Past Purchases Research Articles

Identifying Innovators for the Cross-Selling of New Products

With recent advances in information technology, most companies are amassing extensive customer databases. The wealth of information in these databases can be useful in identifying those customers most likely to purchase a new product and in predicting when this adoption may take place. This can assist database marketers in determining when individuals should be targeted for the promotion of a new product, which may increase the efficiency of manufacturing and distribution, and assure a faster return on investments. For this purpose, we propose a model that considers the timing of past purchases across multiple product categories and produces estimates of each customer's propensity of ever purchasing in a particular product category and of the timing of their purchases. The model is designed to help managers identify the best prospects for a new offer in one of multiple categories based on generalizations obtained from past offers. The proposed model also provides projections of aggregate penetration for new brands within the database, based on sample estimates.

Customer repurchase intention

This paper develops a general service sector model of repurchase intention from the consumer theory literature. A key contribution of the structural equation model is the incorporation of customer perceptions of equity and value and customer brand preference into an integrated repurchase intention analysis. The model describes the extent to which customer repurchase intention is influenced by seven important factors – service quality, equity and value, customer satisfaction, past loyalty, expected switching cost and brand preference. The general model is applied to customers of comprehensive car insurance and personal superannuation services. The analysis finds that although perceived quality does not directly affect customer satisfaction, it does so indirectly via customer equity and value perceptions. The study also finds that past purchase loyalty is not directly related to customer satisfaction or current brand preference and that brand preference is an intervening factor between customer satisfaction and repurchase intention. The main factor influencing brand preference was perceived value with customer satisfaction and expected switching cost having less influence.

Upgrading, Degrading, and Intertemporal Price Discrimination

The paper studies monopoly pricing of a vertically differentiated durable good in a two-period model. It provides an explanation for seemingly unusual practice of a firm selling a "degraded good," arguing that the presence of Coasian dynamics may lead to the sale of the degraded good that is not less costly to produce than a high-quality good. The main finding is that when the firm can identify previous customers only if they voluntarily reveal their past purchases, it sells the degraded good along with the high-quality good in the first period. When the firm sells an upgrade of the degraded good, the price of the high-quality good cannot be "too low" in the second period, since otherwise the upgrading customers would pretend to be new customers. Thus the firm can enhance first-period sales while mitigating consumers' incentive to wait until the next period.

Psychosocial and Demographic Variables Associated with Consumer Intention to Purchase Sustainably Produced Foods as Defined by the Midwest Food Alliance

Objective To examine psychosocial and demographic variables associated with consumer intention to purchase sustainably produced foods using an expanded Theory of Planned Behavior. Design Consumers were approached at the store entrance and asked to complete a self-administered survey. Setting Three metropolitan Minnesota grocery stores. Participants Participants (n = 550) were adults who shopped at the store: the majority were white, female, and highly educated and earned ≥ $50,000/year. Participation rates averaged 62%. Main Outcome Measures The major domain investigated was consumer support for sustainably produced foods. Demographics, beliefs, attitudes, subjective norm, and self-identity and perceived behavioral control were evaluated as predictors of intention to purchase them. Analysis Descriptive statistics, independent t tests, one-way analysis of variance, Pearson product moment correlation coefficients, and stepwise multiple regression analyses ( P<.05). Results Consumers were supportive of sustainably produced foods but not highly confident in their ability to purchase them. Independent predictors of intention to purchase them included attitudes, beliefs, perceived behavioral control, subjective norm, past buying behavior, and marital status. Conclusions and Implications Beliefs, attitudes, and confidence level may influence intention to purchase sustainably produced foods. Nutrition educators could increase consumers' awareness of sustainably produced foods by understanding their beliefs, attitudes, and confidence levels.

Customer Reactions to Service Provider Overgenerosity

Existing research in services marketing has mostly focused on service encounters in which the service provider fails to meet customer expectations. Service encounters in which customer expectations are exceeded have not been as thoroughly examined. This article focuses on service encounters characterized by service provider generosity: the act of giving customers value beyond their expectations. Utilizing an experimental research design, the authors demonstrate that consumer reactions to such behavior are moderated by the organizational level of the service provider as well as the existence of a past purchase history. Counter to conventional belief, under certain conditions, overgenerosity may hinder customer assessments of a service. The article concludes with a discussion of the implications for services marketing strategy and research.

