No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
On the Prediction Power of Market Share Attraction Models
Abstract
Logical consistency of market share models has received considerable attention in the recent marketing literature. The reports suggest that, at least theoretically, attraction-type specifications are more easily justified than the ones commonly used in empirical work. The authors examine whether the theoretical superiority also implies a better predictive performance. The empirical results give a tentatively affirmative answer.
... Only recently has empirical evidence been introduced to resolve the issue. Specifically, Naert and Weverbergh (1981), using data from two different markets (seven brands of gasoline and three brands of electric shavers), compared several specifications of the linear, multiplicative, and attraction models and concluded, using predictive power as the sole criterion, that the attraction model was superior. This conclusion must be regarded as tentative, however, in view of the limited sample size employed, a fact acknowledged by the authors when they state that "further empirical study is needed to confirm our conclusions, to refine them, or to refute them" (Naert and Weverbergh 1981, p. 152). ...
... We describe such an attempt, using data representing 15 brands in three different markets to compare the performance of nine different market share specifications. Following Naert and Weverbergh (1981), we tested both unconstrained and restricted forms of the linear, multi-plicative, and attraction models with OLS as well as GLS estimates. We used a methodology similar to that of Naert and Weverbergh to enhance comparability of results. ...
... This issue has received considerable attention recently. Beckwith (1973), Koehler and Wildt (1977), McGuire et al. (1968), McGuire and Weiss (1976), Naert and Bultez (1973), Schmalensee (1973), Weverbergh (1976), and Weverbergh, Naert, and Bultez (1981) all investigated whether logical consistency could be achieved by imposing parameter restrictions. Naert and Bultez (1973) demonstrated that linear market share predictions can be made to sum to one if one is willing to assume that the response parameters for the different marketing instruments do not vary across brands, a restriction which does not have much intuitive appeal. ...
Linear and multiplicative market share models are compared with the attraction specification on the basis of both predictive and descriptive measures. The data used represent 15 brands in three different markets. The results show that linear and multiplicative models with brand-specific parameters estimated by OLS or GLS may perform as well as or better than the attraction specification. The failure of parameter restrictions and ex-post normalization to improve prediction power is explained in terms of the heterogeneity in the estimated parameters. This finding is consistent across brands and models and suggests that the hypothesis of equal response parameters across brands should be tested formally before such restrictions are imposed. The combined results suggest that the attraction model may have considerable intuitive appeal, but that its superiority for descriptive and predictive purposes is not yet established.
... One important question is the type of functional relationship to be specified in the model. Though linear and multiplicative specifications have been used commonly in the past, several authors have suggested attraction-type models (Bultez and Naert 1975;Naert and Weverbergh 1981;Nakanishi and Cooper 1974). Another issue is the assumption about the properties of the error terms and the corresponding statistical procedure used to estimate the model. ...
... Should the models contain different response parameters for every brand or a common parameter for all brands, or should some intermediate level of parameterization be used? Though models with brand-specific response parameters are most meaningful for decision makers, Naert and Weverbergh (1981) have found that constraining the parameter values improves the predictive performance of market share models. ...
... On the basis of his study, Rao concluded that model performance was sensitive to the way in which variables were defined, but the method of estimation had no significant impact. Naert and Weverbergh (1981) conducted an extensive comparison of 13 market share models using data for gasoline and electric razor brands. The three basic model specifications tested were linear, multiplicative, and attraction-type models (Bell, Keeney, and Little 1975). ...
The authors compare market share models differing in terms of functional form, estimation procedure, and level of parameterization. They evaluate the models by testing the accuracy of predicted shares as well as the validity of the estimated parameters. The tests indicate that valid parameter estimates and similar forecasts are obtained with linear, multiplicative, and attraction models. Better share forecasts generally are obtained from models with brand-specific parameterization and OLS estimation. Most of these conclusions are contrary to previous findings. Several implications for researchers are discussed.
... 1. La première raison est que certains auteurs ont montré empiriquement que les modèles d'attraction ne donnent pas des résultats significativement meilleurs que les autres modèles en termes de qualité d'ajustement et de prédiction (voir par exemple Ghosh et al. [20] et Leeflang et Reuyl [36]). Cependant, Naert et al. [51] ont montré l'inverse quelques années avant, suggérant que cette affirmation dépend de l'application considérée. 2. La deuxième raison est que l'estimation d'un modèle d'attraction n'est pas évidente : c'est un modèle non linéaire qui peut être linéarisé par une transformation, généralement la transformation log centrée, également appelée transformation log ratio centrée (CLR) dans la littérature de l'analyse des données de composition (voir Aitchison [1]). ...
... The first one is that some authors have shown empirically that attraction models do not give significantly better results than the others in terms of fitting and prediction accuracy (see for example Ghosh et al. [20] and Leeflang et al. [36]). Nevertheless, Naert et al. [51] have made the opposite claim a few years ago, suggesting that the conclusion can depend on the considered application. 2. The second reason is that the estimation of an attraction model is not straightforward: it is a non-linear model which can be linearized by a transformation, generally the log-centering transformation, also called centered log-ratio transformation (CLR) in the compositional data analysis literature (see Aitchison [1]). ...
... The first one is that some authors have shown empirically that attraction models, as the multiplicative competitive interaction model (MCI) for example, do not give significantly better results than the others in terms of fitting and prediction accuracy (see for example Ghosh et al. [20] and Leeflang et al. [36]). Nevertheless, Naert et al. [51] have made the opposite claim a few years ago, suggesting that the conclusion can depend on the considered application. ...
