Academic journal article Academy of Marketing Studies Journal

A Multi-Objective Optimization Approach Using the Rfm Model in Direct Marketing

Academic journal article Academy of Marketing Studies Journal

A Multi-Objective Optimization Approach Using the Rfm Model in Direct Marketing

Article excerpt

(ProQuest: ... denotes formulae omitted.)

INTRODUCTION

Direct marketing is all about customer data: their characteristics, their buying habits, and their buying potential. Data is obtained from many sources, including internally generated data, public databases, and third party list vendors. The widespread use of data analytics by many direct marketing firms allows them to use this customer data to fine-tune their marketing strategies with precision and accuracy. Data analytics involves the strategic and extensive use of data and quantitative analysis to improve business decision making (Davenport and Harris, 2007, 2010). Customer data and data analytics are especially important in direct marketing because they are used to help firms improve response rates, conversion rates, and campaign profitability (Davenport and Harris, 2007; Dyer, 2003; Hambleton, 2013).

One particular analytical tool used frequently in direct marketing is the RFM model. The recency-frequency-monetary value (RFM) framework leads to highly effective direct marketing campaigns by enabling companies to categorize customers into homogenous segments based on their previous purchasing behavior and then design highly customized promotional campaigns to reach those customers. According to this approach, customer data on the recency of purchase (R), frequency of purchase (F), and monetary value of purchase (M) are captured and stored for each customer. Then, customers with similar values are grouped together, and targeted promotional offers are created to reach them. For example, if a given customer segment shows a low value for recency and relatively high values for frequency and monetary value, these customers are typically approached with a "we want you back" marketing strategy. If a given customer segment shows a low monetary value and high values for frequency and recency, a more relevant "up-selling" marketing strategy could be designed to generate additional sales revenue.

The RFM model typically assumes unlimited marketing resources, however, and suggests that a company can reach all its customers, even customers with less than optimal RFM scores. Clearly, most organizations operate under yearly budget constraints, and therefore such assumptions are impractical. Adding optimization to the well-known RFM approach to help allocate resources most effectively was recommended by Fader et al. (2005b) as an important next step for future research.

In addition, the importance of the R, F, and M components in the RFM approach for a given marketing campaign might not be the same. For example, a company trying to improve its customer retention rate might be interested primarily in recency, i.e., prioritizing the return of lost customers who may have defected to the competition. For the same campaign, frequency and monetary values might be second and third priorities, respectively. When confronted with both spending limits and differing goals, marketing managers should allocate marketing resources toward those customers with the greatest long-term profit potential.

This research proposes a multi-objective optimization methodology based on a goal programming (GP) approach to profit maximization for direct marketers using RFM data. One unique characteristic of this (GP) model is the inclusion of varying direct marketing objectives as well as corresponding budget constraints.

In addition to balancing marketing priorities with marketing budgets, companies must also strive to achieve a balance between two types of errors for any given campaign: Type I and Type IL A Type I error would occur when organizations ignore customers (mistakenly) who could have returned and repurchased, thereby providing the firm with additional revenue and profit. Type II errors occur when companies (unknowingly) target customers with their marketing campaigns who are not ready to purchase (Venkatesan & Kumar, 2004). The model proposed in this research creates a balance between a Type I and a Type II error by identifying the proper RFM segments to target. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.