Loading...

Puzzle versus Mystery

Published: June 24, 2010 by Guest Contributor

By: Kennis Wong

Several weeks ago, I attended and presented at Experian’s sold-out annual conference, Vision, in Phoenix, Arizona. One of the guest speakers was Malcolm Gladwell, best-selling author of The Tipping Point, Blink, Outliers and What the Dog Saw: And Other Adventures.

Since I’ve read three of his four books, I could be considered a fan. And yes, his hair did look as wild in person as it appears in the pictures on the insides of his book covers. But that was not why I was so impressed by his speech. The real reason was that his topic was so relevant to how Experian Decision Analytics delivers value to our clients.

Gladwell spent the whole hour addressing the difference between “puzzle” and “mystery”, providing abundant examples for both. The puzzle-versus-mystery topic was from one of his articles in The New Yorker. To solve a puzzle, one or more pieces of information are needed. The source of the problem is that insufficient data is available to have a conclusive answer to the question. An example would be finding Osama Bin Laden’s whereabouts. We simply do not have enough information to locate him, and we need more intelligence.

On the other hand, a mystery is not solved by simply gathering more information. It is a matter of making sense out of a massive amount of data available, using analysis and judgment. Enron’s creative accounting was an example of a mystery. All the information was out in the open. Pages and pages of SEC filings and annual reports were there for anyone who was willing and able to analyze them. All that was needed to solve the mystery was to make sense out of the data.

In the Fraud and Identity Solutions team, we satisfy clients’ needs by providing solutions for both puzzles and mysteries to fend off fraudsters. Besides the core credit bureau data, we have demographic data, fraud consortium data, past application data, automotive data and much more. We also have strategic partnerships to deliver demand deposit account, cell phone, and device data. All these data sources ensure that our clients get the data they need to piece the puzzle together.

Our consulting and analytics, on the other hand, help clients to solve mysteries. Looking at individual pieces of disparate data is inefficient and provides little or no value. That’s why our numerous scoring solutions combine the available data in a way that is most predictive of various fraud outcomes. For example, our Precise ID Score and Fraud Shield Score Plus predict first- and third-party fraud; our BustOut Score predicts the likelihood of bust outs; our Never Pay score predicts the likelihood of a consumer never making a payment. As more data are available, we incorporate them into existing or new models if it increases the effectiveness of the models.

So we have both the puzzle and mystery grounds covered.

A note to Malcolm Gladwell: Great job at Vision! If you write a book about this topic, I’ll definitely buy it.