Commercial Clutter: Effects of 15-Second Television Ads on Consumer Recall

ABSTRACT - Fifteen-second television commercials are gaining acceptance among advertisers in spite of concerns about increased "clutter" in the television advertising environment. This laboratory research examines effects on copy point recall of configurations of chains of commercials containing 15-second commercials, and product relatedness. Results show both factors affect recall for 15- and 30-second commercials.

Citation:

Scott Ward, David Reibstein, Terence A. Oliva, and Victoria Taylor (1989) ,"Commercial Clutter: Effects of 15-Second Television Ads on Consumer Recall", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 473-478.

Advances in Consumer Research Volume 16, 1989      Pages 473-478

COMMERCIAL CLUTTER: EFFECTS OF 15-SECOND TELEVISION ADS ON CONSUMER RECALL

Scott Ward, University of Pennsylvania

David Reibstein, University of Pennsylvania

Terence A. Oliva, Rutgers University

Victoria Taylor, Research Associate

[The authors would like to thank Prof. Ray Burke, University of Pennsylvania, and three anonymous reviewers for their helpful comments.]

ABSTRACT -

Fifteen-second television commercials are gaining acceptance among advertisers in spite of concerns about increased "clutter" in the television advertising environment. This laboratory research examines effects on copy point recall of configurations of chains of commercials containing 15-second commercials, and product relatedness. Results show both factors affect recall for 15- and 30-second commercials.

Advertisers are currently engaged in a debate about the effectiveness of 10- and 15-second television commercials. From the point of view of advertising economics, short commercials are attractive to the extent that they are proportionately less expensive than longer commercials. The crucial issue, however, is the relative effectiveness of short commercials. Some have argued that the main function of television commercials is to register brand name recognition, and the "short" commercials are appropriate and efficient for that task (see Bogart and Lehman, 1983); other argue that short commercials will appear in increasingly long chains of television commercials, and the result will be increasing advertising "clutter" and decreased effectiveness (Ray and Webb, 1986).

Still others argue that today's debate about the relative effectiveness of 10- and 15- second commercials merely mirrors the debate in the late 1960's when 30-second spots were touted as "less expensive but at least as effective" as traditional 60-second commercials. However, research in the area of advertising "clutter" raises the issue of the impact of short commercials, since they will appear on American television as part of increasingly long and complex chains, or "pods" of commercials. Data from large-scale surveys reinforce the fact that such concerns are justified: as the number of network and spot TV commercials increased 1 Co% between and 1981 (from 1,856 to 4,079 and 2,413 to 5,300 per week, respectively) in one major market, recall of the "last commercial seen" within the last four minutes decreased from 26% in 1965 to 17% in 1974 and 12% in 1981 (Bogart and Lehman, 1983).

Despite these concerns, the advertising industry has sponsored considerable large-scale research aimed at the specific question of short commercial effects, without, perhaps, considering the larger issue of television clutter they certainly imply (Ray and Webb, 1986). Recent industry data suggest that 15-second commercials are between 50-80% as "effective" as 30 second commercials, at least in terms of recall (Mord and Gilson, 1985; Gay, 1985; Bogart and Lehman, 1983). Moreover, data suggest that effectiveness of 15-second commercials is not influenced by the length of commercials that precede or follow them, and, conversely, that 15-second spots enhance the effectiveness of surrounding 30-second commercials (although recall and affect measures of 30-second commercials are-highest when a 15-second commercial follows, rather than precedes, a 30-second spot). However, the same data show that attitudinal responses to 15-second spots are quite negative, especially among 18-24 year-olds (Mord and Gilson, 1985). From a theoretical point of view, concepts from various psychological and information processing research streams are quite useful in helping us to understand processes underlying consumer responses to short television commercials, and to formulate hypotheses for research. The issue of short commercials' effects is what Ray (1974) first called "initial processing," to mean cognitive processes that occur during and just after exposure to information, up to and including short-term memory. Research by Webb and Ray (1979) and by Webb (1979) examines multiple cognitive responses to television advertising environments varying in "clutter"--operationally defined in terms of number of commercials in a chain, ad length, and inclusion of other materials (such as program credits). While their research included only 30-second commercials, their findings refuted a commonly-accepted notion from earlier research (Burke, 1972), well-grounded in early learning theory (Ebbinghaus, 1902), that commercials at the beginning or the end of a chain were more effective than commercials in the middle. Webb (1979) found a monotonic relationship with highest attention scores initially, declining in the middle of a chain, but no "recovery" at the end.

