Building Implementable Marketing Models

by Philippe A. Naert, Peter S.H. Leeflang

Overview

The observation that many models are built but few are used has almost become a commonplace in the management science and operations research literature. Nevertheless, the statement remains to a large extent true today, also and perhaps even more so where marketing models are concerned. This led Philippe Naert, now about four years ago, to write a concept text of a few hundred pages on the subject of how to build imple­ men table marketing models, that is, models that can and will be used. One of the readers of that early manuscript was Peter Leefiang. He made suggestions leading to a more consistent ordering of the material and pro­ posed the addition of some topics and the expansion of others to make the book more self-contained. This resulted in a co-authorship and a revised version, which was written by Peter Leefiang and consisted of a reshuffling and an expansion of the original material by about fifty per cent. Several meetings between the co-authors produced further refinements in the text and the sequence of chapters and sections, after which Philippe Naert again totally reworked the whole text. This led to a new expansion, again by fifty per cent, of the second iteration. The third iteration also required the inclusion of a great deal of new literature indicating that the field is making fast progress and that implementation has become a major concern to marketing model builders.

Product Details

ISBN-13:	9789020706741
Publisher:	Springer US
Publication date:	05/31/1978
Edition description:	Softcover reprint of the original 1st ed. 1978
Pages:	418
Product dimensions:	5.51(w) x 8.50(h) x 0.03(d)

About the Author

Philippe A. Naert (1945) is the Dean of TIAS Business School and Professor of Marketing, Tilburg University, The Netherlands. Peter S.H. Leeflang (1946) is Professor of Marketing at the Department of Economics at the University of Groningen, The Netherlands, and member of the Royal Netherlands Academy of Arts and Sciences.

Table of Contents

I: Marketing Modelsl.- 1. Introduction.- 1.1. Purpose and outline.- 1.2. The model concept.- 2. Classifying Models According to Their Degree of Explicitness.- 2.1. Implicit models.- 2.2. Verbal models.- 2.3. Formalised models.- 2.4. Numerically specified models.- 3 Benefits From Using Marketing Models.- 3.1 Are marketing problems quantifiable?.- 3.2. Benefits from marketing decision models.- 3.3 Building models to advance our knowledge of marketing.- 3.4. On the use of a marketing model: A case study.- 4. A Typology of Marketing Models.- 4.1. Intended use: Descriptive, predictive, normative models.- 4.2. Demand models: Product class sales, brand sales, and market share models.- 4.3. Behavioural detail: None, some, a substantial amount.- II: Building Marketing Models.- 5. Elements of Model Building.- 5.1. The model building process.- 5.2. Some basic model building terminology.- 5.3. Specification of behaviour equations: Some simple examples.- 5.3.1. Mathematical forms.- 5.3.1.1. Models linear in parameters and variables.- 5.3.1.2. Models linear in the parameters but not in the variables.- 5.3.1.3. Models nonlinear in the parameters but linearizable.- 5.3.1.4. Models nonlinear in the parameters and not linearizable.- 5.3.2. Modelling marketing dynamics.- 5.3.2.1. One explanatory variable.- 5.3.2.2. Several explanatory variables.- 6. Implementation Criteria with Respect to Model Structure.- 6.1. Introduction.- 6.2. Implementation criteria.- 6.2.1. Models should be simple.- 6.2.2. Models should be built in an evolutionary way.- 6.2.3. Models should be complete on important issues.- 6.2.4. Models should be adaptive.- 6.2.5. Models should be robust.- 6.3. Can non-robust models be good models?.- 6.4. Robustness related to intended use.- 6.5. Robustness related to the problem situation.- 7. Specifying Models According to Intended Use.- 7.1. Descriptive models.- 7.2. Predictive models.- 7.3. Normative models.- 7.3.1. A profit maximization model.- 7.3.2. Allocation models.- Appendix The Dorfman-Steiner theorem.- 8. Specifying Models According to Different Levels of Demand.- 8.1. Product class sales models.- 8.2. Brand sales models.- 8.3. Market share models.- 9. Specifying Models According to Amount of Behavioural Detail.- 9.1. Models with no behavioural detail.- 9.2. Models with some behavioural detail.- 9.3. Models with a substantial amount of behavioural detail.- 10 Shastic Consumer Behaviour Models.- 10.1. Brand choice and purchase incidence models.- 10.1.1. Brand choice models.- 10.1.1.1. Markov and Bernouilli models.- 10.1.1.2. Learning models.- 10.1.2. Purchase incidence models.- 10.2. Response models based on shastic models of consumer behaviour.- 10.2.1. Response models based on Markov models.- 10.2.2. Response models based on learning models.- 10.2.3. Response models based on purchase incidence models.- 10.3. A normative model based on a shastic consumer behaviour model.- 11. Parameterization.- 11.1. Organizing data.- 11.2. Estimating parameters in models with no behavioural detail.- 11.2.1. The linear model: Ordinary least squares.- 11.2.2. The linear model: Generalized least squares.- 11.2.3. Estimating parameters in simultaneous equation systems.- 11.2.4. Nonlinear estimation.- 11.3. Parameterization of models with some behavioural detail.- 11.3.1. Parameterization of models of intermediate market response.- 11.3.1.1. Descriptive brand choice models.- 11.3.1.2. Brand choice response models.- 11.3.2. Parameterization of aggregate flow models.- 11.4. Parameterization of models with a substantial amount of behavioural detail.- 11.5. Subjective estimation.- 11.5.1. A justification for subjective estimation.- 11.5.2. Obtaining subjective estimates.- 11.5.2.1. Point estimation.- 11.5.2.2. Probability assessment.- 11.5.3. Combining subjective estimates.- 11.5.3.1. The mathematical approach.- 11.5.3.2. The behavioural approach.- 11.5.4. Combining subjective and objective data.- 11.5.4.1. Formal analysis.- 11.5.4.2. Informal analysis.- 12. Validation.- 12.1. Measure reliability and validity.- 12.2. Face validity.- 12.3. Statistical validation criteria.- 12.3.1. Goodness of fit.- 12.3.2. Reliability of the estimates.- 12.3.3. Multicollinearity.- 12.3.4. Assumptions concerning the disturbance term.- 12.4. Validation criteria related to intended use.- 12.5. Validation of models with no behavioural detail.- 12.6. Validation of models with some behavioural detail.- 12.7. Validation of models with a substantial amount of behavioural detail.- 12.8. Validation of subjective estimates.- III: Building Implementable Marketing Models.- 13. Determinants of Model Implementation.- 13.1. Organization validity.- 13.1.1. Matching model and user.- 13.1.2. The model user-model builder interface.- 13.1.3. Contingencies specific to the organization.- 13.2. Elements of implementation strategy.- 13.2.1. Evolutionary model building.- 13.2.2. Model scope.- 13.2.3. Ease of use.- 14. Some Cost-Benefit Considerations in Marketing Model Building.- 14.1. The cost of building models.- 14.2. Measuring benefits.- 14.3. Some qualitative examples.- 14.4. General conclusion.- Author Index.