Landmark paper: ‘Towards an integrated theory of consumer behaviour’, A.S.C. Ehrenberg
13 January 2016

You can read the landmark paper here: ‘Towards an integrated theory of consumer behaviour’, A.S.C. Ehrenberg.

MRS celebrates its 70th anniversary in 2016, and I feel we cannot pass this milestone by without considering the work of the late Andrew Ehrenberg. Andrew was Chairman of MRS in the mid 1960s and was a driving force in establishing the original MRS journal, then called ‘Commentary’. He was also a leading academic in developing marketing science, especially in exploring factors associated with brand share and consumer purchase behaviour, leading to the formulation of the NBD (Negative Binomial Distribution)-Dirichlet model explaining how buyers vary in their purchase propensities.

This model is described by Byron Sharp in ‘How Brands Grow’ as one of marketing’s first true scientific theories. And, the model continues to be applied in new ways - we’ve published two papers in recent years where it has been applied in exploring the impact of product harm on brand shares in China, for the baby milk and car markets (‘The effects of product-harm crisis on brand performance’, Ma et al, IJMR Vol.52, Issue 4, 2010; ‘The impact of a product-harm crisis on customer perceived value’, Ma et al, IJMR Vol. 56 Issue 3, 2014) .

Andrew’s name, and his huge contribution to the development of marketing science, lives on in the Ehrenberg-Bass Institute for Marketing Science based at the University of South Australia from which IJMR continues to benefit via strong links that provide a steady stream of submissions (a paper by Rachel Kennedy et al, ‘When ‘significant’ is not significant’, published in 56/5, being a Finalist for the MRS Silver Medal in 2015) and members of the Editorial Advisory Board, reviewing the submissions of others.

Having looked through the body of work we’ve published by Ehrenberg in the Journal over the years, I felt that I can do no better than recommend the paper first published way back in Vol.11 Number 4 (October 1969) in JMRS (an issue unfortunately missing from my collection of MRS journals), re-printed in Vol. 38 No. 4 (October 1996), one of two special issues containing reprints of previous ‘classic’, influential papers to celebrate the 50th anniversary of MRS.

In introducing the re-print in 1996, which Ehrenberg calls ‘this youthful folly’, he was struck by how much was already known about consumer behaviour by 1969, bearing mind that in many respects marketing can be described as a child of the 1960s. This was the period when Philip Kotler published his first edition of ‘Marketing Management’ (and introduced us to the ‘Ps’ of marketing strategy) and Jagdish Sheth (his work being cited in the paper) developed his theory of buyer behaviour as a further cornerstone that led to the birth of marketing research.

Ehrenberg’s concern with the original paper was its focus on research, and a lack of potential applications to marketing, but that is not entirely missing from the paper as Table 10, for example, lists practical problems that can be addressed by studying buyer behaviour. He also commented that the use of ‘Toward’ in the title had been overtaken by the establishing of the NBD-Dirichlet model by 1996, and that by then there were numerous applications out there, plus a book developing the ideas introduced in the paper. He also felt that it was taking time for the principles to be accepted and adopted, but you could argue that a considerable proportion of marketing spend today still ignores the ‘truths’ laid down in this paper and the subsequent model about consumer buying behaviour.

This was also the period of motivational research, pioneered by Dichter in the USA in the 1950s, and brought to the UK by Bill Schlackman in the 1960s (see Patterson & Malpass in 57/2, 2015), leading to the the emergence of unease about the implications of this new knowledge in exploiting consumers as exemplified by Vance Packard in the ‘Hidden Persuaders’, published as early as 1957.What we were witnessing in that period was the application of knowledge gained in other social sciences, and psychology, to marketing and the emergence of an overall strategic approach to the market which was beginning to usurp the operationally orientated functions of sales and advertising as the only ways to engage with consumers, with powerful consequences that linger on today.

The brand as a core intangible asset of a firm, and strategy to build image, market share, and thereby increase the value of the company was becoming king. Demand was replacing production as the key to success in a market.

What can be seen in the paper are the way-markers towards the development of the Dirichlet model. As Ehrenberg discusses in the paper there was a lack of empirically-based theory: ‘We do not know whether consumer attitudes change before consumer behaviour’, or any real understanding of the relative importance of what influences behaviour and how such factors might interrelate. And, as seen in many of the citations within the paper, the importance of Ehrenberg’s key associates Chatfield and Goodhardt’s input to the development of theory in this field is fully recognised.

The early model described in the paper, the NBD/LSD (Logarithmic Series Distribution) provides an empirically based analysis of repeat buying; quantities purchased; rates and pack size rates of buying; multi-brand buying and sole buyers; consumer attitudes - all at this early stage based on FMCG sector data, and media consumption.

The paper also includes an intriguing diversion into a discussion of the impact on consumers of product quality, citing a study into how the sensory perception of white fish was affected by a wide range of factors related to the fish itself, plus catch location and storage, the point being that it can be difficult to assess how product quality affects consumer preference. Ehrenberg concludes that: ‘the essential step of relating the results of product testing to consumer acceptance in the marketplace is still largely missing. The principle statistic here remains in the saying that ‘Seven new products out of ten fail’.

The author takes issue with the move towards formulating comprehensive mathematical ‘models’ before any organised empirical knowledge has been established - ‘Sonking’, the Scientification of Non-Knowledge’ – in order to try ‘to bring order out of complex datasets’! Contentiously, Ehrenberg believes that the sizable literature on model-building had yet to stand up to any basic test of scientific theories: ‘Take away the mathematical language and what generalised factual knowledge of the process in question remains? If the answer is none, the mathematical symbol for that is very simple’! The ‘I sonk, therefore I am’ approach had left, in the author’s view, no tangible results.

I would argue that this phenomenon is still apparent in some of the models described in rejected papers submitted to IJMR, especially in cases where the data feeding the model, and the research design, are below the standards we expect to see in submissions to IJMR.

Ehrenberg concludes with twelve relationships in the data that can be expressed quantitatively based on the proportion of the population buying a particular brand and the average purchases all within a given time period. These models of the ‘equilibrium position’ provide a foundation for exploring the ‘non-stationery’ trends in consumer behaviour. The key call to action is the need to establish a theory of consumer dynamics.

In recent years, the role, added-value and accountability of marketing has been questioned, with a trend towards fewer senior marketeers being board level appointments. However, the principles established in the latter half of the twentieth century established scientific principles of how to measure consumer buying behaviour and use this knowledge within marketing strategy that are as relevant today as when first theorised. The problem is that they are often ignored in the search for a new way, but if you look at Binet & Field’s ‘Marketing in the era of accountability (Warc, 2007), an analysis of the IPA dataBANK database of IPA award entries, the authors state: ‘The dataBANK findings on buyer behaviour echo those of Ehrenberg (2000). His analysis based on his many years of research, suggests that penetration has a much bigger impact than loyalty on market share, and that attempts to increase loyalty without increasing penetration are largely futile’, and the authors refute criticisms that his work is too fmcg orientated to have established a more generic rule for markets.

This is a fitting tribute to Ehrenberg’s work and his legacy, and if only more marketers would take heed of this advice when planning their strategy and setting their budget priorities!

Read the landmark paper: ‘Towards an integrated theory of consumer behaviour’, A.S.C. Ehrenberg.