The aim of this CIFRE thesis, realized with the market research institute BVA in collaboration with the automobile manufacturer Renault, is to build a model in order to measure the impact of media investments of several channels (television, outdoor, etc.) on the brands' market shares, taking into account the competition et the potential cross effects and synergies between brands, as well as controlling for average price and regulatory context (scrapping incentive).Market share models have been developed in the marketing literature, especially the GMCI model (generalized multiplicative competitive interaction model), inspired from the aggregated conditional MNL (multinomial logit) model. In the statistical literature, the compositional data analysis (CODA) allows to analyze share data respecting their nature (a vector of D shares subject to the unit sum constraint is a composition and belongs to the dimension D simplex space). Regression models for dependent and explanatory compositional variables exist but are rarely used in practice. Finally, the Dirichlet covariate model allows to model a simplex valued dependent variable.In the first chapter, these different models avec compared from a theoretical and empirical point of view. It is shown that all of them can be expressed with a similar formulation using the notions of attraction and of simplicial expected value. The GMCI model appears to be a particular case of the CODA model, such that these two specifications can be combined into a unique model. The complexity of Dirichlet and CODA models turns out to be necessary in order to capture the diversity of competitive relationships. In the second chapter, emphasis is given to the interpretation of models which is not very well developed in the CODA literature. Different types of interpretations are presented, but it is demonstrated that the calculation of the elasticities of market shares relative to media investments in particularly relevant from a mathematical point of view and from a practical perspective. Indeed, we prove that elasticities are consistent with C-derivatives of simplex valued functions of another simplex. Moreover, these elasticities can be easily interpreted by car manufacturers and can be used for advertising budgeting optimization (Dorfman-Steiner theorem). In the third chapter, a practical application to the B segment of the French automobile market is presented for the purpose of measuring the impact of the different advertising channels on the market shares of the three leaders of this segment and of the group of other brands, taking into account the lagged effects of advertising (adstock function) and the competitive cross effects. The media investments elasticity of the brand market share varies from one brand to another and from one channel to another. Synergies between some brands can be highlighted. The last chapter opens the discussion on different directions to be explored in order to improve the proposed model and to provide further answers to the considered issue.
... The paranKter a, (0 s a s 1), would determine the rate of convergence of the process. A system of such equations, one for each competing firm, would represent the competitive dynamics of the mar-ke^lace in a logically consistent manner (McGuire and Weiss 1976, Naert and Bultez 1973, Naert and Weverbergh 1981. Under this model, if each firm maintained a constant share of strategic investments, i.e., Xj,, = Xj, it may be verified that mlj = Xj at equilibrium. ...
... The paranKter a, (0 s a s 1), would determine the rate of convergence of the process. A system of such equations, one for each competing firm, would represent the competitive dynamics of the mar-ke^lace in a logically consistent manner (McGuire and Weiss 1976, Naert and Bultez 1973, Naert and Weverbergh 1981. Under this model, if each firm maintained a constant share of strategic investments, i.e., Xj,, = Xj, it may be verified that mlj = Xj at equilibrium. ...
... The preceding discussicKi has shown that nonlinear equilibrium response fiinctions such as (3) are theoretically plausible rejsesentations of the relationship between share of nutrket and share of nrmrketing investments. There is also considerable empincal support for such formulations (Kotier 1980, Little 1981, Naert and Weverbergh 1981. ...
This article critiques the "new paradigm of marketing strategies" proposed by Cook {JM, Spring 1983). It is shown that his formulation of the strategic marketing ambition and differential advantage concepts does not provide general or actionable links between a firm's market performance goals and its marketing mix decisions. The consequent limitations of Cook's framework are discussed, and alternative formula- tions that may overcome these limitations are suggested.
... In this case one might wish to use the type of logit model applied by McFadden (1974). Naert and Weverbergh (1981) demonstrated that the predictive power of logically consistent market share models is superior to that of nonlogically consistent We readily note that the main contribution to the conditional logit of brand D (relative to brand C) entails reductions in its own price; in particular, brand D's third lowest price shows a very large contribution (0.72 compared with the base of zero), after which the effect attenuates with further reductions in price." Price changes in brands A and B show no significant effect on brand D's conditional logit and the lowest price reduction in brand C shows only a modest effect. ...
... 9 Though only 16 pricing conditions were employed in the design, the model could be used to make 4 4 = 256 models. The logically consistent market share models satisfy the basic properties of market shares, namely having a value between zero and one (the range constraint) and summing to one across brands (the sum constraint) (Naert and Bultez 1973). The linkage between the model we describe and the logically consistent market share attraction models can be easily established. ...
The authors describe an approach to measuring self- and cross-price/demand relationships via conjoint analysis. The modeling “trick” is to characterize each brand as a separate price-related factor in which respondents are asked to indicate their subjective likelihood of buying a set of J brands, each priced at some experimental level. A first-order Markovian model is postulated to represent price-induced brand switching. The price-demand model is then fitted (by generalized least squares) to share of choices, estimated after T time periods. The model is illustrated empirically in the context of a consumer nondurable good.
... Market response may be formulated as a linear additive, multiplicative, or attraction function. Each has its theoretical as well as its practical strengths and weaknesses (Barnett 1976;Beckwith 1972;Brodie and de Kluyver 1984;Ghosh, Neslin, and Shoemaker 1984;Karnani 1983;Leeflang and Reuyl 1984;McGuire and Weiss 1976;McGuire et al. 1968;Naert and Bultez 1973;Naert and Weverbergh 1981). Additive functions have the benefit of computational simplicity but are known to be biased in several ways, particularly in their failure to account for interactions among marketing variables. ...
