In Praise of the 15-second Advertisement

  • REPORT 48
  • Kate Newstead and Jenni Romaniuk

Abstract

15-second advertisements have traditionally been the “poor cousins” of their 30-second counterparts, used begrudgingly by marketers when they do not have the budget for a more expensive 30-second spot, or after launching a campaign with a 30-second spot. We think this disdain is misplaced.

We find that recall and likeability scores for 15-second advertisements are approximately 80% that of 30-second advertisements. Further, 15- and 30- second advertisements were equal on correct brand identification levels, because of the better branding execution of 15-second advertisements. Given that 15-second spots have a lower per-unit media cost than 30-second spots, our research findings indicate that using 15-second advertisements (and therefore increasing media efficiency) does not necessarily result in a major loss of advertising effectiveness. However, the relative strength of the 15-second ad might be an artefact of the poor quality of 30-second ads in branding execution.

INTRODUCTION

The length of an advertisement is a key decision marketers must make when commissioning television advertising. Ad length constrains what is possible in terms of building a story, conveying a message or making branding links. While the 30-second spot has been a standard for many decades, there is a trend towards alternative, particularly shorter, advertisements. 15-second (and 10- second in the UK) are the most popular alternatives. Often advertisers launch a campaign with 30-second spots, and switch to shorter 15-second spots later in the campaign to prolong it. However, it is now also common to forego standard 30-seconds spots all together and use 15-second advertisements for the entire duration of the campaign. 15-second spots are relatively cheaper than 30-second spots. This means advertisers who use them can stretch out the media budget and extend the life of the campaign.

To what extent does this increase in efficiency result in a decrease in effectiveness?

Much of our knowledge of the relative effectiveness of 15- and 30-second advertisements is based on research that took place in the 1980s and early 1990s, when 15-second advertisements were still relatively novel. The ever-increasing cost of buying media on television is making shorter advertising lengths more attractive. This is making their implementation increasingly common. Therefore it is timely to examine the relationship between advertising effectiveness and advertising length.

EFFECTIVENESS METRICS

We now review existing research that has examined the relative performance of 15- versus 30-second advertisements. We draw on advertising recall, liking and correct brand identification. These are three metrics commonly used by practitioners to assess the performance of advertising, particularly once it has gone to air.

Execution Cut Through/Recall

In the competitive viewing environment, an advertisement needs to cut-through the clutter in order to be noticed by the viewer and processed in memory. A 15-second advertisement gives the audience half as much time as a 30-second advertisement to notice and process it. Furthermore, with a 15-second ad, creatives have half as much time to build a story, repeat information or make branding links. This leads us to suggest that a 30-second advertisement should be more memorable than a 15-second advertisement.

Most studies report that 15-second advertisements are approximately 80% as effective in achieving recall as 30- second commercials (du Plessis 1995; Mord and Gilson 1985; Patzer 1991; Singh and Cole 1993; Stanton and Burke 1998). However, many prior studies were conducted in forced exposure conditions. Given the increasing amount of competitive clutter, in real-life viewing situations 15-second advertisements might not be able to maintain such a high relative recall. So we ask: What is the level of recall of 15-second advertisements relative to 30-second advertisements amongst respondents exposed under natural viewing conditions?

Advertising Likeability

Advertisement liking is commonly used to evaluate advertising. Liking, or having a positive emotional response or feeling towards an advertisement, can enhance the memorability of the actual advertisement and its content. This is because people give more attention to and are less likely to avoid advertisements they like (Biel 1990; du Plessis 1998).

At the 1991 ARF copy testing validity summit, advertisement likeability was shown to be the single best intermediary variable at distinguishing between campaigns that were successful and unsuccessful in terms of sales results (Haley and Baldinger 1991). This stimulated its popularity as an advertising effectiveness metric, something validated by a number of other researchers.

There has been limited direct testing of the relative likeability of 15-versus 30-second advertisements. Mord and Gilson (1985) reported that consumers find 15-second ads less interesting, less believable, less warm, less informative, more irritating, and more confusing than longer commercials. Stanton and Burke (1998) speculated that 15-seconds might not be long enough to develop a ‘likeable’ creative execution. This leads us to ask: What is the relative likeability of 15-second advertisements compared to 30-second advertisements?

Correct Brand Identification

Correct brand identification measures the proportion of those respondents who recalled having seen the test advertisement that could correctly identify the brand advertised. Given the huge cost that advertising creative and media scheduling incurs, once an advertisement reaches a customer and is noticed, it is imperative that the brand be correctly identified.

