A comparative study of market share models using disaggregate data

Prior research assessing the predictive validity of alternate market share models produced conflicting results and often found that econometric models performed worse than naive extrapolations. However, contributors to IJF's recent issue on market share models suggested that such models are often misspecified, in part because they exclude promotional variables and are estimated on aggregate data. Thus, we used weekly scanner data to assess full, reduced, and naive forms of linear, multiplicative, and attraction specifications across different levels of parameterization. Consistent with specification-based arguments, (1) econometric models were superior to naive models, (2) GLS estimates of attraction models were superior when models were fully specified, (3) OLS estimates of linear models were superior when models omitted important variables, and (4) attraction models predicted best overall. Moreover, in general, unconstrained models yielded superior forecasts relative to constrained models because brand-specific parameters were heterogeneous for the product category tested.

[1]  P. Naert,et al.  On the Prediction Power of Market Share Attraction Models , 1981 .

[2]  Vithala R. Rao,et al.  Alternative Econometric Models of Sales-Advertising Relationships , 1972 .

[3]  John D. C. Little,et al.  Feature Article - Aggregate Advertising Models: The State of the Art , 1979, Oper. Res..

[4]  Marcel Weverbergh,et al.  Market Share Specification, Estimation, and Validation: Toward Reconciling Seemingly Divergent Views , 1985 .

[5]  J. Armstrong,et al.  Forecasting Methods for Marketing: Review of Empirical Research , 1987 .

[6]  Frank M. Bass,et al.  Pooling Issues and Methods in Regression Analysis with Examples in Marketing Research , 1975 .

[7]  Robert E. Lucas,et al.  Estimation and Inference for Linear Models in Which Subsets of the Dependent Variable are Constrained , 1968 .

[8]  Lee G. Cooper,et al.  Voting for a Political Candidate Under Conditions of Minimal Information , 1974 .

[9]  A. Buse,et al.  Elements of econometrics , 1972 .

[10]  R. Brodie,et al.  Attraction versus Linear and Multiplicative Market Share Models: An Empirical Evaluation , 1984 .

[11]  Peter S. H. Leeflang,et al.  On the Predictive Power of Market Share Attraction Models , 1984 .

[12]  D. Aaker,et al.  The sophistication of naive modeling , 1987 .

[13]  Roderick J. Brodie,et al.  Reply to the commentary , 1987 .

[14]  Frank M. Bass,et al.  Misspecification and the inherent randomness of the model are at the heart of the Brodie and de Kluyver enigma , 1987 .

[15]  Scott A. Neslin,et al.  A Comparison of Market Share Models and Estimation Procedures , 1984 .

[16]  Henri Theil,et al.  Economic Forecasts and Policy , 1962 .

[17]  G. Erickson.  Marketing managers need more than forecasting accuracy , 1987 .

[18]  Cornelis A. de Kluyver,et al.  A comparison of the short term forecasting accuracy of econometric and naive extrapolation models of market share , 1987 .

[19]  Dick R. Wittink,et al.  Causal market share models in marketing: Neither forecasting nor understanding? , 1987 .

[20]  Vincent P. Norris.  The Economic Effects of Advertising: A Review of the Literature , 1984 .

[21]  H. Theil.  Principles of econometrics , 1971 .

[22]  R. Parks,et al.  Efficient Estimation of a System of Regression Equations when Disturbances are Both Serially and Contemporaneously Correlated , 1967 .

[23]  A. Zellner.  An Efficient Method of Estimating Seemingly Unrelated Regressions and Tests for Aggregation Bias , 1962 .

[24]  K. Boyer,et al.  Informative and Goodwill Advertising , 1974 .