The data in Table 4.1 (from Tauber 1981) support this view. They reflect consumers’ intentions to buy expressed during concept testing and their perceptions and behavior gauged 25 weeks after the beginning of adver­ tising for the new brand. Awareness is clearly related to some extent to purchase intention and both probably reflect interest in the product class. O f those who became aware, the prior expression of definite purchase intention was associated with trial. But significant amounts of trial also came from the other respondents who were less sure they would buy or convinced they would not. Moreover, among triers, the expression at concept stage of a purchase intention, positive or negative, weak or strong or noncommittal, appears not to affect the probability of repeat purchase. Tauber is adamant in indicting tests which rely on the assumption that

measures of attitude and intention can predict behavioral choices:

Concept tests and product tests do not work. Take a look at their history. Most marketers are aware of the high failure rate of new consumer products. Yet the A. C. Nielsen Company, in a study of the introduction of many new products, concluded that for products that survived the first eight months, test markets were a good prediction (correct 75 per cent) of national introductory results. So when do all of these failures happen? Obviously, they occur at the test market stage. Thus the track record of pre-test-market screening is poor. And what are the techniques generally used for this screening? - Concept testing and product testing.