An interview-based measure of short-term advertising effects

Often, advertisers do not have real (electronic) single-source data available for the evaluation of their advertising campaigns. In this article a system is described which may prove useful for such advertisers. It is useful because real purchases are substituted by self- reported past purchase behaviour and real exposure is substituted by self-reported recalls of past exposures. In doing so, the concept of attitudinal short-term advertising strengths (ASTAS) is introduced. ASTAS is an attitudinal measure of short-term advertising strengths (STAS)-like effects of different campaigns, brands and media. It was developed in the Nordic countries in the absence of ‘real single-source data’ and within a media environment dominated by print advertising. As it turns out, the measure contains promising qualities that may be used outside the Nordic region too.In essence, the measure looks at the way in which people report advertising recall and past purchases, both when advertising is being broadcast and when it is not. In this way, it provides useful feedback for advertisers in small markets as well as others who do not have access to single-source data.

A Guadagni-Little Likelihood Can Have Multiple Maxima

Despite many advances in marketing models, the Guadagni–Little (1983) model is still in widespread use by both practitioners and academics. For many new marketing models, the Guadagni–Little model serves as a benchmark. The key variable that allows the Guadagni–Little model to accurately fit data is the loyalty variable, which is an exponential smoothing of past purchases. In this paper, I show that inclusion of this variable in the logit model may result in a likelihood function that can have multiple maxima. I am able to demonstrate this using simulated data and actual household scanner panel data. In addition, I document a systematic relationship between the loyalty coefficient and the loyalty smoothing parameter. Insight for this systematic relationship and the multiple maxima is obtained by recognizing a trade-off between capturing household heterogeneity and state dependence. Finally, in the Guadagni–Little model extreme parameter values capture two different idealized forms of consumer behavior. However, reported studies rarely find these extreme parameter values. I show that procedures commonly used to initialize loyalty biases against these extreme parameter values. This bias offers some explanation for the observed empirical regularity in Guadagni–Little parameter estimates and suggests that researchers should be cautious concluding these parameters capture regularity in consumer behavior.

The Demand for Private Medical Insurance in the UK: A Cohort Analysis

This paper examines the determinants of the demand for private health insurance in the United Kingdom from 1978 to 1996 using a pseudo-cohort panel. The focus is on the impact of public and private sector quality, generational change, and past purchase on demand. The results indicate that there has been generational change in buying behaviour, that the number of senior doctors employed in the public sector impacts upon demand for the private alternative, and that there is limited impact of habit in purchase. Changes to the structure of labour contracts in the NHS may affect demand for the private alternative. Since the mid-1980s governments in the United Kingdom have sought to increase the extent of private finance for welfare provision. The state has withdrawn from the public finance of housing and made widespread changes to pension provision. While public funding for hospital and primary health care has been maintained, the scope of state financed care has been reduced at the margins. The result has been a considerable growth in the numbers buying private health care (Burchardt et al., 1999). This paper examines the determinants of the demand for private health insurance in the United Kingdom. Understanding the purchase of private health insurance in the United Kingdom is of general interest because many governments are trying to encourage greater private finance of health care, but is of more specific interest because it provides an understanding of the interaction between public and private welfare provision. Private medical insurance in the United Kingdom is supplementary to NHS care. Purchase is optional and purchase does not preclude use of the NHS nor, with the exception of the minor amount of tax relief to the over 60s, does it reduce the buyer's contributions to the tax-financed NHS. At a theoretical level, models of

Segmentation-based competitive analysis with MULTICLUS and topology representing networks

