, although it is not the only one that did it, at least in Europe.
The researchers took into account 10 data that users provided passively , as the device and its operating system, the time at which they accessed the web and where they had done it, among other things. They left aside other conventional factors that retailers usually take into account when granting a loan, such as, for example, if the user has returned the integrity of the loans that had previously been granted.
The conclusions are devastating: the difference in Default rates between Android and iPhone users is “Equivalent to the difference in the default rates between an average FICO score and the 80th percentile of the FICO score” . Translated in plain language, If you have an Android you have all the ballots so you can not return a loan , Or what is the same, Having an iPhone is synonymous with economic solvency.
More interesting facts Customers who buy through mobile They are more likely to default on their payments than those who buy from the computer. Those who use emails completed in Hotmail or Yahoo , and that has numbers in their mails ([email protected], for example), have a higher default rate. Those who misspelled their mail missed their payments 5.09% of the time. Finally, half of those who came to the web through price comparison websites also failed.
On the other hand, Tobias Berg, principal author of the study , he says, as they collect in, what “Most consumers in Germany do not know that information, such as the type of device they use, is sometimes taken into account in loan concessions, even though it is explained in the terms of the service agreements of the retailers Almost nobody reads that, and nobody really understands what it literally means ” .
Will this FICO scoring system come into being at any time? Possibly not , at least not by itself. Think that this can be manipulated relatively easily. I can not have a hard but have a second-hand iPhone, an email without numbers and have reached the web by my own means. That could “cheat” the system , so these fingerprints, by themselves, are not entirely reliable. The ideal, according to experts, would be a combination of both.
It seems that Big Data is evil , Nothing could be further from the truth. The benefits of Big Data, like everything in this life, can be misappropriated and used for questionable purposes, but they also have a huge friendly face . Through data analysis you can help diagnose diseases more quickly , build hospitals in key sites, improve tourism, the management of cities and government decisions, etc.
Big Data is the future , and together with Machine Learning and Artificial Intelligence, big data are destined to change the world. The case of the banks is merely anecdotal. It can not be implemented in the medium-long term, much less in the short term, and if it was done, it would have been polished so that it would not be so excessively discriminatory.