An appraisal of media weight tests

Abstract

This report considers the topic of media weight testing: that is,the examination of the effect of different levels of media advertising spending on behaviour in the market. In a simple test, for example, the level of expenditure is increased above normal in one area and the sales, brand shares or attitudes in that area are compared to those in other, comparable areas where normal weights of advertising are maintained. Although media weight tests are a much practised marketing exercise they would seem rarely to produce conclusive, or useful results. The MCRU has found that within the body of experiences of Sponsor companies only about one in twenty media weight tests have produced conclusive results. Similarly, discussion with other market researchers has suggested the same order of magnitude for the success ration. Also a study reported by Clancy (1972,1638) suggests the situation is not much better in the U.S.A. In order to formulate guidelines for successfully conducting media weight tests, it is necessary to establish, illustrate and understand the reasons giving rise to the usual failures. These reasons can be categorised in two ways: A - statistical, quantifiable reasons B - conceptual and managerial reasons, which tend to be less easily quantifiable After the problem has been defined, this report is set out in three parts:¬ Part A considers and establishes why the circumstances of the market place may render it unlikely that many media weight tests will produce any detectable changes in the market place; Part B considers the market and organisational context in which this sort of exercise is usually conducted and against which it has to be assessed; and Part C discuss the implications of these findings and presents guidelines for successful media weight tests.