Capital One Financial Corporation: Response Modeling

Abstract

This case describes the use of response modeling at Capital One. Response modeling refers to the building of models to forecast the likelihood a prospect will respond to a direct solicitation. The use of response modeling is one component of the Capital One information-based marketing strategy. A data set is available from the author; it affords students the opportunity to try their hand at response modeling and to receive a score based on how will their chosen solicitation strategy fared. Excerpt UVA-M-0507 CAPITAL ONE FINANCIAL CORPORATION: RESPONSE MODELING Signet Banking Corporation, a regional bank based in Richmond, Virginia, offered its first bank card in the 1950s, making it the oldest continuously operating bank-card issuer in the United States. In 1988, the credit card division had over one million cardholders and $ 1.25 billion in outstanding loans. Shortly thereafter, the division hired Nigel W. Morris and Richard D. Fairbank, two principals of a consulting group that had devised a new strategy for marketing credit cards. Under their guidance, the division experienced explosive growth. In February of 1995, Signet spun off its card business as Capital One Financial Corporation with Fairbank as CEO and Morris as president. In June of 1995, Capital One was the 10th largest credit card company in the United States with 6 million cardholders and $ 9 billion in outstanding loan balances. By 1996, Capital One was widely recognized as one of three “category killers” in the credit card business. A key component of the Capital One strategy involved collecting and maintaining individual-level data to support a variety of modeling efforts. The results of these modeling efforts were used to improve the efficiency with which Capital One acquired and managed its portfolio of bank-card accounts. Response modeling was one key component of this information-based strategy. Response Modeling . . .