It would appear that viewers do not know when to increase their attention during a chain of commercials, in anticipation of the program's return, and this problem will become greater with increasing use of short commercials. Webb's finding suggest that Krugman's (1968) view of the "bleary-eyed TV viewer" is correct, in that boredom and fatigue set in during viewing, and that either a high-involvement message or repetition is necessary to arrest decay, and regain attention and subsequent recall.

One possibility presented by 15-second commercials is that two short commercials for the same product could be used in the same "pod" or chain of commercials. Literature on advertising repetition invariably focuses on longer term effects. However, psychological theories of information-processing are helpful. In particular, the notion of stimulus trace (Hebb, 1949, 1959) would suggest that 15-second commercial effectiveness might be heightened when two commercials for the same product appear in the same pod. Conceptually, the notion of stimulus trace is similar to the "halo effect" in other areas of consumer behavior research.

FIGURE 1

CONSUMER ATTENTION DURING A COMMERCIAL CHAIN

The concept of stimulus novelty (Berlyne, 1959; 1960) would suggest that 15-second commercials appearing in a chain of commercials might temporarily increase arousal and attention because they present marked variations in the audiovisual environment. Indeed, such variations and their novel effects have been noted in children's attention to television, and even credited with the success of "Sesame Street" (Lesser, 1982).

Another psychological theory that is helpful in understanding potential effects of short commercials is interference theory (McGeoch, 1932; Postman, 1961; Sawyer and Ward, 1978). The essential idea is that association between two items, A and B, is not lost because of decay of associative strength, but rather that alternative associations gain strength in the absence of continued training on A-B. For the present problem, interference theory suggests that recall should be greater for two ads for the same product in a pod, compared to when all commercials are different, and opportunities for "interference" are, therefore, maximized.

Based on previous research and on these theoretical notions, our conceptualization of the effects of 30- and 15-second commercials in a one-minutes "pod" is that attention (and subsequently, recall) decays over the minute, but that 15-second commercials will "revive" attention and recall. While the strengths of these hypothesized relationships should vary depending on the particular configuration of 15's and 30's with a pod, and on whether or not the products in the commercials are related, our general model looks like Figure 1.

From some level of attention at the onset of a chain of commercials, attention will generally decrease, but will be "revived" when novel stimuli (new commercials) appear. Barring other influences, the extent to which new commercials revive attention will generally not reach previous attention levels. We assume that these theoretically-based patterns of attention will be reflected in levels of subsequent recall.

THE STUDY

Hypotheses

H1: Recall will be greater for related ("related" means 2 commercials for the same brand) 15-second spots than for unrelated ("unrelated" means 2 commercials for different brands) 15-second spots; additionally, recall for the 30-second spot will be greater when the 15-second commercials are related. Memory traces should be stronger for commercials for the same brand than for unrelated brands. Moreover, short commercials for the same brand should result in less degradation of recall of an adjoining 30-second commercial than when the other commercials are unrelated, since opportunities for interference are less when commercials are for the same brand.

H2: Recall for 15-second spots will be greater when they are separated by a 30-second commercial compared to when they are blocked. Adjacent commercials should blur memorable features of advertising, but the relative novelty of a third 15-second spot should revive attention and result in higher recall for separated 15-second spots.

H3: Recall for 15-second spots will be greater for unrelated brands than for related brands when they are separated by a 30-second commercial. Recall of separated 15-second spots should be stronger when the brands are unrelated since there is a larger pool of information to recall when three brands are different, as opposed to two. Additionally, the first commercial does not serve as a source of memory trace for subsequent commercials when they are unrelated, and opportunities for memory interference are high. The third commercial should gain more attention because it is novel, and there can be no blurring of brand-related features when the commercials are for different brands.

