Consumer Mailing Lists: Using Known Vs. Modeled Credit Data

by Dave Hare

There are several instances where companies may want to offer individuals special offers to re-finance their homes, offer a loan consolidation offer or credit repair. Using known credit data is very accurate because, beyond the FICO score itself, you can select or omit based on debt amounts on open trade lines, interest rates on existing loans, late payment history. This data is also highly regulated; to get access to this kind of data, you need to make a bonafide offer of credit. For example a letter might say, “Dear X, according to our records and your good payment history, we are able to offer you a lower interest rate of 3.9% for refinancing up to $450,000 for a term of 15 years.”

Offers which do not make a firm offer of credit are required to use modeled data. Modeled data is created using various characteristics of individuals (age, income, marital status, credit history and proprietary demographic information) and applying this on a zip+4 level; which can be as small as a few households up to over 100 in densely populated areas. Some models start with known credit data on a few hundred individuals with extensive demographic details and then apply the model to find other individuals with the same demographics and then model a credit score.

Several companies make their models available to mailers and telemarketers and most have certain strengths and weaknesses in their data, depending on the sources used and how different companies consider what attributes are more important than others.