... Market response may be formulated as a linear additive, multiplicative, or attraction function. Each has its theoretical as well as its practical strengths and weaknesses (Barnett 1976;Beckwith 1972;Brodie and de Kluyver 1984;Ghosh, Neslin, and Shoemaker 1984;Karnani 1983;Leeflang and Reuyl 1984;McGuire and Weiss 1976;McGuire et al. 1968;Naert and Bultez 1973;Naert and Weverbergh 1981). Additive functions have the benefit of computational simplicity but are known to be biased in several ways, particularly in their failure to account for interactions among marketing variables. ...
This article evaluates the dynamic interaction of market share and profit. The determinants of market share are compared with published empirical research and illustrated in a consumer product industry. The effects of market growth, maturity and decline, competitive attack, price leadership, and heavy user strategies on market share and profit are analyzed theoretically and simulated over a 10-year period.
... The coefficients of the attraction model (a,ß,s,d) reflect the differential effectiveness of elements of the marketing mix (Naert and Weverbergh, 1981;Leeflang and Reuyl, 1984;Ghosh, Neslin, and Shoemaker, 1984). We assume that the relative effects of the marketing mix ...
... Finally, market condition M3 suggests that the contribution of each marketing mix variable to attraction is concave. In other words, there is a decreasing marginal effect of any changes in marketing variables (Naert andWeverbergh 1981, Ghosh, Neslin andShoemaker 1984). This assumption is supported by academic and industry research (Lambin, 1976;Simon and Arndt, 1983;Zielske, 1986). ...
We operationalize Aspiration Adaptation Theory in the context of incumbent response to entry. We show that bounds on aspiration levels affect the direction and magnitude of marketing mix responses to entry. Using scanner panel data, we estimate a spatial brand share attraction model in the Marion, IN coffee market to examine how well the alternative bounded and unbounded aspiration schemes reflect the observed marketing mix changes. We find that these alternative aspiration schemes (bounded and unbounded) capture the observed firm behavior, which suggests that Aspiration Adaptation Theory is a useful approach to modeling firm decision making.
... It is usually assumed that, by imposing in-sample specification restrictions, the out-of-sample forecasting accuracy will improve. Exemplary studies are Brodie and Bonfrer (1994), Danaher (1994), Naert and Weverbergh (1981), Leeflang and Reuyl (1984), Kumar (1994) and Chen et al. (1994), among others. A summary of the relevant studies is given in Brodie et al. (2001). ...
... The competitive structure can be studied using either sales or shares. We choose to analyze market shares and to measure the effectiveness of marketing instruments of different brands using the familiar market share attraction model (Naert and Weverbergh, 1981;Leeflang and Reuyl, 1984;Cooper and Nakanishi, 1988;Bronnenberg et al., 2000). An advantage of using market shares is that these are in general not influenced by category expansion or contraction. ...
The present availability of large databases in marketing, concerning, for example, store-level sales or individual purchases, has led to an increased demand for appropriate econometric models to deal with these data. The typical database contains information on revealed preferences, measured by for example sales, market shares or brand choices. In this thesis we study econometric models for some of these series, to be precise, we consider market shares, purchase timing and brand choices. These models allow us to, for example, gain insight into the effect of marketing instruments on consumer behavior. Examples of topics we discuss are heterogeneity in decision processes and the development of easily interpretable models to capture dynamical features in market shares and interpurchase times. Additionally, we contribute to the econometric literature by extending and developing models and estimating techniques.
... We assume the brand availability is a concave function of actual spending on distribution (Dist j ). The coefficients of the attraction model (δ, α, β, σ) reflect the differential effectiveness of elements of the marketing mix (Naert and Weverbergh 1981). We assume that the relative effects of the marketing mix variables are the same for each brand as in the Gupta (1988) model of the coffee market. ...
... Finally, market condition M3 suggests that the contribution of each marketing mix variable to attraction is concave. In other words, there is a decreasing marginal effect of any changes in marketing variables (Naert and Weverbergh 1981). Each of these sufficiency conditions is empirically validated in the Marion retail coffee market. ...
A new brand entering a market often finds itself in competition with sibling brands (those owned by the same parent company).
In a case study of a retail coffee market, the authors examine how these brand relationships might influence the sibling and
competitor brands' responses to entry. Using an empirically validated brand-share attraction model, the authors compare the
actual responses to entry with the optimal responses under different incumbent objectives. The authors find that the responses
by sibling brand are more accommodating than those of unrelated brands whose responses are consistent with the preservation
of preentry levels of sales.
... As mentioned in the introduction, we will use the general framework of gravity-type spatial interaction models, which were first introduced by Huff [18] to represent customer choice decisions. These models belong to the general class of stochastic utility models-also known as the ''brand share models'' in the marketing literature (see Naret and Weverbergh [25], and Nakanishi and Cooper [24]). ...
... Moreover, if a solution to (25) exists, then it is unique. ...
We consider a spatial interaction model for locating a set of new facilities that compete for customer demand with each other, as well as with some pre-existing facilities to capture the "market expansion" and the "market cannibalization" effects. Customer demand is assumed to be a concave non-decreasing function of the total utility derived by each customer from the service offered by the facilities. The problem is formulated as a non-linear Knapsack problem, for which we develop a novel solution approach based on constructing an efficient piecewise linear approximation scheme for the objective function. This allows us to develop exact and α-optimal solution approaches capable of dealing with relatively large-scale instances of the model. We also develop a fast Heuristic Algorithm for which a tight worst-case error bound is established.