H4: Recall for a 30-second commercial will be greater when 15-second commercials surround it, as opposed to when they are blocked before it. The rationale is that greater cognitive effort is required to maintain attention through a chain of 2 short commercials, followed by a longer commercial, compared to attending to only 1 short- commercial before the longer one. Consequently, attention and recall will decay by the time viewers are exposed to the 30 second commercial when it is in the last part of the chain.

H5: When a 30-second commercial follows two blocked 15-second commercials for the same brand, recall for the 30-second spot will be greater than when the blocked commercials are for unrelated brands. Again, greater cognitive effort is required in more cluttered situations: two unrelated short commercials should greatly inhibit recall for a 30-second commercial that follows them.

Methodology

A laboratory experiment was designed to test our hypotheses. Subjects were shown a specially designed videotape, containing an episode from "Tales Prom the Dark Side," and eleven 30-second commercials in three commercial break periods, roughly approximating normal viewing conditions. The test minute, containing two 15-second commercials, and one 30-second commercial, was inserted approximately eight minutes into the program. Following Webb and Ray (1979), four versions of the tape were developed to execute the experimental design. Pour food commercials were used to develop the four test minutes:

BY: 30-second commercial for Bryers' yogurt;

WW1: 15-second commercial for Weight Watchers Chocolate Treats:

WW2: 15-second commercial for Weight Watchers Bread;

H57: 15-second commercial for Heinz 57 Sauce.

The Weight Watchers commercials were specifically designed to be 15-second spots--an important feature considering Ray and Webb's (1986) admonition that previous research has typically used 15-second commercials that are edited version of 30-second ads, instead of commercials specifically designed and executed for 15-second airings. The four test minutes were configured as follows:

Cell I: Related with 30-second gap

WW115-seconds

BY30-seconds

WW215-seconds

Cell II: Related with 15's blocked

WW115-seconds

WW215-seconds

BY30-seconds

Cell III: Unrelated with 30-second gap

WW1 15-seconds

BY30-seconds

H5715-seconds

Cell IV: Unrelated with 15's blocked

WW115-seconds

H5715-seconds

BY30-seconds

We do not believe order effects other than those dealing with relatedness, blocking, or gapping (e.g., WW1 always came first) had a significant impact on result, since the experimental commercial pod was inserted 8 minutes into the 30-minute videotape, which contained eleven 30-second commercials. Moreover, S's were tested following exposure. We believed the intervening commercials and the elapsed time before testing mitigated any such effects.

Subjects were told that the study was about reactions to television, and that they would be asked questions about the tape they were to see. Following viewing, respondents were asked a series of questions, including the copy point recall items used as dependent measures in this research. Our operational measure of copy point recall was a broad one, including selling points, brand name, and executional events if they were related to selling points. For example, in an ad for yogurt, recalling "a car" in the ad would not be a copy point; recalling "grabbing the yogurt container because it tastes so good" would be since it is an executional event related to the selling point. Subjects were debriefed following the procedure, and none indicated knowing the true purpose of the experiment. Responses were recorded on cassette tape, and transcribed verbatim from the tapes to a reporting form constructed for the research. A limitation of our study is that we did not measure S's prior familiarity with the brands and commercials; moreover, we did not compute total copy points in the commercials.

The sample consisted of 78 graduate or undergraduate students recruited from classes at a large, private Eastern university. Forty-four were males and 34 were females. Respondents were randomly assigned to one of the cells. Total time per respondent averaged 60 minutes.

TABLE 1

OVERALL ANOVA RESULTS

Findings

Overall ANOVA results are presented in Table 1. Since this is exploratory research, the test of significance level was set at .10.

