Threshold effect of advertising has received wide attention in the marketing literature, but little formal representation of this phenomenon has been developed. On the basis of monthly time-series and cross-sectional data, the author tests for the existence of a threshold effect in a consumer good market. Alternative models are developed and estimated. A switching regression model which assumes a concave function to a cutoff point seems consistent with the evidence. The cutoff point below which advertising has little effect is estimated and the shift of advertising elasticity is assessed. The results show that advertising elasticity increases above the cutoff level, price elasticity is larger for heavily advertised premium brands than for less advertised low-price brands, and a minimum threshold level on advertising share is consistent with the theory.
BEMMAOR, A.C. (1984). Testing Alternative Econometric Models on the Existence of Advertising Threshold Effect. Journal of Marketing Research, 21(3), pp. 298-308.