Advertising Research - Recognition
Advertising Research - Recognition
One level of testing is whether a respondent can recognize the advertisement as one that he or she has seen before. An example of recognition testing is the Bruzzone Research Company (BRC) tests by mail of television commercials.1 Questionnaires, such as the one shown in Figure 23-1, are mailed to 1000 households. The sample is drawn from the Donneley list of all households that either have a registered automobile or a listed telephone. Interest in the subject matter and a dollar bill payment usually generates a sample of 500. The recognition question is shown at the top. At the bottom is the brand association question, a critical dimension of most campaigns. On the average 60 percent recognize the commercial and 73 percent of these can correctly select the correct brand from the list of three. Test—retest correlations of .98 have been reported.
Another firm, Communicus, determines similar recognition measures of both television and radio commercials by showing respondents brief (10 seconds) edited portions of the commercial (excluding advertiser identification). Respondents are asked if they have seen or heard it before, to identify the advertiser, and to play back other identifying copy points. The percentage of those who can recognize the commercial and identify the sponsor has been decreasing. It fell from 59 percent in 1974 to 50 percent in 1980, perhaps because of increased clutter.2
In print, Starch has been measuring advertisement recognition since 1923. In the Starch survey, respondents are taken through a magazine and, for each advertisement, asked if they saw it in the issue. The noted score is the percentage who answer affirmatively. Two companion measures are seen-associated (note the name of the advertiser) and read-most (read more than 50 percent of the copy). Studies using the Starch data indicate that advertisement recognition will depend on the product class, the involvement of the segment in the product class, and on such media-option variables as the advertisement size, color, position, and copy approach. Although Starch scores are highly reliable in a test-retest sense, there is concern about their validity. The respondent can claim readership where none exists to please or impress the interviewer, or because of confusion with prior advertising for the brand. Further, this bias can be difficult to predict for a particular advertisement.
Logically, recognition is a necessary condition for effective advertising. If the advertisement cannot pass this minimal test it probably will not be effective. In one study of inquiries received by an advertiser of electronic instrumentation, those with low Starch scores were also low in inquiries received. Of course, high recognition does not guarantee effectiveness.
One level of testing is whether a respondent can recognize the advertisement as one that he or she has seen before. An example of recognition testing is the Bruzzone Research Company (BRC) tests by mail of television commercials.1 Questionnaires, such as the one shown in Figure 23-1, are mailed to 1000 households. The sample is drawn from the Donneley list of all households that either have a registered automobile or a listed telephone. Interest in the subject matter and a dollar bill payment usually generates a sample of 500. The recognition question is shown at the top. At the bottom is the brand association question, a critical dimension of most campaigns. On the average 60 percent recognize the commercial and 73 percent of these can correctly select the correct brand from the list of three. Test—retest correlations of .98 have been reported.
Another firm, Communicus, determines similar recognition measures of both television and radio commercials by showing respondents brief (10 seconds) edited portions of the commercial (excluding advertiser identification). Respondents are asked if they have seen or heard it before, to identify the advertiser, and to play back other identifying copy points. The percentage of those who can recognize the commercial and identify the sponsor has been decreasing. It fell from 59 percent in 1974 to 50 percent in 1980, perhaps because of increased clutter.2
In print, Starch has been measuring advertisement recognition since 1923. In the Starch survey, respondents are taken through a magazine and, for each advertisement, asked if they saw it in the issue. The noted score is the percentage who answer affirmatively. Two companion measures are seen-associated (note the name of the advertiser) and read-most (read more than 50 percent of the copy). Studies using the Starch data indicate that advertisement recognition will depend on the product class, the involvement of the segment in the product class, and on such media-option variables as the advertisement size, color, position, and copy approach. Although Starch scores are highly reliable in a test-retest sense, there is concern about their validity. The respondent can claim readership where none exists to please or impress the interviewer, or because of confusion with prior advertising for the brand. Further, this bias can be difficult to predict for a particular advertisement.
Logically, recognition is a necessary condition for effective advertising. If the advertisement cannot pass this minimal test it probably will not be effective. In one study of inquiries received by an advertiser of electronic instrumentation, those with low Starch scores were also low in inquiries received. Of course, high recognition does not guarantee effectiveness.
Comments
Post a Comment