One factor that can influence correct brand identification is the branding execution within the advertisement. Stanton and Burke (1998) found that including at least three brand name mentions was equally important for brand recall in both 15- and 30-second advertisements. However, 30-second advertisements were more likely to include three or more mentions than 15-second advertisements. Because of their length restriction, 15-second advertisements may struggle to incorporate an adequate amount of branding compared to their 30-second counterparts.

However, some elements of branding quality, such as early branding, are not a function of advertisement length. Early brand presence is important to set the scene for the advertisement and means that the viewer may still notice the brand if the advertisement is turned off partway through. This means that advertisement length may not be a key factor in the level of correct branding. This leads us to ask: How do 15-second advertisements compare to 30-second advertisements for correct brand identification?

RESEARCH METHOD

During a weekend, respondents were interviewed about the advertisements that they had seen over the preceding week in their own homes. The first part of the questionnaire established which of the test advertisements the respondent had a prior opportunity-to-see (OTS). This ensured that all respondents who were asked about an advertisement had at least one confirmed OTS for that advertisement.

Respondents were then questioned about execution-prompted advertising recall, advertisement liking and correct brand identification for relevant advertisements.

  • Advertising recall – Interviewers read out a description of each advertisement that the respondent had an OTS and asked if they remembered seeing the advertisement. e.g. “Have you seen the TV advertisement with a pink background and the packet has ‘love potions’ on it? The voiceover talks about how it is double chocolate and raspberry and that it is love at first sight.” Recall of the test advertisements ranged from 8% to 95%.
  • Advertising likeability – Those who remembered the advertisement were asked to assess the advertisement’s likeability using a 5-point scale as used in the ARF Copy Research Validity Project (Haley & Baldinger, 1991). The mean likeability is calculated from only those respondents who remembered the advertisement. This ranged from 2.2 to 4.3.
  • Correct brand identification – Again only those who remembered the advertisement were asked to recall which brand was being advertised. Mean correct brand identification ranged from 0% to 93%.

For each advertisement that they could recall, each respondent was asked to approximate how many times they had seen the advertisement overall. This is to test for any confounding effects associated with either 15- or 30-second advertisements being played more frequently. We found no meaningful differences.

THE ADVERTISEMENT SAMPLE

The objective of the study was to include as many currently on-air advertisements as possible. We excluded advertisements if:

  • The length of the advertisement was neither 15- nor 30-seconds.
  • The advertisement was shown in both 15- and 30-second versions, in which the advertisement creative was the same. We would be unable to know which length was recalled.
  • The advertisement was part of a long-running campaign; due to high levels of accumulated viewing frequency for these campaigns.
  • The advertisement was for TV programs or TV networks, which tend to be played at high frequency, and lend themselves to primacy/recency effects since they are almost always the first or last shown in an advertisement pod.
  • The advertisement was for government and/or charity organizations, since such campaigns have different objectives.
  • Dual-branded advertisements, as such advertisements present difficulties in terms of comparing their correct branding score to single-branded advertisements.
  • The advertisement was for a highly targeted product that would not have broad category usage (e.g., Diapers/ Nappies).

The final set of advertisements came from a wide range of product and service categories, including beer, packaged food, financial services and automobiles.

The final dataset has a total of 143 ads, of which 108 were 30-seconds and 35 were 15-seconds. The different number of 30- and 15-second advertisements reflects their natural incidence.

RESULTS

In this study, we compare the mean performance of 30-second advertisements to the mean performance of 15-second advertisements for each of the three-effectiveness metrics. As seen in Table 1, 15-second advertisements perform at 80-90% for execution recall and likeability. Not a bad result given they are half the length. However, of surprise to us were the equal correct branding levels.

Table 1: Mean Execution Recall, Likeability and Correct Branding for 15- and 30-second Advertisements
Execution
Recall		Like-abilityCorrect Branding
Score
15 sec ads393.242
30 sec ads483.540
Relative Performance#81%91%105%

Why similar correct branding scores?

To explore this similarity in correct branding results further, we compared 15- and 30-second advertisements across some branding execution dimensions: Brand Frequency; Brand Entry Timing; and Last third only branding.

