Pooling and Dynamic Forgetting Effects in Multitheme Advertising: Tracking the Advertising Sales Relationship with Particle Filters

Abstract

Firms often use a pool or series of advertising themes in their campaigns. Thus, for example, a firm may employ some of its advertising to promote price-related themes or messages and other of its advertising to promote product-related themes. This study examines the interdependence that can occur between pairs of themes in a pool (i.e., ), the impact of these pooling effects on the allocation of advertising expenditures, and the factors that can affect forgetting rates (or, conversely, carry-over rates) in a multitheme advertising environment. The study measures pooling, wear out, and forgetting (carry-over) effects for a campaign that uses five different advertising themes. To obtain these measures, I extend the linear Nerlove-Arrow (NA) (1962) model to a nonlinear model of advertising theme quality and goodwill and estimate the extended model using Markov chain Monte Carlo (MCMC) and particle filtering ideas. Particle filtering belongs to a class of sequential Monte Carlo (SMC) methods designed to estimate nonlinear/nonnormal state space models. Results show that forgetting (or carry-over) rates may be time varying and a function of prior goodwill (past advertising) and other advertising variables. Results show, moreover, that pooling effects can reduce theme wear out and, in turn, significantly improve advertising efficiency.nonlinear state space model, particle filtering/smoothing, sequential Monte Carlo (SMC), sequential importance resampling (SIR), Markov chain Monte Carlo (MCMC), metropolis hastings, aggregate advertising models, pooling effects, forgetting effects

Firm Expansion, Size Spillovers, and Market Dominance in Retail Chain Dynamics

“…In our most flexible specification, W imt includes X mt , the squares of the components of X mt , the firm's own Z imt , the average of rival Z −imt values, and pairwise interactions between X mt and the Z imt , and rival Z −imt values. 25 Following the literature, we use the following simple, but flexible linear specification for y * imt that includes higher-order terms and interactions:…”

How does awareness evolve when advertising stops? The role of memory

“…Marketing literature contains several models of awareness formation (see Mahajan et al 1984;Mahajan and Muller 1986;Bass et al 2007;Bruce 2008;Naik et al 2008;Srinivasan et al 2010), of which the Nerlove-Arrow (NA) or autoregressive model is the most commonly used in theoretical and empirical analyses (see Fig. 1 for the frequency of published marketing studies using different dynamic specifications in the last 5 years in marketing journals).…”

“…Previous marketing literature extensively studied how advertising affects awareness formation (e.g., Zielske and Henry 1980;Mahajan and Muller 1986;Batra et al 1995;Naik et al 1998;Dube et al 2005;Bruce 2008;Srinivasan et al 2010). Based on extant awareness formation models, awareness declines immediately and gradually in the absence of advertising.…”

“…But if consumers remembered ads for 1 month (say), then the awareness loss today would depend on the awareness level prevailing a month ago rather than the current awareness level. This memorableness of ads, which delays forgetting, is ignored even in recent advertising models (see Batra et al 1995;Naik et al 1998;Dube et al 2005;Bruce 2008;Srinivasan et al 2010).…”

Robust Dynamic Estimation

“…to market outcomes (e.g., sales, share, etc.). Existing methods to estimate such dynamic models assume that the forecast errors between the model's predictions and the market data follow some probability distribution, which is typically multivariate normal (e.g., Naik et al 1998, Vakratsas et al 2004, Lachaab et al 2006, Ataman et al 2008, Bruce 2008, Rubel et al 2011, van Heerde et al 2013, Liu and Shankar 2015, Kolsarici and Vakratsas 2015. The data generating process can, however, differ from the assumed probability distribution.…”