Two neural network approaches, Kohonen's self-organizing (feature) map (SOM) and the topology representing network (TRN) of Martinetz and Schulten are employed in the context of competitive market structuring and segmentation analysis. In an empirical study using brands preferences derived from household panel data, we compare the SOM and TRN approach to MULTICLUS, a parametric latent vector multi-dimensional scaling (MDS) model approach which also simultaneously solves the market structuring and segmentation problem. Our empirical analysis shows several benefits and shortcomings of the three methodologies under investigation. As compared to MULTICLUS, we find that the non-parametric neural network approaches show a higher robustness against any kind of data preprocessing and a higher stability of partitioning results. As compared to SOM, we find advantages of TRN which uses a more flexible concept of adjacency structure. In TRN, no rigid grid of units must be prespecified. A further advantage of TRN lies in the possibility to exploit the information of the neighborhood graph for adjacent prototypes which supports ex-post decisions about the segment configuration at both the micro and the macro level. However, SOM and TRN also have some drawbacks as compared to MULTICLUS. The network approaches are, for instance, not directly accessible to inferential statistics. Our empirical study indicates that especially TRN may represent a useful expansion of the marketing analyst's tool box. Scope and purpose Determination of competitive market structure among rival brands and market segmentation represent well-known concepts in strategic marketing planning. During the last decade, approaches that combine the two interrelated tasks into one single model have been introduced into marketing literature. Most of them respect consumer heterogeneity by including ‘fixed’ parameters (e.g., demographic or past purchase behavior variables) for each individual or by assuming consumer parameters to be distributed according to a (mixture of) probability distribution(s). However, the key to the success of these statistical modeling approaches seems to lie in the proper choice of parametric model assumptions and/or heterogeneity distributions. Due to its non-parametric nature, the neuro-computing methodology presented in this article imposes less rigorous assumptions on data properties and derives segment-specific patterns of competitive relationships between brands in a purely data-driven way.

Market Segmentation of Green Consumers in Hong Kong

Market segmentation is one of the useful tools for marketers to define target markets. An intercept sample survey of 704 shoppers in Hong Kong was conducted to segment the market based on the past purchase of environmentally friendly as well as not-so-friendly products. Chi-square analysis and stepwise discriminant analysis were conducted to differentiate heavy and light green consumers using demographics and other environmental variables including green consumerism knowledge and perception about environmentally friendly products. It was found that heavy green consumers were more likely to have a higher education and a higher household income. They perceived that environmentally friendly products were good for their health and helped to save resources. Heavy green consumers were more likely to report that they perceived influence from other persons, the government and the green groups. They had a strong self-identity and think of themselves as green consumers and as someone who was concerned with environmental issues. They possessed a better knowledge about green consumerism and more frequently used the mass media for environmental news. Light green consumers found environmentally friendly products difficult to access. Implications for green marketers and manufacturers are discussed.

Markov chain Monte Carlo and models of consideration set and parameter heterogeneity

In this paper the authors propose an integrated consideration set-brand choice model that is capable of accounting for the heterogeneity in consideration set and in the parameters of the brand choice model. The model is estimated by an approximation free Markov chain Monte Carlo sampling procedure and is applied to a scanner panel data. The main findings are: ignoring consideration set heterogeneity understates the impact of marketing mix and overstates the impact of preferences and past purchase feedback even when heterogeneity in parameters is modeled; the estimate of consideration set heterogeneity is robust to the inclusion of parameter heterogeneity; when consideration set heterogeneity is included the parameter heterogeneity takes on considerably less importance; the promotional response of households depends on their consideration set even if the underlying choice parameters are identical.

Defence implications of East Asia's crisis

East Asia's economic crisis has greatly reduced the arms-procurement programmes of several leading states, with important results for regional security. Because China has so far been relatively immune to the crisis, it has been able to maintain its programmes, therefore increasing its regional strategic weight and re-emphasising the military dependence of other states on the US. On the other hand, many of the past military purchases by states in the region were expensive, unnecessary and provocative, and the new economic climate may bring with it a welcome new sobriety.

Consumer Preferences for Geranium Flower Color, Leaf Variegation, and Price in Five U.S. Markets

Consumers in five U.S. markets evaluated photographs of geranium plants with regard to purchase likelihood. Photographic images were colored electronically to produce uniform geranium plants with five flower colors (pink, white, red, lavender, and blue) and three leaf variegation patterns (dark zone, white zone, and no zonal pattern). Photographs were mounted on cards with five selected price points ranging from ($1.39 to $2.79). We randomly generated an orthogonal array, partial-factorial design for consumers to rate a reduced number of choices. Consumers shopping in cooperating garden centers located in Dallas, Texas; Montgomery, Ala.; Athens, Ga.; Charlotte, N.C.; and Wilmington, Dela., rated 25 photographs on the basis of their likelihood to purchase the plants shown. Conjoint analysis revealed that customers in the Georgia garden center placed the highest proportion of their decision to buy on leaf variegation (29%), while customers in the Alabama outlet placed the most emphasis on price (46% of the decision). Shoppers in Texas valued flower color most highly (58% of their decision to buy). Demographic characteristics and past purchase behavior also varied widely, suggesting diverse marketing strategies for geraniums.

