MARKETING
Professor Emeritus of Marketing
Robert C. Blattberg is the Polk Bros. Professor of Retailing, Professor of Marketing and Director of the Center for Retail Management. Professor Blattberg joined the faculty in 1991. Prior to his appointment at the Kellogg School, Blattberg served as the Charles H. Kellstadt Professor of Marketing and Director of the Center for Marketing Information Technology at the Graduate School of Business at the University of Chicago. He received his PhD in industrial administration from Carnegie Mellon University.
His primary research is in the areas of marketing information technology, database marketing, sales promotions, pricing and retailing. His articles have appeared in the Journal of Marketing Research, Management Science, Marketing Science, Econometrica, Journal of Marketing, Journal of Direct Marketing, and other leading academic journals. He has co-authored four books, including Sales Promotions (Prentice-Hall) and Customer Equity (Harvard Business Press). Professor Blattberg has consulted to a wide variety of firms including American Express, Kroger, Best Buy, Rite Aid, IRI, and A.T. Kearney. He has won both the John D.C. Little award for best paper in Marketing Science and the Robert B. Clarke Award from the Direct Marketing Educational Foundation as Educator of the Year.
Professor Blattberg is currently working on a fifth book about the lifetime value of customers. He serves as a Director of First Horizon National Corporation (Memphis), Price Chopper (Schenectady, NY), Mannatech, Incorporated (Dallas) and Gapbusters (Australia).
- Recent Media Coverage
Expansion (Mexico): ¿Cliente leal o potencial? - 6/2/2008
Expansión (Mexico): ¿A quiénes recompensar? - 5/26/2008
See all Kellogg in the Media
For both academics and practitioners, the dominant focus of customer relationship management has been customer retention. The authors assert that customer winback should also be an important part of a customer relationship management strategy. Customer winback focuses on the reinitiation and management of relationships with customers who have lapsed or defected from a firm. In some cases, firms engage in extensive efforts to reacquire lapsed customers or defectors, and a common tactic is lowering the price to reacquire a customer. This investigation goes beyond the reacquisition pricing strategy and also examines the optimal pricing strategy when the customer has decided to reinitiate the relationship. By simultaneously modeling reacquisition and duration of the second tenure with the firm, the authors determine that the optimal pricing strategy for their application involves a low reacquisition price and higher prices when customers have been reacquired. In addition to pricing strategy, they also discuss the implications of their findings for targeting lapsed customers for reacquisition.
Consider the problem of estimating a price-sensitivity parameter in a demand model. Depending on the context in which the estimate will be used, traditional squared-error loss may be inappropriate. The authors consider the situation in which the estimate will be used by a manufacturer to set the price. Effectively, the manufacturer's goal of profit maximization induces a loss function that turns out to be asymmetric. Estimates that perform well with respect to such loss functions are obtained. A real example is considered in which, compared to traditional estimation under squared-error loss, this approach leads to smaller price-sensitivity estimates, suggesting higher optimal prices.
Focuses on the areas of potential behavioral research and its relation to consumer behavior. Sophistication of the marketing research; Development of theories on consumer behavior; Indication of topics for the research.
It's a marketer's dream - the ability to develop interactive relationships with individual customers. Technology, in the form of the database, is making this dream a reality. Now companies can keep track of customer preferences and tailor advertising and promotions to those needs. For instance, a grocery store system could note that you recently purchased a sample size of dishwashing detergent and could offer you a coupon to buy the large size. Blattberg and Deighton explore the impact of this development on marketing practice and give practical advice on designing a marketing database and staffing an interactive marketing department. They also address consumer fears and the public debate over marketing and privacy. (Reprinted by permission of the publisher.)
Consider the problem where a retailer or manufacturer wants to estimate product price and promotional elasticities based on supermarket scanner data. Classical linear modeling suffers from the following aggregation dilemma. Price and promotional elasticities appear to vary considerably among chains and brands so that one overall model is too restrictive. Alternatively, the use of a different model for each chain and brand leads to noisy and often nonsensical estimates of separate elasticities because of excessive data variation. To resolve this dilemma, shrinkage estimation procedures are proposed. By borrowing strength across chains and brands, these procedures reduce variability while providing flexibility that allows for separate elasticity estimates. Application of these procedures to a large data set yields not only more reasonable model estimates but also improved predictive power.
