Why liking matters. (includes appendix) (Special Issue: Copy Testing)

Are likable TV commercials more effective? Is liking a useful measure for evaluative pretesting? Previous research shows liking can work in more than one way to influence viewer response, and some studies suggest that liking may be a valid indicator of relative sales performance. Analysis of a major copytesting database demonstrates that liking is moderately but significantly correlated with other, validated measures of effectiveness. Used in conjunction with other appropriate measures, liking measures add substantial value to the assessment and optimization of advertising effectiveness.

It's natural enough to want to know whether consumers like advertising, and it's easy enough to find out by asking them. The more vexing question--whether likable advertising is inherently more effective than less-likable advertising--has been debated for decades within and from outside the advertising industry.

In its most extreme form, one point of view is that advertising, occurring in the context of entertaining television programming, must be likable to be effective and that positive attitudes toward an advertisement lead to favorable attitudes toward the advertiser or advertised brand. On these grounds, a substantial number of practitioners advocate the use of liking as a pretest measure to evaluate an ad's potential effectiveness.

Recently, the value of liking measures has been the subject of animated discussion. Much of the controversy stems from studies that have defined the issue in different terms, both theoretically and operationally, sometimes with seemingly conflicting results. The authors propose that useful insight can be gained by examining liking ratings in relation to other commonly accepted pretest measures, and by viewing these relationships in the light of previous research.

What follows is a summary of key theoretical positions for and against the use of liking measures, explanations of how ad liking works to influence other consumer perceptions, a review of relatively recent research on liking and advertising effectiveness, and, finally, an account of our own research on liking.

Theoretical Approaches to Commercial Liking

Extensive academic research offers substantial support for the notion that attitude toward the ad can influence attitude toward the brand. Numerous studies suggest that ad liking is one of the most important predictors of brand liking, second only to in-going or prior brand attitude (see review by Thorson, 1991; Moore and Hutchinson, 1985; Lutz, 1985).

Perhaps less widely known is theory and research on the reverse, where attitude toward the brand influences attitude toward the ad (MacInnis and Jaworski, 1989; Madden and Ajzen, 1991). Indeed, Alwitt (1993) has shown that attitude toward the ad can be influenced by the mere presence of the brand in the execution, as well as by other executional factors not necessarily related to the product or message. This suggests that attitude toward the ad and attitude toward the brand have a chicken-and-egg relationship: it may be difficult to distinguish the cause from the effect.

However, recent work by Brown and Stayman (1992) has gone a long way to clarify the processes by which attitudes toward the commercial influence attitudes toward the brand. Results from a meta-analysis across studies on attitude toward the ad were consistent with a "dual mediation" model. According to this model, attitude toward the ad has an impact through two avenues: (1) it influences attitude toward the brand which, in turn, influences brand-purchase intention; and (2) it modifies brand cognitions which also affect brand attitudes, which again influence purchase intention.

In broader terms, there are two primary hypotheses to explain how liking might contribute to advertising effectiveness. The first has to do with rational response, or cognitive processing of the advertising message. The idea is that if consumers like the advertising, they are more likely to notice and to pay attention--and more likely to assimilate and respond to the message.

The second hypothesis has to do with more emotional, or affective, response. This theory of "affect transfer" asserts that if viewers experience positive feelings toward the advertising, they will associate those feelings with the advertiser or the advertised brand.

These two processes are not mutually exclusive, and academic research (Madden, Allen, and Twible, 1988) finds evidence for both. Both of these theories seem

reasonable, and we believe many practitioners of advertising and advertising research would agree that likable executions might work in either way (or both) to influence attitudes.

In contrast, an alternate position is that getting the sales message across through a unique selling proposition (USP) is paramount. Advertising works if consumers buy the product; it doesn't matter if they like or dislike the advertising. From this position one can readily argue that very much liked or disliked executional characteristics of the commercial may detract from communication of the USP, a phenomenon that Rosser Reeves (1986) dubbed "vampire video."

A related notion, prevalent in the industry, is that well-liked ads are "too soft" to break through the competitive clutter: what they need is an element of irritation to be effective (cf. Bartos, 1980, 1981). Adherents of this view often cite examples of effective commercials that consumers "love to hate" (e.g., the Wisk "Ring Around The Collar" and Mr. Whipple "Please Don't Squeeze the Charmin" campaigns). Alwitt (1987) argued against this conclusion:

The few dramatic cases in which disliked advertising has been associated with increased sales are sometimes emphasized out of proportion to their prevalence and without credit given to media weight, distribution, and other non-executional factors which may also have contributed to sales.

Previous Research on Commercial Liking

In recent years there has been an increased interest, among both …