A Dynamic Analysis of Market Structure Based on Panel Data

Internal market structure analysis infers brand positions in an attribute space from preference and choice data, given a market in which consumers have heterogeneous tastes for attributes. Previous market structure models have adopted a static framework (e.g., Elrod 1988, Chintagunta 1994, Elrod and Keane 1995). Furthermore, they assumed that consumer perceptions of brand attributes do not vary across consumers. Yet, these approaches may render inaccurate representations of market structure if there is state dependence in consumer choice behavior. This paper attempts to incorporate consumer choice dynamics into market structure models by specifying the source of choice dynamics explicitly. In particular, the process by which past purchases affect current choices is modeled in a framework which captures both consumer habit persistence and variety seeking behavior. More specifically, consumer preferences for brand attributes are modeled to depend on the attributes of brands bought on the previous purchase occasion. Furthermore, the modeling approach adopted incorporates heterogeneity in both consumer preferences and perceptions of brand attributes. The audience of this paper includes practitioners and academics interested in understanding consumer choice processes and inferring market structure from consumer choice data. The proposed models are estimated on Nielsen scanner panel data for margarine, peanut butter, yogurt, and liquid detergent using simulated maximum likelihood techniques. The empirical results suggest that accounting for choice dynamics improves both in-sample and out-of-sample fit. The results indicate that the average consumer is habit persistent in all the product categories studied. This result is consistent with the findings of Kannan and Sanchez (1994), who conducted an aggregate analysis of consumer variety seeking behavior across product categories. However, the results obtained in this paper suggest that consumers are heterogeneous with respect to the processes by which past purchases affect current purchases. These results provide strong evidence for habit persistence and variety seeking in brand attributes to be the behavioral source of consumer choice dynamics in food categories. Thus, consumer tastes (utility weights) seem to be affected by the attributes of the brands consumed in the past. Given the empirical result that a large proportion of consumers are habit persistent, this suggests that tastes are reinforced by the brand attributes consumed in the past. The empirical results also show that not accounting for state dependence in market structure models for panel data may produce misleading results, that is, depending on consumer behavior patterns, models that do not account for state dependence may distort the true nature of competition among brands. More specifically, the results confirm the expectation that if there is habit persistence, that is, if consumer tastes are reinforced by attributes of brands consumed in the past, models that do not capture this choice dynamics will overestimate the distance between (similar) brands. Furthermore, the policy experiments conducted suggest that (1) static models overestimate the short-run impact of a price cut on the sales of the brand on promotion, (2) price cuts hurt the sales of the more similar brands more, and (3) free samples affect relatively less similar brands the most. Finally, this paper studies variety seeking and habit persistence across brands over purchase occasions. However, variety seeking behavior may also involve the purchase of a portfolio of brands or items at a purchase occasion. Consumers may buy multiple items knowing that prior to the next trip they may want to consume different items (Simanson 1990, Walsh 1995). This type of behavior can be modeled within the context of dynamic expected utility maximization with forward-looking consumers. The development and estimation of market structure models that include forward-looking consumers who maximize expected-utility over a planning horizon, incorporating their future tastes and needs, and shopping for an inventory of brands, remain an important future research issue.

