Title: Performance evaluation of Adstock models using market drivers in the consumer packaged goods (CPG) industry
Authors: K. Antony Arokia Durai Raj; Balasubramanian Kanagasabapathi; Suyashi Shrivastava; Kunal Krishnan; Mitul Shah
Addresses: Center for Knowledge Driven Information Systems, Infosys Labs, Infosys Limited, Plot No. 44, Electronics City, Hosur Road, Bangalore – 560 100, India. ' Center for Knowledge Driven Information Systems, Infosys Labs, Infosys Limited, Plot No. 44, Electronics City, Hosur Road, Bangalore – 560 100, India. ' Infosys Limited, Plot No. 44, Electronics City, Hosur Road, Bangalore – 560 100, India. ' Infosys Limited, Plot No. 44, Electronics City, Hosur Road, Bangalore – 560 100, India. ' Retail, CPG & Logistics Practice, Infosys Consulting, Bangalore, 560 100, India
Abstract: The impact that advertising has over consumer awareness and in turn on sales volume is called as Adstock. It means stocking the advertising activity by integrating prior advertising expenditures into a stock variable and considering 'carry-over effect' over the years. Adstock is not an aggressive plan to augment sales. Rather, it is a planned activity to keep the sales steady over a given period, increase awareness and strengthen brand equity. Planned scheduling and execution of budget spending in achieving this advertising goal dictates the growth or decay in sales of a product. This research focuses on measuring the Adstock using television gross rating points (TV-GRP) and non-TV media variables by using different methods available to model Adstock. In this study, we investigated the models available in the literature describing the 'decay-effect' over a given period on live data from the consumer packaged goods (CPG) company. In addition, we have presented modified forms of the S-shaped curve model and the exponential model. The proposed exponential model with both TV-GRP and non-TV media as independent variables has shown superior performance over the other models. All the results have been computed using the Statistical Analysis System (SAS) Enterprise Guide 4.1.
Keywords: adstock models; consumer purchase behaviour; advertising carry-over; GRPs; gross rating points; forecasting; consumer awareness; television; TV-GRP; half-life; market mix modelling; performance evaluation; CPGs; consumer packaged goods; sales volumes; advertising activity; prior expenditure; advertising expenditure; stock variables; brand equity; budget spending; advertising goals; product sales; sales growth; sales decay; non-TV media; media variables; decay-effect; live data; exponential models; S-curve model; independent variables; SAS; Statistical Analysis System; electronic marketing; e-marketing; electronic retailing; e-retailing.
International Journal of Electronic Marketing and Retailing, 2012 Vol.5 No.2, pp.173 - 186
Accepted: 31 Aug 2012
Published online: 14 Dec 2012 *