Malcolm Wright

Purchase Loyalty is Polarised into either Repertoire or Subscription Patterns

Abstract We have observed that competitive repeat purchase markets are polarised into two radically different structures. The first and best known we call repertoire markets; these have few solely loyal buyers as most buyers allocate their category requirements across several brands in a steady fashion. The other we call subscription markets; these have many solely loyal buyers as most buyers allocate category requirements entirely to one brand. This is an empirical difference rather than a theoretical distinction, and surprisingly there appear to be no markets which occupy the middle ground between these two extremes. The repertoire-subscription distinction turns out to be an important boundary condition for some well-established generalisations about repeat purchase behavior. Despite this, the NBD-Dirichlet model of purchase incidence and brand choice fits both types of markets, and the differences in loyalty are adequately captured by the Dirichlets switching parameter, S. This represents an important extension of the generalisability of the Dirichlet, allowing the insights gained from repertoire market analysis to be applied to customer churn analysis in subscription markets.

Market statistics for the Dirichlet model: Using the Juster scale to replace panel data

Abstract The well-known NBD-Dirichlet model of purchase incidence and brand choice is usually estimated from aggregate market statistics such as penetration, average purchase frequency, and market share. In practice, panel data have been the only accurate source of this information. This research developed estimators of these market statistics based on the Juster scale, an 11-point purchase probability scale that can be applied using a survey. A validation study, involving comparisons with panel data for 16 brands in three categories, showed that the Juster-based estimators performed very well. This result allows the data requirements of the NBD-Dirichlet model to now be satisfied using a survey.

A contingency model of marketing information

Marketing information systems have usually been analysed in terms of market research, market intelligence and computerised modelling and analysis systems. Both empirical and theoretical studies are making it increasingly clear that this approach is inadequate, especially as it takes too little account of the role of organisation design methods and environmental contingencies in marketing information processing. A new approach is suggested which draws on these areas to provide guidelines for research and intervention into marketing information flows. Exploratory case studies of three diverse organisations broadly confirmed this new approach, opening the way for further empirical work.

Descriptive and evaluative attributes: what relevance to marketers?

A growing number of studies have suggested that consumers hold very similar beliefs about the brands they use within a product category. This implies that experience, rather than marketing activity, leads consumers to associate attributes with brands. Replicates and extends earlier studies and addresses methodological criticisms directed at that work. Our findings reveal that descriptive attributes can be successfully predicted and they confirm that the usage level of a brand typically determines the proportion of consumers who hold favourable attitudes about that brand. The results question the popular emphasis on positioning and brand differentiation and we conclude by suggesting more behaviourally oriented strategies.

The dubious assumptions of segmentation and targeting

Discusses Mitchell’s recent proposal for astrological segmentation which demonstrates the extremes to which segmentation and targeting can be taken if we uncritically accept their core assumptions. Proposes that although Mitchell’s proposal can be subjected to a number of minor criticisms, it can only be finally disposed of by critically examining whether astrological segments really are associated with a stable set of preferences, and whether targeting these segments actually gives a higher return than other approaches. Once the stability of segments, the logic of targeting, and the empirical evidence are examined, it turns out that not only is Mitchell’s approach unsupported, but so are most other segmentation and targeting efforts.

It's a Dirichlet World: Modeling Individuals' Loyalties Reveals How Brands Compete, Grow, and Decline

ABSTRACT The Dirichlet is one of the most important theoretical achievements of marketing science. It provides insights into the distribution of consumer loyalties and is used widely in industry for benchmarking and interpreting brand performance. The Dirichlets implications run counter to some well-entrenched marketing pedagogy and so, unsurprisingly, it has attracted criticism arguing that it cannot adequately reflect the dynamic nature of consumer choice. The authors address these criticisms by discussing how consumer loyalties are manifested and examining whether changes in consumer loyalties do, in fact, disrupt Dirichlet buying patterns. To the best of our disciplines knowledge, based on extensive empirical and theoretical work, brands compete in a Dirichlet world.

Are Australasian brands different

Over the last 30 years a range of empirical generalisations has been developed about the performance of competitive brands in frequently purchased product categories. These generalisations have been based mainly on European and US data, and this paper addresses the question of whether they also hold in Australia and New Zealand. We examined consumer panel data from four different markets (supermarkets, department stores and retail fuel in Australia and retail fuel in New Zealand) and found similar patterns to those in Europe and the USA, although there were some minor exceptions, and also some interesting variations between markets. Our results suggest that there is much that Australasian marketers can learn from using models such as the Dirichlet, which was developed in the Northern hemisphere, to identify norms and exceptions in their own markets.

Predicting blood donation behaviour: further application of the theory of planned behaviour

Purpose – The purpose of this paper is to test whether, in the context of blood donation, the predictive ability of the theory of planned behaviour (TPB) extends from behavioural intention to actual donation behaviour, and whether extended versions of the TPB perform better than the standard version.Design/methodology/approach – Intentions to donate blood predicted by the TPB are compared with an accurate measure of blood donation behaviour obtained following a mobile blood drive by the New Zealand Blood Service.Findings – When the observed outcome is donation behaviour rather than behavioural intention, the TPB models performance drops. Extending the variables in the model to include moral obligation and past behaviour does not improve its predictive ability, and neither does the use of belief‐based variables.Practical implications – The TPB is much less effective in predicting blood donation behaviour than it is in predicting intentions to donate blood. But only actual donation behaviour yields medical...

Double jeopardy in brand defection

Purpose – The purpose of this paper is to test whether brand defection shows double‐jeopardy effects, and whether stochastic models provide useful benchmarks of expected brand defection rates.Design/methodology/approach – The approach takes the form of an empirical study of brand defection in four markets using panel data, comparing the performance of simple OLS models with a stochastic model.Findings – Brand defection shows double jeopardy. Almost all brand defection in the markets studied could be explained by the category examined and the market share of the focal brand. A stochastic model of choice fits these data well, and provides many further practical and theoretical applications.Practical implications – The study provides improved benchmarks for brand defection, allowing managers to spot whether their brand is performing better or worse than expected. It also allows better analysis of market structure for subscription services, especially under dynamic conditions, and better estimation of custome...

Does the duplication of viewing law apply to radio listening

Purpose – There has been long‐standing interest in the duplication of audience between media vehicles, starting with work by Agostini and later developed by Goodhardt, Ehrenberg and Collins into the “duplication of viewing law”. The aim of this paper is to further extend duplication analysis to radio listening. As radio markets are believed to have many partitions, the paper considers whether an un‐partitioned duplication analysis provides an adequate description of market structure.Design/methodology/approach – The paper reports the results of a weekly radio diary with 1,129 responses in a regional New Zealand radio market. This data has special characteristics suitable for this research: the market has experienced rapid expansion in station numbers with substantial attempts at format segmentation, providing a strong test of the un‐partitioned nature of the duplication analysis; use of a single regional market avoids the aggregation bias inherent in national data; use of primary research allows the inclu...