The Value of Purchase History Data in Target Marketing

An important aspect of marketing practice is the targeting of consumer segments for differential promotional activity. The premise of this activity is that there exist distinct segments of homogeneous consumers who can be identified by readily available demographic information. The increased availability of individual consumer panel data open the possibility of direct targeting of individual households. The goal of this paper is to assess the information content of various information sets available for direct marketing purposes. Information on the consumer is obtained from the current and past purchase history as well as demographic characteristics. We consider the situation in which the marketer may have access to a reasonably long purchase history which includes both the products purchased and information on the causal environment. Short of this complete purchase history, we also consider more limited information sets which consist of only the current purchase occasion or only information on past product choice without causal variables. Proper evaluation of this information requires a flexible model of heterogeneity which can accommodate observable and unobservable heterogeneity as well as produce household level inferences for targeting purposes. We develop new econometric methods to implement a random coefficient choice model in which the heterogeneity distribution is related to observable demographics. We couple this approach to modeling heterogeneity with a target couponing problem in which coupons are customized to specific households on the basis of various information sets. The couponing problem allows us to place a monetary value on the information sets. Our results indicate there exists a tremendous potential for improving the profitability of direct marketing efforts by more fully utilizing household purchase histories. Even rather short purchase histories can produce a net gain in revenue from target couponing which is 2.5 times the gain from blanket couponing. The most popular current electronic couponing trigger strategy uses only one observation to customize the delivery of coupons. Surprisingly, even the information contained in observing one purchase occasion boasts net couponing revenue by 50% more than that which would be gained by the blanket strategy. This result, coupled with increased competitive pressures, will force targeted marketing strategies to become much more prevalent in the future than they are today.

Memory and Desire

Consumer product demand increases with non-informational advertising and with past product purchases. To these stylized facts, a third observation can be added: When agents are in a hurry, they are more likely to buy the product that they bought last time or the product which they saw recently on television. These three observations can be explained by using three stylized facts about human memory. These facts are: 1) Memory contains information about products. 2) Memory takes time to operate. 3) Memory is faster when the information has been recalled more frequently in the past. Remembering information in memory takes time, but time has a cost. Therefore the information in memory has a cost. This information cost rises with the cost of time and decreases with the speed of recall. The speed of recall increases with the frequency that the information has been recalled in the past. Therefore, the cost of information declines with the number of times the information has been previously recalled. This paper hypothesizes that advertising and product use cause agents to recall product information. This, in turn, makes recalling product information faster and cheaper relative to other goods. Therefore, the product is more likely to be recalled and purchased when agents have seen product advertising or used the product in the past. In addition, an increase in recall speed, caused by advertising and product use, reduces the relative cost of recalling product information more when the cost of time is higher. So, advertising and past product use will have larger marginal effects on product demand when the cost of time is higher. That is, when agents are in a hurry, they will be more likely to buy highly advertised or frequently used products than when their cost of time is lower. The process of recall can be viewed as an optimal stopping problem. So, the formal framework of this model is similar to the standard sequential search model'. However, the spirit of the paper is closest to Salop and Stiglitz (1977) and Hauser and Wernerfelt (1990). They use a Stigler-type

Decision-Making Under Uncertainty: Capturing Dynamic Brand Choice Processes in Turbulent Consumer Goods Markets

We construct two models of the behavior of consumers in an environment where there is uncertainty about brand attributes. In our models, both usage experience and advertising exposure give consumers noisy signals about brand attributes. Consumers use these signals to update their expectations of brand attributes in a Bayesian manner. The two models are (1) a dynamic model with immediate utility maximization, and (2) a dynamic “forward-looking” model in which consumers maximize the expected present value of utility over a planning horizon. Given this theoretical framework, we derive from the Bayesian learning framework how brand choice probabilities depend on past usage experience and advertising exposures. We then form likelihood functions for the models and estimate them on Nielsen scanner data for detergent. We find that the functional forms for experience and advertising effects that we derive from the Bayesian learning framework fit the data very well relative to flexible ad hoc functional forms such as exponential smoothing, and also perform better at out-of-sample prediction. Another finding is that in the context of consumer learning of product attributes, although the forward-looking model fits the data statistically better at conventional significance levels, both models produce similar parameter estimates and policy implications. Our estimates indicate that consumers are risk-averse with respect to variation in brand attributes, which discourages them from buying unfamiliar brands. Using the estimated behavioral models, we perform various scenario evaluations to find how changes in marketing strategy affect brand choice both in the short and long run. A key finding obtained from the policy experiments is that advertising intensity has only weak short run effects, but a strong cumulative effect in the long run. The substantive content of the paper is potentially of interest to academics in marketing, economics and decision sciences, as well as product managers, marketing research managers and analysts interested in studying the effectiveness of marketing mix strategies. Our paper will be of particular interest to those interested in the long run effects of advertising. Note that our estimation strategy requires us to specify explicit behavioral models of consumer choice behavior, derive the implied relationships among choice probabilities, past purchases and marketing mix variables, and then estimate the behavioral parameters of each model. Such an estimation strategy is referred to as “structural” estimation, and econometric models that are based explicitly on the consumer's maximization problem and whose parameters are parameters of the consumers' utility functions or of their constraints are referred to as “structural” models. A key benefit of the structural approach is its potential usefulness for policy evaluation. The parameters of structural models are invariant to policy, that is, they do not change due to a change in the policy. In contrast, the parameters of reduced form brand choice models are, in general, functions of marketing strategy variables (e.g., consumer response to price may depend on pricing policy). As a result, the predictions of reduced form models for the outcomes of policy experiments may be unreliable, because in making the prediction one must assume that the model parameters are unaffected by the policy change. Since the agents in our models choose among many alternative brands, their choice probabilities take the form of higher-order integrals. We employ Monte-Carlo methods to approximate these integrals and estimate our models using simulated maximum likelihood. Estimation of the dynamic forward-looking model also requires that a dynamic programming problem be solved in order to form the likelihood function. For this we use a new approximation method based on simulation and interpolation techniques. These estimation techniques may be of interest to researchers and policy makers in many fields where dynamic choice among discrete alternatives is important, such as marketing, decision sciences, labor and health economics, and industrial organization.