... Naert, Philippe A. & Marcel Weverbergh [1981], "On the Prediction Power of Market Share Attraction Models," Journal of Marketing Research, 18 (May), 146-153.Naert, Philippe A. & Marcel Weverbergh [1983], "Market Share Specification, Estimation and Validation: Toward Reconciling Seemingly Divergent Views," Working Paper 83-90, Centrum Voor Bedrijfseconomie en Bedrijfseconometrie, Universiteit Antwerpen,Prinsstraat 13, 2000 Antwerpen Belgium, December. Brodie, Roderick & ...
... Market share attraction models are used for analyzing Interbrand competitive structures. These models have received more and more attention [1][2][3][4][5]. ...
Abstract In this paper, the dynamics of a discrete market share attraction model are investigated. It shows that the system can undergo flip bifurcation and chaos. The stability and bifurcation of a market share attraction model are analyzed by using the bifurcation theory and the center manifold theorem. The system displays complex dynamical behaviors, including period-1, 2, 4, 6, 8, 16 orbits, invariant cycle, a cascade of period-doubling, quasi-periodic orbits, and the chaotic sets. Numerical simulations illustrate the analysis and results.
... Market share models typically are specified as either linear, multiplicative, or attraction models (Naert and Weverbergh 1981). The PIMS measures do not permit an attraction type of specification. ...
In a broad cross section of consumer goods businesses, market pioneers generally have substantially higher market shares than late entrants. In fact, the empirical association between order of entry and market share is almost as strong as the association between market share and return on investment. The authors examine theoretical sources behind this relationship. The empirical results suggest that the higher pioneer shares are derived from firm-based superiority as well as from consumer information advantages. Nine hypotheses are developed and tested empirically. The results also indicate that order of market entry is a major determinant of market share.
... Other quality measures can be used for share data. See for example Kumar [21], Quagrainie [35], Leeflang and Reuyl [23], Naert and Weverbergh [29], Ghosh et al. [12]. Table 5 presents the out-of-sample average quality measures for our four models. ...
... Such flexible choice of customers is represented in the gravity-type spatial interaction models. These models are known as the brand share models in the marketing literature [4]. According to these models the utility u ij for a customer at point i ∈ N of a facility at location j ∈ S can be written as an exponential function. ...
In this paper the location and design problem is considered. The point of this is that a Company is going to open markets to attract the largest share of total customers demand. This share varies flexibly depending on the markets location and its design variant. The Company vies for consumers demand with some pre-existing competitors markets. The mathematical model is nonlinear, therefore, there are difficulties in the application of exact methods and commercial solvers for it. The ways of constructing upper bounds of the objective function are described. Two algorithms based on the Variable Neighborhood Search approach are proposed. To study the algorithms a series of test instances similar to the real data of the applied problem has been constructed, experimental analysis is carried out. The results of these studies are discussed.
... To define A i,c,t , we use a fully competitive interaction model (MCI) [18,19]. Although there is no consensus regarding the superiority of market share attraction models compared to simple linear or multiplicative models [20][21][22][23][24][25][26][27][28], we value the logical consistency of market share attraction models; that is, market shares are by definition greater than zero and sum to unity [29]. ...
In recent years, the expiration of patents for large drug classes has increased the importance of post-patent drug markets. However, previous research has focused solely on patent drug markets. In this study, the authors evaluate the influence of preferred supplier contracts, the German approach to tendering, in post-patent drug markets using a hierarchical market share attraction model. The authors find that preferred supplier contracts are a powerful strategic instrument for generic manufacturers in a highly competitive environment. They quantify the effects of signing a preferred supplier contract and show that brand-name manufacturers are vulnerable to tendering. Therefore, brand-name manufacturers should readjust their strategies and consider including preferred supplier contracts in their marketing mix. In addition, the authors employ a simulation to demonstrate that a first-mover advantage might be gained from signing a preferred supplier contract. Furthermore, their results can be used as a blueprint for decision makers in the pharmaceutical industry to assess the market share effects of different contracting strategies regarding preferred supplier contracts.
... Tellis (1988) conducted a meta-analysis study of models estimating the price elasticity of demand for various products and found that the mean price elasticity of demand is significantly negative. Also, Brodie and Kluyver (1984), Ghosh et al. (1984), Naert and Weverbergh (1981) and Gaur and Fisher (2005) investigated price elasticity issues using different datasets and pricing models. Besides cost-based and market-based pricing methods, market experiments can serve as useful scientific tools for firms to get a better understanding of consumer response to the use of marketing activities such as pricing and promotions, as well as new demand functions after the establishment of experience stores. ...
Methods of contacting customers are constantly evolving. An increasing number of manufacturers have established experience stores, plazas or centers that provide marketing, education, and trade functions to enhance business performance. This paper develops integer programming models to analyze the experience store establishment decision. Demand functions, cost parameters, and firm learning curves that impact on the experience store decision are also examined. In addition, this paper provides two case studies to validate the proposed experience store establishment models. Three proposed models can assist managers in developing marketing strategies related to the experience store establishment.
... Naert and Weverbergh (1981)은 13종류의 모형을 석유와 전자레인지 상품에 적용하여 모수 측정 기법간의 차이를 알아보기 위하여 OLS, GLS등 을 이용하여 분석하였다. 연구결과에서 GLS에 의해 측정된 유인모형이 보다 높은 예측력을 나타내었으 나, 시차를 고려하지 않은 경우에는 OLS에 의한 측정이 더 좋은 결과로 나타났다. ...
The purpose of this article is to present the application of Least Squares Support Vector Machine in analyzing the existing structure of brand. We estimate the parameters of the Market Share Attraction Model using a non-parametric technique for function estimation called Least Squares Support Vector Machine, which allows us to perform even nonlinear regression by constructing a linear regression function in a high dimensional feature space. Estimation by Least Squares Support Vector Machine technique makes it a good candidate for solving the Market Share Attraction Model. To illustrate the performance of the proposed method, we use the car sales data in South Korea's car market.
