Network Past Issues

Frank Bass, a highly respected US-based academic who worked in the area of marketing science, formulated the Bass model that predicts the diffusion of new products or technologies by onetime buyers. Bass made fundamental contributions to the discipline of marketing, in particular to marketing science, including introducing models derived from econometrics changing, thereby, the pre-existing pedagogy of marketing courses in academia and the manner in which it is practised in the industry. He won numerous awards and is one of the most cited researchers in marketing in academic journals.

The Bass Model

While teaching Industrial Administration at Purdue University’s Graduate School of Management in 1961 Frank Bass published his seminal paper on modeling diffusion of new products in 1969. Subsequently, the model enunciated in the paper became a landmark in marketing and is now referred to as the Bass diffusion model.

The early 1960s was an exciting time for researchers as far as new product diffusion was concerned. The era saw the emergence of a few mathematical models. In 1962 Professor Everett Rogers’ seminal book Diffusion of Innovations was published, which was descriptive and devoid of mathematical formulations. However, the book did lay down the foundations of the diffusion of innovation theory.

It was in this backdrop that Bass presented an article at the Winter Conference of the American Marketing Association on the basics of the Bass Model without offering any empirical evidence to validate it in 1963. In his 14-page article the model occupied less space than two pages. While developing the model, in line with that of Rogers, he classified adopters into two categories (against five by Rogers): innovators – who buy a new product irrespective of whether others buy them or not and imitators – people who buy a product depending on whether others have bought it or not.

There is an interesting anecdote regarding how Bass arrived at the model. This was in 1963 when Bass was a teacher at the Krannert School of Management at Purdue University. He was reading about the impact of word-of-mouth on sales of new products in Rogers’ newly-released book when a student approached him to express Rogers’ idea mathematically. It was in response that Bass thought, “The probability of adopting by those who have not yet adopted is a linear function of those who had previously adopted.”

While it is almost certain that Rogers’ book influenced him in pursuing this line of thought it is also possible that the epidemiological models he had read about gave an impetus to the formulation of his new model.

The main assumption of the model was that the product was a singlebuy product, such as a consumer durable.

Subsequently, Bass wrote a working paper justifying the model empirically. The mathematics was substantially the same as the 1963 paper but this paper carried empirical details of 11 consumer durables with data spanning a decade or two supporting the model. It also considered coloured TV sets for which data spanning three years was available. On this basis sales were forecast for five years. Since the forecast differed to an extent from industry expectations and forecasts it ruffled feathers in the industry and Bass received angry telephone calls and letters from the industry and Wall Street. Bass then subjected his model to the test of falsifiability. His paper, carrying two pages of handwritten computations, also used data for two products that were not consumer durables- a weed spray and a new drug.

This was followed by the seminal article by Bass in 1969, which resembled the 1967 working paper very closely. The paper remained with the publishers for about 22 months prior to publication. Bass recounts an interesting anecdote in this context. The title of the paper had a typographic error. The published article was titled “A New Product Growth for Model Consumer Durables” instead of “A New Product Growth Model for Consumer Durables”.

The paper forecast sales of 5.8 million in 1967 and a peak of 6.7 million in 1968. The model had gone against the industry and Wall Street wisdom and cleared the falsifiability test.

The 1969 paper was nominated by INFORMS Institute for Operations Research and the Management Sciences as one of the ‘ten most influential papers’ published in the first 50-year history of Management Science, its flagship journal. The article is among the most frequently cited in the history of marketing literature. Google Scholar (2017) showed 7302 citations on February 15, 2017.

Parsimony of the Model

The Psychology Wiki (2017) defines parsimony as the “least complicated explanation for an observation”. It further explains that mathematical models with the least possible number of parameters are preferred as each additional parameter introduces an element of uncertainty to the model. Parsimony is considered to be a theoretical virtue. The physicist Newton is reported to have said, “Nature is pleased with simplicity and affects not the pomp of superfluous causes”.

The Bass model is a parsimonious one that uses only three parameters to forecast diffusion. In the basic model it does not use marketing-mix variables like price and promotion. The Generalized Bass Model formulated later does use marketingmix variables but gives adequate reasons about why the Bass model is generally successful without the inclusion of variables. Parsimony is considered to be one of the strengths of this model.

Inadequate quality data in emerging markets is an established fact. It was meant to access new product diffusion data in India, which is nonexistent in many cases. That is why this model’s requiring only sales’ data over a period of time can be seen as one of its major merits.

Subsequent developments

The model has gone on to spawn a huge body of literature over the past years or so attracting researchers from all parts of the world including India. Bass himself was part of a substantial portion of such developments including the paper he co-authored with Krishnan and Jain in 1994 (the Generalized Bass model) explaining why the Bass model works well without marketing-mix variables. The model has, since, been extended to include technology generations, repeat purchases, marketing-mix variables, diffusion across international borders, and the influence of socio-economic factors on diffusion.

Bass nurtured r e s e a r c h e r s including doctoral candidates. He was thesis chair to 58 students who, in turn, guided the work of other students many of whom are leading researchers in the field. Bass himself was supervisor to such researchers as Dan Horsky, Dipak Jain, and John Norton. The healthy tradition that Professor Bass started lives on after his death through his doctoral tree.

Tributes

Bass was the recipient of numerous awards including the Paul D Converse Award (1986), the American Marketing Association/ Richard D Irwin/McGraw-Hill Distinguished Educator Award (1990), the Gilbert A Churchill, Junior Award (2002), and the Charles Coolidge Parlin Award. He served as Editor of the Journal of Marketing Research during 1972-75. He was also president of the Institute of Marketing Science during 1978-79.

Over the past 45 years or so his model has guided new product developers in various sectors of the economy... The Ehrenberg-Bass Institute for Marketing Science, has written on its webpage:

“Professor Frank Bass pioneered the establishment of marketing as a science in which well-tested ma t h e m a t i c a l models could be used to predict the behavior of future markets.”

Trichy Krishnan, a leading researcher in the field and a former student of Professor Bass fondly recalls him saying, “Dr. Bass always believed that everything in the world should be explainable in simple words. To him, simplicity and clarity were two critical qualities of truth. His thoughts remain highly infectious.

I always remember his Texan accent when he would say “beautiful” on seeing a good marketing model. What a beautiful mind his was!”

By: Suddhachit Mitra FPRM Participant

.