We focus on ways of combining simple database models with managerial intuition. We present a model and method for isolating managerial intuition. For five different business forecasting situations, our results indicate that a combination of model and manager always outperforms either of these decision inputs in isolation, an average R[sup 2] increase of 0.09 (16%) above the best single decision input in cross-validated model analyses. We assess the validity of an equal weighting heuristic, 50% model + 50% manager, and then discuss why our results might differ from previous research on expert judgment.
This research focuses on how price changes influence the observed pattern of brand competition. The paper begins with a basic utility model formulation and examines the implications of three major classes of preference distributions on the expected patterns of competition. A price-tier model is proposed to operationalize the theory and to allow predictive testing. The price-tier model is estimated on 28 brands across four product categories. The results show a specific asymmetric pattern of price competition. Higher-price, higher quality brands steal share from other brands in the same price-quality tier, as well as from brands in the tier below. However, lower-price, lower-quality brands take sales from their own tier and the tier below brands, but do not steal significant share from the tiers above. The results are consistent with a bimodal preference distribution, with the regular price indifference point being located toward the lower-quality end of the preference distribution for the categories analyzed.
Direct marketing tests often generate expectations that are not realized during rollout. One reason for this is that the variability of the response rates tends to be greater than that predicted by the standard binomial model. An alternative model is proposed which accounts for this extra variability, and implications with respect to test design and evaluation are discussed. Results from an actual direct marketing test are used to illustrate different aspects of the proposed model. Recommendations for improving current testing procedures are also examined.
Trade promotions have become an increase important element of the marketing mix. Yet, there is very little research describing how to measure the profitability and effectiveness of trade promotions. This paper describes how retailers behave when trade promotions are offered Then, a model is developed to capture the two key components of the process, the consumer and the retailer. An example is given showing how to apply the model to actual manufacturer and retail sales data. Then estimates of the profitability for different items in a product category are calculated. Many research questions ute raised in this paper which can serve as future directions for research. Why are trade generally unprofitable? How can scanner data improve The estimates given? How do different types of trade promotions after the retailer and ultimately the consumer? Which brands and items should be trade promoted? (Modelling: Trade Promotion: Scanner Data)
The purpose of this paper is to derive a model of advertising effects on the firm's sales. A micromodel is postulated and aggregated across individuals and over time to produce a macromodel of the aggregate sales-advertising relationship for a single product. The micromodel postulated is very simple. It incorporates two factors: reach of the ads and rate of decay of their effectiveness over time. This approach to modeling advertising effects is shown to be fruitful in several respects: (1) the coefficients of the aggregate equation are easily interpretable -- in terms of the reach and decay parameters; (2) the model derived is nonlinear yet estimable; (3) a special case of the model is very similar to lag models that have been in use; (4) the model can be used whatever the unit of time is; (5) the carryover effect of advertising (as commonly defined) is not constant, but depends upon the previous spending levels; and (6) the model helps illustrate that the duration of advertising may be greatly overstated if aggregate lagged dependent variable models are simplistically interpreted.
Food retailers regularly offer products for less than normal market price in special sales or deals. This paper briefly examines several common explanations for this phenomenon and finds the analyses to be less than complete. It then presents an explanation for dealing of storable products based on the idea of transferring inventory carrying costs from the retailer to the consumer. An inventory control model is described in which both consumers and the retailer act so as to minimize their own costs. Results derived from this model are then presented. Data relevant to both the consumer and the retailer model are presented and analyzed. The conclusion is that the data are consistent with the predictions of the models. Finally, the strategic implications of the model for manufacturers and retailers are discussed.
Recent literature contains several expositions of the log-linear modeling (LLM) capability of analyzing multiway contingency tables. This method has been proposed as a way of overcoming the deficiencies of traditional models such as ordinary least squares and AID. In order to begin an assessment of the utility of LLM, we report the results of four applications, and then provide a rationale for these empirical findings by examining the different model structures.
The authors examine the issue of determining the market segments to which a new national brand should be targeted. The usual recommendation is that the new brand should be targeted toward those segments that exhibit considerable brand switching. However, a new national brand should also attempt to attract segments that are loyal to existing national brands as well as segments that primarily purchase private labels. These implications follow from an explicit consideration of the changes in pricing and distribution patterns which occur when a new national brand is introduced. The results are illustrated with a set of diary panel data for facial tissue.
The author explains how decision theory can be applied to the design and evaluation of advertising experiments and discusses its application to marketing problems.