... Naert, Philippe A. & Marcel Weverbergh [1981], "On the Prediction Power of Market Share Attraction Models," Journal of Marketing Research, 18 (May), 146-153.Naert, Philippe A. & Marcel Weverbergh [1985], "Market Share Specification, Estimation and Validation: Toward Reconciling Seemingly Divergent Views," Journal of Marketing Research, 22 (November), 453-61.Brodie, Roderick & Cornelius A. de Kluyver [1984], "Attraction Versus Linear and Multiplicative Market Share Models: ...
... This type of model is theoretically appealing because it is logically consistent: it yields market shares that are between zero and one, and sum to one across all the competitors in the market. Market share attraction models have been used frequently in empirical work; see, e.g., Bultez and Naert (1975), Naert and Weverbergh (1981). Moreover, they are prevalent in the economics, game theory, operations research and marketing literature; see, for example, Monahan and Sobel (1994), Monahan (1987), Friedman (1958), Schmalensee (1976), Case (1979, Ch. 4), Cooper and Nakanishi (1988). ...
In this paper we study a class of dynamic promotional competition models, in which firms compete for market share by expending
marketing effort. We investigate two main issues. First, we answer the question if it is possible to give a global characterization
of the stability of the steady state effort allocation. We show that by using the concept of critical curves and an invariance
property of the coordinate axes a characterization of the set of feasible points (points that generate positive trajectories
converging to the steady state allocation) and its changes can be given. Second, we deal with the assumption of homogeneous
firms, which is often made in the literature. We demonstrate that the symmetric model which derives from this assumption exhibits,
in many situations, non-generic dynamical behavior. New phenomena, like Milnor attractors and synchronization of trajectories,
arising in the homogeneous case are illustrated. The introduction of small heterogeneities into the model invalidates many
of the conclusions derived under the hypothesis of homogeneous firms.
KeywordPromotional competition, Homogeneous and quasi-homogeneous firms, Global dynamics, Milnor attractors, Synchronization, Symmetry
breaking.
... As our share-switching model is dynamic by definition, we include lagged market shares in (4.26) and (4.28) in the same way as price and display are included. The literature often considers one common lagged market share parameter for all brands, see, for example, Brodie and De Kluyver (1984), Chen et al. (1994), Kumar and Heath (1990) and Naert and Weverbergh (1981). However, we find that this does not improve the predictions. ...
Thesis (doctoral)--Erasmus Universiteit Rotterdam, 2005.
Choice models are widely applied in psychology, economics, transportation, marketing and operations studies. We review the existing developments on the modeling of consumers' choices, including the attraction model, the utility‐based model, the temporal model, and the rank‐based model. The relationships among different classes of structural models are established and analyzed. Moreover, an operational data analytics (ODA) framework is presented to estimate the general consumer choice model using data. This framework, generalizing the existing estimation methods for specific structural models, strikes a delicate balance between the (likely imprecise) structural knowledge and the data. This is achieved by articulating the domain of validation through extending the structural knowledge and by formulating the data‐integration model based on the associated structural properties. We demonstrate the implementation of the ODA framework to identify the appropriate consumer choice models. The ODA estimate outperforms the existing parametric and nonparametric methods, particularly over the choice sets that are not covered in the data. We also discuss potential future research of developing ODA approaches to study the related aspects of consumer choice models.
The Brazilian Alcohol Program promotes ethanol as an alternative fuel to gasoline. Policymakers want to know the effect of relative fuel prices on demand for gasoline-fuelled and alternative fuel vehicles (AFVs). Considering vehicle engines dedicated to gasoline, ethanol or flex in the Brazilian market, we use market share models to estimate fuel price market share elasticities, both own and cross effects for each technology. In the first phase, when gasoline and ethanol are the only competing technologies, we find that variations in price profoundly impact the market share of new vehicles. The second phase shows how the efficiency of flex engines, reflected by the car cost per kilometre, significantly contributes to the widespread acceptance of this technology. The near-zero elasticities indicate that the market share of flex vehicles is hedged against price fluctuation. The study provides useful suggestions to help policymakers accelerate the socio-technological transition towards renewable and cleaner energy.
In the May 1984 issue of JMR three articles describe replications and extensions of work reported by Naert and Weverbergh in the May 1981 issue of JMR. They address various problems of market share specification, estimation, and especially validation. A rough comparison of the research findings seems to indicate that no general conclusions can be drawn. However, several of the differences in findings can be explained. Providing these explanations is the main purpose of this article.
Expiration dates are used by couponers to limit their financial liability temporally. Traditional wisdom assumes that coupon redemptions are greatest in the period immediately following the coupon drop and decline monotonically. Using regret theory, the authors predict that expiration dates induce a second mode in the redemption pattern just prior to the expiration date. They test this prediction by extending an existing coupon redemption model to incorporate an expiration effect and then estimating both the existing and the expiration models using weekly coupon redemption data for spaghetti sauce from A.C. Nielsen panels in two cities. Results are consistent with their prediction. Implications for practitioners and researchers are discussed.
The authors develop a two-stage forecasting methodology for estimating the sales responses to marketing and environmental variables when it is likely that their impacts will change unpredictably over time. The methodology is based on an integrated least squares procedure that uses regression analysis in Stage 1 to estimate the coefficients of the controllable and environmental variables in combination with a Stage 2 discounted least squares smoothing procedure that updates key parameters in response to changing market conditions. The effectiveness of the methodology is demonstrated by applying it to weekly sales data from a major retail chain.
