Headline Length as a Factor in Magazine Ad Readership

Abstract

It is generally conceded that headlines carry much of responsibility for communicating strategic message in magazine advertisements. Whether there is an optimal headline in terms of form, length or composition is still debatable subject. Perhaps one of most widely read practitioners who has addressed this question is Ogilvy. In his most recent discourse, Ogilvy On Advertising, several suggestions are put forth on writing headlines.1 Concerning length he says: Starch reports that headlines with more than IO words get less readership than short headlines. On other hand, study of retail advertisements found that headlines of 10 words sell more merchanidse than short headlines. Conclusion: if you need long headline, go ahead and write one, and if you want short headline, that's all right too. Other authors in trade and text books have varied opinions on optimal headline length. Caples is often cited as proponent of long headlines for relating important news, for direct response, or for highinvolvement products such as large ticket items where reader is likely to be seeking substantial information. Burton2 asks in text subhead, Is it better to keep headlines brief? He answers, A few years ago this question would have been called silly. answer would have been: Of course it is. Evidence seems to be piling up, however, that long headlines sell better than short ones. In his advertising text Nelson says, While it is best under normal circumstances for headline to be short, there is no reason it cannot run on for several sentences.3 Acceptance of longer headline length is generally hiestant: Most copywriters feel that head has better chance of being read if it's brief and quick .... best advice probably is to watch out for long heads.4 The headline should be as long as it has to be to get job done. Brevity is desirable, but not at expense of clarity or persuasiveness.5 This study addresses question of optimal headline length for magazine advertising using actual performance data for sample of advertisements from general consumer magazines. This was done by correlating headline length with an industry standard advertising recall procedure in form of readership scores collected by Starch INRA Hooper. Starch's description of their procedure for evaluating readership is that minimum of 100 magazine issue readers are shown advertisements from issue and asked questions about their recall of ad and its parts such as, Did you see or read any part of this advertisement? Method sample of advertisements used in this study was 136 advertisements evaluated by Starch in three issues of general interest consumer magazines. Sports Illustrated, Woman's Day and Reader's Digest. This sample produced 84 advertisements judged to be directed toward women and 52 directed toward men. This judgement was obvious for Sports Illustrated and Woman's Day, since only men and women readership scores are recorded. Reader's Digest has scores for both sexes for each advertisement, though some are obviously directed at only one audience. To make judgment for which gender's score to report, larger of two scores was used. independent variables were number of characters and number of words in each headline. In addition, variable was computed which consisted of number of words greater than three characters long. This was done to reduce impact on word totals of formalistic words, such as the a and, as opposed to significant words in headlines. dependent variables were headline readership score as reported by Starch and computed variable which attempted to mediate relations between headline score and interest in advertisement in general, as represented in Ad-as-aWhole Noted score. This follows Starch's recommendation that product interest be taken into account in using these measures. …