A model of consumer buying behavior is used to identify household characteristics that should affect deal proneness. The model treats household purchasing and inventory decisions like those of a firm. In other words, the household's purchasing decisions are assumed to be based on such factors as transaction costs, holding costs, and stockout costs in addition to product price. Household characteristics then are related to these cost parameters to identify households that are likely to be deal prone. The predictions are tested empirically by use of panel data for five frequently purchased products. The empirical results indicate that deal prone households can be identified and that the key variables affecting deal proneness are household resource variables such as home ownership and automobile ownership.
A new product forecasting model is described which uses survey data (not panel data) to predict year-end test market sales from early test market results (usually three months). In addition to offering a sales forecast, the model is designed to provide diagnostic information about a new product's strengths and weaknesses. By relating advertising expenditures, price, and perceptions of performance/acceptability of the product to sales, the model indicates how an unsuccessful product can be redesigned or the marketing mix changed to make possible a successful introduction. The model also can be used for new product planning. Given a media plan, price, sampling level, couponing, and some estimate of repeat usage, a pre-test market forecast of year-end sales can be made which allows management to evaluate different marketing plans to see which best meets profit or sales goals.
The modeling of buyer behavior by stochastic brand choice models has typically involved the use of a single model to represent the behavior of all consumers though consumer heterogeneity is recognized by allowing the model’s parameters to vary across the population. However, analysis of panel data for several frequently purchased products indicates the existence of several distinct consumer segments which are difficult to represent by a single model. It is shown, instead, that in order to describe adequately the behavior of these segments, it is necessary to use several different models while allowing consumers within a segment to have different model parameters. It is further shown that simple heterogeneous multinomial and Markov models appear to be adequate to represent the behavior of most of the segments.
This article describes the degree to which consumers use identical or similar brand and store choice strategies across product categories. The analysis is based on data on two pairs of frequently-purchased products and the results indicate that consumers frequently use identical or similar purchasing strategies across product categories. The principal implication of these results is that buying behavior may be governed by general household characteristics such as demographics to a greater extent than past research in this area had indicated.
This paper describes a Bayesian model-discrimination procedure which determines for each consumer the stochastic model of brand choice which is best supported by his purchasing behavior. The Bayesian technique is illustrated by means of two Markov models and two Bernoulli models. We first discuss in some detail how we set priors for the four models. Then we use simulated consumer panel data to demonstrate that the Bayesian technique is a good discriminator among the four models. The technique is then applied to some actual aluminum foil purchase data. We provide estimates of the proportion of aluminum foil consumers in each of the four model segments and the degree to which the size of each segment changes over time. We also show that model discrimination at the individual consumer level has important implications for market segmentation, pricing, and promotion.
The authors use data on aluminum foil purchases to empirically evaluate several aspects of a new, potentially more useful, market segmentation strategy. The proposed strategy should help the marketing manager define homogeneous purchase segments, determine the size of each segment, estimate each segment's response to changes in the marketing mix, and direct the company's marketing efforts to the appropriate segments.
This paper investigates the small sample properties of minimum chi-square estimates of the parameters of stochastic brand choice models. It also describes and evaluates a statistical test which is appropriate for discriminating between two stochastic brand choice models when one is a constrained version of the other.
If advertising by public transportation companies is to be useful, an effective method of advertising evaluation must be available. This paper presents such a method, based on a mathematical modeling approach, and the results of its application to a transit advertising campaign.
The authors discuss U.S. and monetary policy regarding inflation in the 1960s using the example of government intervention regarding copper prices in 1965. The Consumer Price Index (CPI) rose an average of 1.2 per cent annually from 1958 to 1964: however, in 1965 the CPI increased 2 per cent, the largest annual increase since 1957-1958. In 1962, the U.S. government adopted a policy of using wage-price guideposts as a way to determine inflation. This policy was used to reverse price increases in copper. The authors support a market-driven economy.
The true value of a marketing contact is the sum of any resulting short-term profit, the change in long-term customer value due to change in customer status and the strategic value of having more information about the customer. The first two components of the value of a contact have been studied extensively in the literature; we explore the third component in this paper. We operationalize the value of information as the increase in discounted future customer value due to policy improvement from additional customer information. We demonstrate how a marketing strategist, using Bayesian dynamic programming techniques, can improve performance by taking the value of information into account while making marketing contact decisions.
This seminar confronts students with significant problems, issues and theories at the leading edge of the marketing field. Presentations and discussions are designed to stimulate thinking on important areas of research and the development of new theoretical viewpoints.
PHONE: 847-491-2721
FAX: 847-491-2498
Jacobs Center Room 455