The authors test Naert and Weverbergh's empirical finding that attraction-type specifications are characterized by a predictive power greater than that of their nonrobust counterparts. The empirical results, obtained by use of data on the German cigarette market, do not support the conclusions of Naert and Weverbergh.
The effectiveness of a sales promotion can be examined by decomposing the sales “bump” during the promotion period into sales increase due to brand switching, purchase time acceleration, and stockpiling. The author proposes a method for such a decomposition whereby brand sales are considered the result of consumer decisions about when, what, and how much to buy. The impact of marketing variables on these three consumer decisions is captured by an Erlang-2 interpurchase time model, a multinomial logit model of brand choice, and a cumulative logit model of purchase quantity. The models are estimated with IRI scanner panel data for regular ground coffee. The results indicate that more than 84% of the sales increase due to promotion comes from brand switching (a very small part of which may be switching between different sizes of the same brand). Purchase acceleration in time accounts for less than 14% of the sales increase, whereas stockpiling due to promotion is a negligible phenomenon accounting for less than 2% of the sales increase.
This article critiques the “new paradigm of marketing strategies” proposed by Cook (JM, Spring 1983). It is shown that his formulation of the strategic marketing ambition and differential advantage concepts does not provide general or actionable links between a firm's market performance goals and its marketing mix decisions. The consequent limitations of Cook's framework are discussed, and alternative formulations that may overcome these limitations are suggested.
The rise in vertical marketing systems (VMS) has had a profound impact on the nature of retail competition. The increase in intrasystem competition between outlets of the same VMS points to the need for increased sophistication in market area analysis models to aid in strategy development. This article develops a set of models to address critical managerial issues concerning retail locational strategy in an intrabrand competitive setting. Findings demonstrate the need to evaluate outlet-specific determinants of patronage behavior, in addition to those of a generic nature, in analyzing intrabrand competition.
We consider a decentralized supply chain (DSC) under resale price maintenance (RPM) selling a limited‐lifetime product to forward‐looking customers with heterogeneous valuations. When customers do not know the inventory level, double marginalization under RPM leads to a higher profit and aggregate welfare than without RPM under a two‐part tariff contract (TT). Both RPM and TT profits are higher and aggregate welfare is lower than in a centralized supply chain (CSC). When customers know inventory, RPM coincides with CSC. Thus, overestimation of customer awareness may lead to overcentralization of supply chains with profit loss comparable with the loss from strategic customers. The case of RPM with unknown inventory is extended to an arbitrary number of retailers with inventory‐independent and inventory‐dependent demand. In both cases, the manufacturer, by setting a higher wholesale price, mitigates the inventory‐increasing effect of competition and reaches the same profit as with a single retailer. The high viability and efficiency of RPM in using double marginalization as a strategic‐behavior‐mitigating tool may serve as another explanation of why manufacturers may prefer DSC with RPM to a vertically integrated firm.
This article is protected by copyright. All rights reserved.
Dissertação de Mestrado: Mensuração da estrutura de preferência do consumidor: uma aplicação de Conjoint Analysis em Marketing
In this chapter we give some examples of marketing models which have been estimated using the general linear model. Most of these models have been estimated using aggregate demand data. Aggregate
demand refers to the demand across a sample of customers or households and can be measured at levels such as store, chain and market demand.
Empirische Befunde legen nahe, dass Preis-Absatz-Funktionen entgegen der oft unterstellten Modellannahme eines linearen Verlaufs häufig eher einen nicht-linearen, s-förmigen Verlauf aufweisen. Vor diesem Hintergrund werden sowohl die theoretischen Grundlagen als auch die Ergebnisse empirischer Untersuchungen solcher nichtlinearer Funktionen vorgestellt. Darüber hinaus wird die Ermittlung von gewinnoptimalen Preisen mithilfe des Newton-Raphson-Verfahrens sowie mit MS-Excel erläutert und anhand von Beispielen illustriert.
The supply of scallops in the United States in recent years has remained relatively stable and the supply source has shifted noticeably from domestic production to imports, especially from China and Japan, where aquaculture production of scallops has been successful. During the past decade, the market share of imported scallops has increased drastically. The scallop fishery in the U.S. is now facing potential competitions from imports. Continuing efforts to provide information on the effect of increasing imports on the U.S. scallop market is warranted. In addition, the impacts of supply increases on domestic scallop prices, either through wild stock enhancement programs or aquaculture operations, merit further investigation. An Annington model is developed in this study to assess the demand for scallops in the United States. Time series data for 1980-1998 are used in the estimation of the model. Results indicate that the demand for U.S. domestic scallops is less elastic in the short run than in the long run. The substitution elasticities and the cross-price elasticities of U.S. scallops with respect to scallop imports are relatively small, indicating that they may be imperfect substitutes. U.S. domestic scallops and scallop imports from other countries may serve different market segments, and there exists little direct competition between them.
By reviewing the literature we developed principles to guide market analysts in their use of econometric models to forecast market share. We rely on the general principles for econometric forecasting developed by Allen and Fildes (2001) to arrive at specific principles. The theoretical and empirical evidence indicates that they should use econometric market share models when
1.
effects of current marketing activity are strong relative to the carryover effects of past marketing activity,
2.
there are enough observations,
3.
the models allow for variation in response for individual brands,
4.
the models are estimated using disaggregate (store-level) data rather than aggregate data,
5.
the data exhibit enough variability, and
6.
competitors’ actions can be forecast with reasonable accuracy.
In most situations the first five conditions can be satisfied. Condition 6 is more difficult to satisfy and is a priority area for further research. If one or more of the conditions are not satisfied then an extrapolation or judgment forecasting method may be more appropriate.
Purpose
– The purpose of this paper is to see the role of advertising in the middle of sales-price relationship of two entrepreneur brands in India.