We found that:

  • 15-second advertisements tended to have the brand present more frequently; once per 4 seconds, versus once per 8 seconds for 30-second advertisements.
  • More 15-second advertisements introduced the brand earlier; shown after on average 3 sec versus 11 sec for 30-second advertisements.
  • More 30-second advertisements had last third only branding; 29% versus 6% of 15-second advertisements.

This suggests that 15-second advertisements are matching the performance of 30-second advertisements, despite their shorter length, because they have stronger branding execution.

Table 2: Branding Execution in UK, US and Australian Advertisements
LengthnFrequency
(secs per mention)		Entry timing
(secs elapsed)		Last third
branding
%
10 secUK104310
15 secUS200439
AUS35456
20 secUK307821
30 secUS20081127
AUS10881129
UK50101330
Average6819

To examine if this finding was also evident in other countries we examined the branding execution in the US and UK as well. We found that 15-second advertisements consistently had more frequent, earlier branding. Also 30-second advertisements were 3 to 4 times more likely than 15-second advertisements to only reveal the brand in the last third of the advertisement *1. In the UK, where 10- and 20-second advertisements tend to be more common, 10-second advertisements resembled the 15-second advertisements from other countries, while 20-second advertisements sat somewhere in the middle.

*1. i.e. brand present in only the final 5 or 10 seconds of the advertisement for 15- and 30- second advertisements respectively.

DISCUSSION

We find that 15-second advertisements are approximately 80% as effective as 30-second advertisements in terms of recall and likeability. Therefore use of 15-second advertisements does not result in a major compromise in effectiveness.

However, for correct brand identification we find that 15-second advertisements are equally as effective as 30-second advertisements. Further branding execution analysis suggests that 15-second advertisements have earlier and more frequent branding, relative to 30-second advertisements.

The latter are more likely to rely on branding in the last third of the advertisement only.

At the same time 15-second advertisements do achieve slightly lower likeability scores than 30-second advertisements. Therefore although advertisers might be incorporating branding elements in 15-second advertisements at the expense of developing a story/theme; this sacrifice is not resulting in extreme reductions in advertisement liking. Furthermore, while 30-second advertisements have double the amount of time of 15-second advertisements to include branding execution, they do not necessarily capitalise on this opportunity. This begs the question, is our result because 15-second advertisements work well, or is it because 30-second advertisements are often so poorly branded?

For marketing practitioners, the findings of this study have important implications. Firstly, in terms of recall and/or likeability, a 15-second advertisement is a fair buy at 80% the price of a 30-second advertisement; and at any price lower than this it is good value. But in terms of correct branding, 15-second advertisements can be even better value than first thought.

Our findings also have implications for the practice of starting with 30-second advertisements and then cutting these executions down to 15-seconds to prolong the campaign. Our results show that the performance of 30-second advertisements was not substantially better than 15-seconds. Therefore is the extra time upfront of any benefit?

Our future research is looking to extend this analysis to buying behaviour, drawing on Single Source data; as well as to television program promotions where we will have the opportunity to look at a wide range of varying lengths.

If you would like to talk to us about assessing the quality of the branding execution within your television advertisements, email Jenni@MarketingScience.info

REFERENCE LIST

Biel, Alexander L. (1990), “Love the Ad. Buy the Product?,” Admap (September), 21-25.

du Plessis, Erik (1995), “An advertising burst is just a lot of drips: a perspective on the reach, frequency, and response curve arguments,” Admap, July, 10.

—- (1998), “Memory and Likeability: Keys to Understanding Ad Effects,” Admap ( July/ August), 42-46.

Haley, Russell I. and Allan L. Baldinger (1991), “The ARF Copy Research Validity Project,” Journal of Advertising Research, 31 (2), 11-32.

Mord, Marvin S. and Edith Gilson (1985), “Shorter Units: Risk-Responsibility-Reward,” Journal of Advertising Research, 25 (4), 9-19.

Patzer, G.I. (1991), “Multiple dimensions of performance for thirty second and fifteen second commercials,” Journal of Advertising Research, 31, 18 – 25.

Singh, Surendra N. and Catherine A. Cole (1993), “The Effects of Length, Content and Repetition on Television Commercial Effectiveness,” Journal of Marketing Research, 30 (February), 91-104.

Stanton, John L and Jeffrey Burke (1998), “Comparative Effectiveness of Executional Elements in TV Advertising 15 versus 30 second Commercials,” Journal of Advertising Research, 38 (6), 7-13.

RELATED CATEGORIES