The Score on Collectors

  Sometimes examining the world around us can help to understand concepts that can apply to other facets of life. And this includes ways to better collect debt.
  Consider this: A male collector between the ages of 37 and 42 whose favorite sport is football will outperform a female collector of comparable age who has no favorite sport when it comes to collecting debt in any city with a National Football League team.
  How do we know? Credit it to something with the unlikely name of the El Pollo Loco theory. The concept, explains its creator, CreditMax CEO Steve Kass, revolves around matching collectors to the environments in which they work best when communicating with debtors.
  The concept came to Kass a few years ago while he was sitting in an El Pollo Loco restaurant. He watched as a veteran cashier, offering each customer a dessert, sold about 40% more desserts than did a rookie cashier, who neither offered nor sold any desserts.
  "I realized what a material impact the person at the point of sale has on the ticket price, and that led me to believe that what's really important in debt-collection arena is the point of sale, when the collector has the debtor on the phone," Kass says.
  For Kass, the collector holds the key to success. So when it comes to improving business operations, it is the collector who should be analyzed and emphasized.
  By scoring the collector, the El Pollo Loco theory bucks the industry standard of scoring the debt. It illustrates not only the industry's increasing reliance on scoring but also its intensifying chase for more accurate methods to predict payment behavior.
  While Kass is zealous about scoring collectors, more traditional scoring gurus say the debt is the thing to score. Regardless of which theory a collector espouses, it is doubtful scoring does not enter into the operations somehow.
  Scoring-rating someone's ability to do something, whether it is to pay a bill or to perform well as a collector-is based on statistics, predictive sciences and common sense. The point is to land the right account with the right collector, and the goal is to determine what pieces of information, or data attributes, accomplish that best.
  "The modeling techniques have been around for a long time," says Arvind Krishnaswami, CEO of Medlytix, a Roswell, Ga.-based collection scoring and consulting company. "It's what feeds the model that's important."
  Scoring models almost all incorporate a mix of internal and external data sources. Internal data include information collected by the agency and the particular way that information is blended with external information to score accounts. External information ranges from generic credit scores to arrest records.
  While many agencies develop their own scoring models, nearly all are searching for more esoteric and predictive sources of information.
  Dan Heisel runs the Cincinnati-based health care collection agency Controlled Credit Corp. Two years ago, he says he fed about 45 pieces of information, or data attributes, into his scoring model. Now he uses 180 attributes.
  The result is a "decrease in the liquidation time line" by about 15%, says Heisel. While the agency is not necessarily collecting more money, it is collecting much faster.
  Simple Start
  The data come from an internal database-Heisel's original scores were index cards saying who paid and who did not-and from such third-party sources as Equifax, the U.S. Census Bureau and the Internal Revenue Service. The accounts come out marked high, medium, low or no-score. Naturally, the accounts that score higher are worked harder.
  "I've been data mining for 12 years," says Heisel. "We started very simply, by saying who paid us and who didn't pay us. It's gotten more scientific and complicated as we moved up. We now truly can see into the financial condition of the consumer."
  Over the years, Heisel says one of the biggest shifts he has noticed is this merging of traditional financial data with client-specific data. "It's a constantly evolving process," he says.
  Applying even simple credit-bureau information to the accounts likely will improve liquidation rates by 25% to 30%, says Krishnaswami.
  Results like that have collectors everywhere getting in on the action. "There's a ton more interest in the space than two years ago," says Steven Seoane, vice president of analytics at LexisNexis.
  Seoane says he has tripled his staff, from five to 15, over the last year to accommodate the growing hunger by the collections industry for scoring models.
  "Scoring providers have gotten more sophisticated. Some are leveraging collectors' notes," he says, referring to the practice of incorporating what collectors type in about their customers. "They look across your collections activity and incorporate it."
  Credit card issuers who got into scoring early regularly score debt to figure out how much they can sell it for. Now debt buyers are doing the same thing, and both sides use scoring as part of the negotiations.
  "[The sellers] run their process, whatever analytics they're going to use, and then we take a look and say, 'what can we do with this?'" says Al Brothers, senior executive vice president for New York-based debt buyers Cavalry Portfolio Services. "It's like we're both using tools to do our jobs and some overlap."
  Public records are another increasingly popular source of information for collectors. "Life-stress stuff is highly predictive," says LexisNexis' Seoane, whose scoring models involve the heavy use of public records. "We're looking at patterns of behavior around moves or interactions. If he lived with a wife in a home and now he lives in an apartment, why is that? The marketplace has been very interested in that as well."
  Tracking change
  Looking at a change in behavior is what Michael Rosenthal, TransUnion director of solutions development for collections, calls "change data." TransUnion's so-called Triggers suite is based on the idea that tracking a change in debtors' behavior helps the collector determine which tack to take with that debtor.
  Rosenthal says TransUnion has built a customized statistical platform from the change data to show how predictive certain changes are. The customer then chooses which changes, or triggers, about which he would like to be notified.
  Indeed, collection scoring is shaped by customization. Most of the models are homemade, so to speak. The more specialized the scoring model, the more effective most collectors feel it will be. There are differences among kinds of debt, and it makes sense that a collector who specializes in credit card debt would like different information built into his model than the one who collects health care debt.
  But Seoane sees this as a disadvantage to the industry. There is no equivalent to the FICA score for the collection industry, he says, and there should be.
  A good generic model would be affordable, force consistency into the industry and drive down risks of litigation, Seoane says. It is what he says is the next level in scoring.
  Before collectors begin working for CreditMax, a West Palm Beach, Fla.-based debt buyer that contracts with collection firms to work their debt, they answer two questionnaires with nearly 200 questions in total. The questions range from a list of the collector's favorites-music, food, vacation spots-to specific questions about their background, preferences and collection history.
  "I would much rather have an outstanding collector attempt to collect a low-scored debt than a higher-scored debt placed in the hands of a low-quality or badly trained collector," Kass says.
  After the collectors' information is processed, the individuals are sent the kind of accounts they like to work. That model, plus generous bonuses and salaries, translates into hard-working and highly successful collectors, Kass says.
  Talent Focus
  "We have people knocking at our door wanting to come to work for us, the best collectors in the country," says Bob Peterson, chief operating officer at Montana-based ARS Recovery Services LLC, one of CreditMax's third-party collectors.
  Collector turnover on CreditMax accounts is very low in an industry plagued by high turnover, says Peterson. The system also produces a secondary benefit: Collectors who do not work CreditMax accounts work their own clients harder, hoping to become CreditMax collectors.
  "I challenge anyone to score debt and give it to us, and I will guarantee that we will be materially different from what they score the debt at," Kass says. "We'll place the debt in the hands of the right collector, and that's where the pedal hits the metal."
  However, some worry about data-privacy laws, which may constrict personal information. More threatening is the possibility of smaller shops being squeezed out, as they cannot afford the more sophisticated scoring models.
  In this age of information, advantage goes to those with the most. But as the collection industry is finding out, it is not just gathering information that is important. It is knowing how to shape it as well.
  (c) 2006 Cards&Payments and SourceMedia, Inc. All Rights Reserved.
  http://www.cardforum.com http://www.sourcemedia.com