Data in-Table 1 show that average copy point recall for two 15-second commercials is greater than recall for the 30-second commercials, across all of the cells. The average of the means is 2.67 for the 15-second spots, versus 2.11 for the 30-second spot. However, the standard errors suggest that there is generally more variance around the means for recall of the 15-second spots. One could argue that we should average copy point recall across the two 15-second commercials (i.e., using "2" as the denominator for recall measure for the 15's, and "1" as the denominator for the 30's. Absent any theoretical reason for doing so, we compare total recall for the 15's with the 30-second commercials, since an important managerial issue is whether or not people recall more from pairs of 15-second commercials as opposed to a single 30-second commercial. In any case--and bearing our position in mind--the results go beyond previous studies that show 15-second commercials are "nearly" as effective as 30's. Our overall means suggest that they are more effective, at least in terms of this gross measure.

Data in Table 2 bear on our first hypothesis, that recall for both lengths of commercials would be greater when the 15-second spots are for related brands. The data provide modest support for our hypothesis: for 15-second spots, the overall mean is 2.80 when brands are related, versus 2.55 when they are unrelated (p < .10), but brand relatedness more strongly affects recall for the 30-second spot. An overall average of 2.41 copy points are recalled from the 30-second spot when the 15-second spots are related, compared to only 1.82 when they are unrelated (p < .05). Our rationale was that stimulus traces are strongest when brands are related and opportunities for interference are minimized, so recall for the 30-second spot would be less degraded by adjacent ads for related brands than for unrelated ones.

Table 3 shows results for our next two hypotheses, first, that recall for the 15-second spots would be greater when separated by the 30-second commercial, compared to when they are blocked. Back-to-back 15's should blur features and attributes in the commercials, but the 30-second commercial should give viewers a chance to "recover," resulting in greater discrimination between the short-ads, and greater recall. The overall means in Table 3 show that the overall data are in the expected direction, but are not significant (2.95 - 2.40, n.s.).

Our third hypothesis was that recall for 15 second spots would be greater for unrelated brands than for related brands when they are separated by a 30-second commercial. Data support our prediction (3.25 versus 2.65, p < .05). There is a larger pool of information from 3 different spots, and less potential for "blurring" of recall when spots are for unrelated brands. Theory also suggests that the novelty value of the commercials would be greatest when brands are unrelated, thereby gaining greater attention and subsequent recall.

Our final two hypotheses concern effects of 15 second commercials on recall of a 30-second commercial. Hypothesis 4 is that recall of the 30 will be greater when it is surrounded by the shorter spots, compared to when they are blocked before it. Data in Table 4 offer modest support for this prediction: overall means for recall of the 30-second are 2.25 when surrounded, versus 2.98 when the 15's are blocked in front. Our expectation was that the third position in the pod was the worst, owing to attention decay for the 60-second pod, but data only weakly support our expectation.

Our fifth hypothesis was the recall of a 30 second commercial that follows two back-to-back shorter commercials will be greater when the blocked commercials are for related brands, as opposed to unrelated ones. Mean recall for the 30-second following blocked 15's was 2.37 when brands in the 15's were related, versus 1.58 when they were not related, supporting our hypothesis (p < .05). Of all the cells, 1.58 is the lowest copy point recall of all, indicating that conditions of maximum clutter (3 commercials for different brands) are maximized, with resulting recall loss, when 2 of the ads are short, and are blocked prior to a third, 30-second commercial.

Discussion and Summary

Overall, our results are consistent with previous research that indicates that 15-second commercials perform well, at least as indicated by copy point recall (and bearing in mind the limitations of this small-scale study). However, our research is not consistent with some previous research in that we do not find that 30-second commercials are consistently "unharmed" by shorter commercials in the same pod. In fact, our results show that the configuration of the pod, and the relatedness of commercials in the pod, have significant implications for recall of both shorter and longer commercials.

TABLE 2

OVERALL MEAN COPY POINT RECALL OF :15S AND :30 BY BRAND RELATEDNESS

TABLE 3

RECALL OF 15-SECOND COMMERCIALS DEPENDING ON POD CONFIGURATION AND BRAND RELATEDNESS

TABLE 4

RECALL OF 30-SECOND COMMERCIAL DEPENDING ON POD CONFIGURATION AND BRAND RELATEDNESS OF 15-SECOND SPOTS

The phenomenon of 15-second commercials is complex, since the duration of the spot is only one variable in the equation. Equally important are length of the entire chain, the configuration of the chain (meaning the sequencing of shorter and longer commercials), and the brand relatedness.

