Markov Chains Applied to Marketing

Styan, George P. H.; Smith, Harry A.

doi:10.1177/002224376400100109

Cited by 37 publications

(9 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This change over time is referred to as nonstationarity. In empirical applications, the presence or absence of nonstationarity in choice probabilities is determined by a statistical test (Anderson and Goodman 1957;Montgomery 1969;Styan and Smith 1964). This phenomenon can be captured simply by including a variable in the choice model that indicates time or purchase occasion number.…”

Section: The Multichannel Marcom Processmentioning

confidence: 99%

“…Second, we include only the most recent prior purchase as a variable in the model. We based this decision on the statistical tests that Styan and Smith (1964) outline, from which we determined that the channel choice at time (t - 1) affects the channel choice at time t. 2 Third, we include the purchase occasion variable. Using a test that Anderson and Goodman (1957) outline, we conclude that channel choice probabilities change over time (i.e., they are nonstationary).…”

Section: Application Of the Marcom Processmentioning

confidence: 99%

See 1 more Smart Citation

Managing Marketing Communications with Multichannel Customers

Thomas¹,

Sullivan

2005

Journal of Marketing

240

182

View full text Add to dashboard Cite

This article presents a marketing communications process that uses customer relationship management ideas for multichannel retailers. The authors describe and then demonstrate the process with enterprise-level data from a major U.S. retailer with multiple channels. On the basis of the results, the authors develop an initial marketing com-). The authors thank the special issue consulting editors, William Boulding and Richard Staelin, for their helpful comments on this article. They are also grateful for the assistance of Len Berry and the Center for Retailing Studies at Texas A&M University. O ver the past decade, customer relationship management (CRM) has proved critical in helping firms make more money by enabling them to identify the best customers and then satisfy their needs so that they remain loyal to the firm. More recently, CRM has grown increasingly complex with the proliferation of retailers expanding their channels of distribution (Cleary 2000). This has led to the need for enterprise-level data, which are the aggregation of data gathered from all firm interactions with their customers across all channels. Given this data-rich, dynamic environment, how can a firm identify the customers who will migrate among its multiple channels and predict their migration patterns? More important, how does the firm communicate with these customers to influence their channel choices and, ultimately, their value? This research focuses on answering these questions.Thus, this article has two general objectives. First, we illuminate a process by which multichannel retailers can leverage enterprise-level data to understand and predict their customers' channel choices over time. Second, we demonstrate how the information gained from this process can be used to develop strategies for targeting and communicating with customers in a multichannel environment. The benefits achieved from the application of this process include increased efficiency in marketing expenditures and enhanced customer value. In the next section, we outline the general process for managing marketing communications (MARCOM) with multichannel retail customers. Subsequently, we demonstrate the application of the process using an enterprise database of a multichannel retailer. We conclude by noting the limitations of the study and ideas for further research in this area. The Multichannel MARCOM ProcessThe process of managing MARCOM in a multichannel environment begins with the identification of relevant factors that differentiate among customers who use different channels. It continues with the development of a communication strategy for existing customers, and it ends with the prediction of the right communications strategy for prospects and new customers. Step 1: Estimate a Segment-Level Channel Choice ModelThe critical aspect of this step of the process is not choosing the model (e.g., multinomial logit or probit) as much as it is specifying the model. It is important that the variables in the model are factors that (1) drive channel choice, (2) help clas...

show abstract

Section: The Multichannel Marcom Processmentioning

confidence: 99%

Section: Application Of the Marcom Processmentioning

confidence: 99%

Managing Marketing Communications with Multichannel Customers

Thomas¹,

Sullivan

2005

Journal of Marketing

240

182

View full text Add to dashboard Cite

show abstract

“…The proportion of brand switching outcomes between the first quarter and subsequent quarters represents a Markov chain, classifying the probability of switching from one state (i.e. modified brand) to another in the subsequent time period (Styan and Smith, 1964; Ehrenberg, 1965; Lattin and McAlister, 1985; Poulsen, 1990; Koschmann and Sheth, 2016). Here, household purchases are aggregated across each quarter of 2011, using the first quarter as the base period.…”

Section: Methodsmentioning

confidence: 99%

Brand line extensions: creating new loyalties or internal variety-seeking?

Koschmann

Sheth

2018

JPBM

View full text Add to dashboard Cite

Purpose The purpose of this paper is to examine whether line extensions (modified brands) create their own loyalties or induce variety-seeking within the brand. Prior research has explored how the branded house strategy (i.e. multiple products bearing the same brand name) retains customers from competing brands. However, this research investigates loyalty within the brand by comparing loyalty and variety-seeking rates of modified brands. Design/methodology/approach Markov chains examine behavioral loyalty and switching rates of panel households in the USA over several quarters for two family brands of carbonated beverages. Emphasis is placed on the consumers who purchase the upper median of volume (heavy half) and constitute a disproportionate amount of brand’s sales (86 per cent of the volume). Findings Three propositions find that loyalty rates are high among modified brands with little switching to other lines within the brand. Further, loyalty and switch to rates are highest for the flagship branded product (the master modified brand). Practical implications Managers segment the market using the branded house strategy, yet loyalty rates vary for each product line. The switching rates can guide managers as to which products have established a loyal consumer base. Originality/value While brand switching is a considerable research stream, this research is believed to be the first to explore loyalty versus variety-seeking in the branded house strategy.

show abstract

“…The bulk of research has been in terms of Markov Chains. In particular, actual research has been carried out on the assumption that brand purchase behavior is a first-order stationary Markov process [2], [99], [100], [129], [154], [160], [223], [231], [232], [293], [324], [364]. For a good statement of the mathematical theory of Markov Chains see Kemeny and Snell [190].…”

Section: Operations Researchmentioning

confidence: 99%

A Review of Buyer Behavior

Sheth

1967

Management Science

View full text Add to dashboard Cite

Post-war advances in technology have changed the concept of marketing management considerably. The core of this change is the shift of emphasis from the firm to the buyer. In this survey article, the author reviews all the important approaches to understanding buying behavior and provides an up-to-date bibliography. Two important issues emerge from the review. First, the existing variety of formulations resembles the variety of responses of seven blind men touching different parts of an elephant and making inferences about the animal which necessarily differ from, and occasionally contradict, one another. Second, the theory which attempts to explain the observed phenomenon of buying behavior and the quantitative techniques which provide adequate definitions and measurements have been developed independently of each other to the detriment of the maturity of the discipline.

show abstract

Markov Chains Applied to Marketing

Cited by 37 publications

References 1 publication

Managing Marketing Communications with Multichannel Customers

Managing Marketing Communications with Multichannel Customers

Brand line extensions: creating new loyalties or internal variety-seeking?

A Review of Buyer Behavior

Contact Info

Product

Resources

About