Econometric Modeling Our empirical setting covers multiple brands in four different categories, over time. Thus we face some critical questions about the stability and the specificity of the relationships we seek to estimate. In particular, we need to test if attitude stickiness and sales conversion are stable over time, or are idiosyncratic to certain time periods. In addition, we need to establish if different brands experience different marketing-attitude response effects, or if the effects are generic to the product category. These distinctions are not only econometrically important, they also have different strategic implications. For example, if the attitude-to-sales conversion parameters, including competitive actions, are found to be similar across brands, then no single brand can claim a competitive sales advantage from lifting an attitude metric.4 Similarly, if an