Overall, our results are reasonably consistent with our theoretical expectations. A chain of commercials presents an array of information that is relatively broad or narrow depending on many factors, including relatedness of the brands advertised in the chain and sequencing of the commercial elements of the chain. In general, ads for related brands narrows the information array and increase recall. However, sequencing has effects, such that recall is heightened even for unrelated brands when commercials for them are separated. Perhaps the novelty of a third commercial, for a brand unrelated to others in the chain, momentarily arrests the slow decay of attention and recall. We also find that duration of a single commercial is not nearly as important as its place in a chain of commercials, and the nature of surrounding advertising.

This study is tentative and exploratory, owing to the small sample size and the laboratory environment, which may have heightened attention more than might have occurred in the "natural" viewing environment. More importantly, our study used only four commercials, and a considerably broader sampling of commercials is required for a valid picture of these complex effects. This is particularly true since previous research has found considerable variation depending on messages used (Mord and Gilson, 1985). Additionally, our design was limited in that we only tested "blocked" 15-second commercials when they preceded a single 30-second ad.

Finally, although our study did use commercials specifically designed as 15-second spots, rather than edited 30-second versions, a significant limitation is that our "chain" was only one minute long. American television viewers are typically exposed to considerably longer chains, and responses may be different in future research that uses longer chains. Hopefully, the present study presents some theoretical ideas on which to base future research, and a pattern of results that can be useful in designing additional work.

REFERENCES

Berlyne, D. (1960) Conflict, Arousal and Curiosity. New York: McGraw-Hill.

Berlyne, D. (1965) Structure and Direction in Thinking. New York: John Wiley & Sons, Inc.

Bogart, L. and C. Lehman (1983) 'The Case of the 30-Second Commercial," Journal of Advertising Research, 23, February/March: 11-19.

Burke Marketing Research (1972) "Viewer Attitudes Toward Commercial Clutter on Television and Media Buying Implications," paper presented at the 18th Advertising Research Foundation Conference, New York, NY.

Ebbinghaus, H. (1902) Grundzuge der Psychologie. Lepzig, Germany: Veit and Co.

Gay, V. (1985) "Marketers Sort Out Shorter Spots," Advertising Age, Nov. 21: 24, 26.

Hebb, D. O. (1949) The Organization of Behavior. New York: John Wiley & Sons, Inc.

Hebb, D. O. (1959) "A Neuropsychological Theory," in S. Koch (ed.) Psychology: A Study of A Science. New York: McGraw-Hill, Inc., 622-643.

Krugman, H. (1968) "Processes Underlying Exposure to Advertising," American Psychologist 23: 245-253.

Lesser, G. (1982) Lessons from Sesame Street. New York: John Wiley & Sons, Inc.

McGeoch, J. A. (1932) "Forgetting and the Law of Disuse," Psychological Review, 39, 352-312.

Mord, M. S. and E. Gilson, (1985) "Shorter Units: Risk; Responsibility; Reward," Journal of Advertising Research, 25, 4, 9-23.

Postman, L. (1961) "The Present Status of Interference Theory," in C N. Coger (ed.) Verbal Learning and Verbal Behavior. New York: McGraw-Hill, 152-179.

Ray, M. (1974) "Consumer Initial Processing: Definitions, Issues and Applications," in Consumer Information Processing, (eds.) G. David Hughes and Michael L. Ray, Chapel Hill, N.C.: University of North Carolina Press, 145-56.

Sawyer, A. and S. Ward (1978) "Carry-over Effects inAdvertising: Evidence and Hypotheses from Behavior Science," in J. Sheth (ed.) Research in Marketing, vol. II., JAI Press.

Webb, P. (1979) "Consumer Initial Processing in a Difficult Environment," Journal of Consumer Research, 6, 3, 225-237.

Webb, P. and M. Ray (1979) "Effects of TV Clutter," Journal of Advertising Research, 19, June: 7-12.

Webb, P. (1986) "Three Prescriptions for Clutter," Journal of Advertising Research, 25, 1, February/March, 69-79.

----------------------------------------