Design/methodology/approach
– Quarterly data on sales (in units), advertising (net of inflation) and price (net of inflation) have been compiled for two entrepreneur brands over a period 2007-2012. First, elasticity of price is estimated by regressing sales on price. Next, the response of price elasticity to advertising is captured using a semi-logarithmic regression model.
Findings
– Results reveal that price and sales are inversely related and advertising influences price elasticity negatively.
Practical implications
– Findings suggest that entrepreneurs/managers should allocate more funds to advertising and at the same time should charge a higher price point in order to increase revenue.
Originality/value
– By showing a new way of how to measure the effectiveness of advertising beyond traditional ones (inform, persuade or remind) of two entrepreneur brands this research definitely adds some value in the literature of marketing.
Expiration dates are used by couponers to limit their financial liability temporally. Traditional wisdom assumes that coupon redemptions are greatest in the period immediately following the coupon drop and decline monotonically. Using regret theory, the authors predict that expiration dates induce a second mode in the redemption pattern just prior to the expiration date. They test this prediction by extending an existing coupon redemption model to incorporate an expiration effect and then estimating both the existing and the expiration models using weekly coupon redemption data for spaghetti sauce from A.C. Nielsen panels in two cities. Results are consistent with their prediction. Implications for practitioners and researchers are discussed.
Market-share analysis focuses on the competitive interrelations between products or brands. Marketing activity may affect the performance of a company's own product and that of its competitors not only within a single time horizon but also over several extended periods. Starting from a static market-share analysis model, the dynamic relationships of market shares between competitive brands are described by multiplicative competitive interaction (MCI) time-series models, in which the problem of logical consistency for estimated shares is resolved. A Bayesian shrinkage estimator solution is applied to the further problem of model-induced collinearity in cross-differential MCI models. Dynamic elasticity is defined and used to measure the delayed and long-term effects of marketing mix variables on market shares. The dynamic relationships of future market shares are predicted by means of predictive density. Strategic simulations are conducted under several scenarios for marketing planning. It is argued that the new dynamic model proposed here, applied to daily national or store tracking data, provides useful insights into dynamic competitive relationships in the marketplace, to the benefit of corporate planners, marketing directors, brand managers and retail strategists.
The decision to build market share has major resource-allocation implications. To aid managers in assessing these implications, research was conducted to determine general relationships between changes in market share and variables representing market strategies and competitive position. The research was based on multiproduct, cross-sectional regression analyses and includes variables that are—or should be—readily available to most businesses.
The concept of generic strategies for gaining competitive advantage has received considerable attention recently in the business policy field. Two generic strategies usually mentioned are low cost position and highly differentiated position. This paper uses a game-theoretic model of oligopolistic competition to provide analytical support for these generic strategies, and, in places, to refine the conclusions drawn from previous research in this field. Another conclusion derived from the model is that a superior cost or differentiation position leads to a larger market share, which in turn leads to higher profitability.
This paper describes an experimentation methodology to measure how demand varies with price and the results of its application at a toy retailer. The same product is assigned different price-points in different store panels and the resulting sales are used to estimate a demand curve. We use a variant of the k-median problem to form store panels that control for differences between stores and produce results that are representative of the entire chain. We use the estimated demand curve to find a price that maximizes profit. Our experiment yielded the unexpected result that demand increases with price in some cases. We present likely reasons for this finding from our discussions with retail managers. Our methodology can be used to analyze the effect of several marketing and promotional levers employed in a retail store besides pricing.
This paper provides some observations on the state of the art in marketing research. The primary and secondary thrust of the
articles published in the Journal of Marketing Research during 1980–1986 is briefly reviewed to identify important areas of
inquiry. In each of these areas, we summarize recent developments, highlight the state of the art, offer some critical observations,
and identify directions for future research. A cross-classification of various techniques and problem areas is also presented
and some observations are made on the application of these techniques to address specific substantive and methodological issues
in marketing research.
Share equations play an important role in applied economic research, notably in marketing and demand analysis. Both market
shares and budget shares have been used as dependent variables in econometric models which were partly motivated by microeconomic
theory. However attempts of econometricians (and other statisticians) to treat share equations adequately led mostly to unsatisfactory
approaches: Some researchers although admitting that shares satisfy a sum constraint simply repressed the fact that shares
cannot be normally distributed. Some researchers looked in vain for a stochastic specification which at the same time is consistent
and allows a flexible covariance structure. Last not least almost nobody has properly taken care of additional problems arising
from dynamic share models. The paper discusses these three issues and proposes a possible way out of this dilemma which was
first suggested by Aitchison (1982) and has been applied to econometric demand analysis by Considine and Mount (1984). Demandtheoretic
implications as well as methods of estimation are discussed. An example using German import data illustrates some of the results.
There are two distinct groups of emerging issues in the area of sales forecasting and decision support systems: methodological
issues and implementational issues. This paper discusses each of them. The discussion of implementational issues includes
some results of a survey of forecasting practice. The paper serves as a preface to the special section on sales forecasting
and decision support systems and introduces the papers included in the special section.
A topic that has received quite some attention in the seventies and eighties is logical consistency of sum-constrained linear models. Loosely defined, a sum-constrained model is logically consistent if the restrictions on the parameters and explanatory variables are such that the sum constraint is automatically satisfied. The literature on logical consistency, however, has not been unambiguous. The main reason for this is that a rigorous definition of logical consistency has not been given in the (marketing) literature. Inspired by an extensive discussion by Koehler andWildt (1981), we therefore present two closely related definitions of logical consistency. By using an elegant and direct approach, we derive necessary and sufficient conditions for the sum-constrained linear model to be logically consistent. We summarize the results on this topic up until now, and we clarify some differences and obscurities. We also show that some generally accepted results of McGuire and Weiss (1976) are not correct.
