In math we trust: the analytical secrets of credit card companies

Introduction

I have a heightened sensibility when it comes to recognizing the use of data mining and analytics.  Having studied computer science in college, I am fairly knowledgeable of those disciplines. So it was no coincidence that I was able to draw attention to the ill-conceived data mining practices of American Express.  Ultimately, my public relations salvo against American Express was not about me being a victim; instead, it was about the credit card company understanding the need to change its algorithm.

I am not sure whether American Express made any adjustments as a result of my efforts and the ensuing public outrage—it said it did—but I do believe that all credit card companies learned a sobering lesson: Mathematical models built on dubious assumptions can and do backfire.  In fact, one of the reasons I think my story was so popular—and continues to be today—is that the current economic crisis occurred, in large part, due to our increased reliance on and faith in esoteric and erroneous mathematical models.

During the height of my notoriety, I received hundreds of e-mails from concerned consumers, asking me and giving me insider information about how these models and algorithms work.  I responded to each inquiry and often posted these responses in detail on my blog.  However, I have never delved into the mathematical concepts of data analysis with an emphasis on how a credit card company might use them. Well, not until now. 

My desire to explain some basic methodologies and models on this blog comes from the realization that we live in an increasingly data-driven society.  From internet searches to phone records, from credit card purchases to tax returns, companies have access to huge data sets which enable them to, in many cases, know us better than we know ourselves. The access to and interpretation of such data empowers companies beyond belief. This tremendous power should go neither unchecked nor overlooked, especially by the very consumers who are often the subjects of its manipulation. 

The next few posts will introduce you to the world of analytics.  They will give you confidence to challenge its applications and assumptions.  But most importantly, they will help you gain more control over your destiny, which is increasingly determined by the use of mathematical formulas.

Comments

[Anna]

I found your blog extremely useful. There is still a great credit freeze and banks are not lending money to the middle class. Their lending practices are absurd and Americans are working hard and not able to purchase a home in return. There are also great issues in collection practices. I have been harassed by credit collection agencies for accounts I am not responsible for - their practices are in large part illegal, and they can ruin your credit. I purchased 3 books on the topic. I think there is a great deal of help you can offer to Americans seeking credit. Perhaps you can offer credit counseling? Maybe pitch your story to news channels or national newspapers and get a little segment? Consumers often feel as if they have done something wrong, when the truth is that the banks and their affiliated organizations can be crooked and unjust. It helps consumers to know they are not alone.

[SP1MURAD12]

Hey i am murad hossain .i am a simple internet user .I used many web site but among them a know a useful and helpful that is dreamtouch.biz . we can use it great to our business. we also can publish many kinds of advertise in this site for free . It is very important for any kinds of business. so we have to visit this site for our business value. Ok ………………..

[Sue]

When is the series coming?

[sarah h.]

i'm looking forward to it. awesome job kevin.