Modeling Consumers’ Choices of Multiple Items

Contemporary choice models focus on choice opportunities in which consumers purchase a quantity of a single item in a product category. However, failing to recognize the possibility of assortments of multiple-Item purchases can lead to incorrect conclusions about the impact of past purchase behavior on current choices. The authors propose a model that allows for multiple-item shopping trips and apply it to scanner data for powdered soft drinks. The model provides descriptions about the influence of (1) consumer's prior behavior across previous shopping trips, (2) behavior within the same shopping trip, (3) the in-store shelf assortment available at the time of purchase, and (4) marketing mix variables on multiple-item shopping trip choices. The authors’ model provides better choice predictions in a holdout sample at the aggregate and assortment composition levels than a traditional, single-purchase choice model. Using the model, they present simulation and naturally occurring experiment results in the powdered soft drinks category. Finally, they discuss the value of their results for understanding the consequences of consumers’ choices and their implications for manufacturers.

Structural models of brand loyalty with an application to the automobile market

Although several studies have found that past purchases are important determinants of a consumer's current purchase, there is disagreement whether past purchases are capturing brand loyalty or consumer heterogeneity. This article addresses the question of consumer heterogeneity, explores the relationship between brand loyalty and singleton choice sets and estimates alternative structural models that are based on the common assumption that consumers are of two types: brand loyals and shoppers, who face the entire set of available choices. Employing data from a 1989 J. D. Power and Associates survey of new vehicle buyers, aggregate switching and logit captivity/DOGIT models are estimated and comparatively evaluated in holdout testing. Among the results, there is evidence that new vehicle purchasers are brand loyal but that brand loyalty may be overestimated if the model fails to control for availability constraints. Further, a parametrized logit captivity model finds that male and college-educated consumers are important determinants of singleton choice sets.

Customer Base Analysis: An Industrial Purchase Process Application

Customer base analysis is concerned with using the observed past purchase behavior of customers to understand their current and likely future purchase patterns. More specifically, as developed in Schmittlein et al. (1987), customer base analysis uses data on the frequency, timing, and dollar value of each customer's past purchases to infer • the number of customers currently active, • how that number has changed over time, • which individual customers are most likely still active, • how much longer each is likely to remain an active customer, and • how many purchases can be expected from each during any future time period of interest. In this paper we empirically validate the model proposed by Schmittlein et al. In doing so, we provide one of the few applications of stochastic models to industrial purchase processes and industrial marketing decisions. Besides showing that the model does capture key aspects of the purchase process, we also present a more effective parameter estimation method and some results regarding sampling properties of the parameter estimates. Finally, we extend the model to explicitly incorporate dollar volume of past purchases. Our results indicate that this kind of customer base analysis can be both effective in predicting purchase patterns and in generating insights into how key customer groups differ. The link of both these benefits to industrial marketing decision making is also discussed.