Most research and writing in the area of retail store location and market analysis has been undertaken by marketing researchers and urban geographers. The purposes of this review are to present this eclectic area of study as a unified body of knowledge, to encourage real estate researchers to become more involved in this area, and to indicate where recent research in real estate and urban economics is an outgrowth of study in this area. The field can be viewed as a hierarchy of investigations, including modeling consumer behavior at the micro level, using models of consumer behavior in models of store location, and developing approaches for choosing among markets at the macro level. The hierarchy of investigations should in turn be viewed as the study of economic activity in space. Recent research in the areas of allocation of space in shopping centers and price discrimination in retail leasing is an outgrowth of work in the area of retail store location and market analysis.
Least squares estimation techniques are developed for a special multiplicative model based on the Luce choice axiom whose potential usefulness in marketing applications justifies estimation techniques which can be easily implemented.
Least squares estimation techniques are developed for a special multiplicative model based on the Luce choice axiom whose potential usefulness in marketing applications justifies estimation techniques which can be easily implemented.
In this paper we first generalize the Dorfman-Steiner theorem to the case of an oligopoly with multiple competitive reactions and expansible industry demand. We then review the literature through a series of corollaries. On the basis of the theorem developed, we attempt to shed light on a oligopolistic market over a period characterized by the raise of a leader.
Many marketing models use variants of the relationship: Market share equals marketing effort divided by total marketing effort. Replacing marketing effort with its resulting “attraction,” the relationship is derived from the assumptions: (1) attraction is nonnegative, (2) equal attractions imply equal shares, and (3) a seller’s share is affected the same if the attraction of any other seller increases a fixed amount.
Many marketing models use variants of the relationship: Market share equals marketing effort divided by total marketing effort. Replacing marketing effort with its resulting “attraction,” the relationship is derived from the assumptions: (1) attraction is nonnegative, (2) equal attractions imply equal shares, and (3) a seller's share is affected the same if the attraction of any other seller increases a fixed amount.
The purpose of this report is to discuss, explain, and position the more notable literature on logically consistent market share demand models. More specifically, the report explains and provides examples for the analysis by McGuire et al., discusses the recent paper by Naert and Bultez, and comments on a recent related article by Beckwith.
The sales responses of five competing brands to advertising were investigated, with parameters of the simple market response functions for all brands estimated jointly. The effectiveness of advertising was found to differ significantly among brands. Optimal advertising expenditures implied by the model were determined for each of the brands.
In this paper a method of estimating the parameters of a set of regression equations is reported which involves application of Aitken's generalized least-squares [1] to the whole system of equations. Under conditions generally encountered in practice, it is found that the regression coefficient estimators so obtained are at least asymptotically more efficient than those obtained by an equation-by-equation application of least squares. This gain in efficiency can be quite large if “independent” variables in different equations are not highly correlated and if disturbance terms in different equations are highly correlated. Further, tests of the hypothesis that all regression equation coefficient vectors are equal, based on “micro” and “macro” data, are described. If this hypothesis is accepted, there will be no aggregation bias. Finally, the estimation procedure and the “micro-test” for aggregation bias are applied in the analysis of annual investment data, 1935–1954, for two firms.
The sales responses of five competing brands to advertising were investigated, with parameters of the simple market response functions for all brands estimated jointly. The effectiveness of advertising was found to differ significantly among brands. Optimal advertising expenditures implied by the model were determined for each of the brands.
Consistent sum-constrained models have often been considered intrinsically nonlinear. We study here a linearizable variant which can be applied to a wide variety of situations of interaction between socioeconomic flows. The properties of the error term of such a model are derived and we show how its parameters can be estimated using a generalized least squares procedure proposed by McGuire et al. [14]. It is further demonstrated how adding realism to the model may cause serious multicollinearity problems. We then compare it with more classical multiplicative forms. The various models discussed are tested for the market of an inexpensive consumer durable product.
Subsets of the dependent variable of a linear model often conform to known constraints; for example, this situation arises with Engel curves and with models which estimate transition probabilities, market shares, or other classes of proportions. These models imply a variety of restrictions on the parameters, explanatory variables and residuals which must be treated explicitly in estimating and testing hypotheses about the parameters. This paper discusses these restrictions and develops an estimating procedure which yields consistent estimates which are asymptotically efficient, unbiased and normal.
The purpose of this report is to discuss, explain, and position the more notable literature on logically consistent market share demand models. More specifically, the report explains and provides examples for the analysis by McGuire et al., discusses the recent paper by Naert and Bultez, and comments on a recent related article by Beckwith.
Characterization of Admissible Logically Consistent Parameters for Constrained Linear Models and their Estimation
- G J Koehler
- A R Wildt
Koehler, G. J. and A. R. Wildt (1977), "Characterization
of Admissible Logically Consistent Parameters for Constrained Linear Models and their Estimation," Working
Paper, College of Business Administration, University of
Florida, August.
“Restrictions on Linear Sum-Constrained Models: A Generalization,”
- M Weverbergh
“Estimating Gravitational Market Share Models,”
- A V Bultez
- P A Naert
Working Paper 73-36, European Institute for Advanced Studies in Management
- A V Bultez
- P A Naert
Bultez, A. V. and P.A. Naert (1973), "Estimating Gravitational Market Share Models," Working Paper 73-36,
European Institute for Advanced Studies in Management,
Brussels.
--and--(1975), "Consistent Sum-Constrained
Models," Journal of the American Statistical Association,
70, 529-35.