In a previous article about the relationship between film editing speed and advertising performance, MacLachlan and Logan (1993) analyzed the relationship between commercial shot length and commercial day-after-recall and persuasion. They noted a substantial increase in the cutting speed or number of shots per commercial from the late seventies to the early nineties so that by the time of their analysis there was an average of 13.2 camera shots in a typical commercial. They also pointed out that the average length of the shot in commercials was much shorter than the shot length of the TV programs in which they were embedded. In an analysis of 641 commercials, they found a strong negative or inverse correlation between the number of shots or cutting speed of the film and day-after recall scores. They found a similar negative relationship between fast edit speed and persuasion scores, measured as pre/post shift in brand preference. In addition, this negative relationship was found to be relatively independent of the age of adult respondents. Their conclusion was that “advertisers are loading their commercials with too many camera shots, and persuasion and recall are suffering as a result.” Their advice was to reduce the number of shots in commercials.
Faster Than Ever
Slowing down the cutting speed of television commercials to reduce their visual complexity seems to be a clear and unambiguous implication of MacLachlan’s research. But despite these findings over a decade ago, advertisers continue to produce fast-cut commercials which, like the rest of the world, seem to be moving faster than ever.
Are we to assume then that agency creatives and their clients are indifferent or immune to the lessons of good research? Or is it that their creative intuition is telling them that research was somehow missing an essential part of the story? To answer this question, we decided to investigate the issue of film speed and advertising performance further.
In reaching their conclusion that advertising was moving too fast these researchers were drawing on the work of the film theorist Raymond Spottiswoode (1967) who proposed a theory of affective cutting tone which suggested the optimal time at which to make a cut. The idea was to cut at the “peak” of the content curve, which is the point at which the audience has just assimilated all the information in the shot. Cutting after the peak of the content curve produces boredom and a sense of dragging time; cutting before the peak produces frustration and a failure to convey the necessary information to follow the storyline. In terms of advertising film, too slow a cutting rate in TV commercials would result in boredom and a probable fall-off in attention; while too fast a cutting rate would produce confusion and undermine learning, with a resulting fall-off in persuasion and recall.
Fast-cut editing of a commercial is a way of “speeding” through information. A useful metaphor for visualizing this is to think of your television set as the windshield of your car as you look out at the road ahead. If you push down on the gas pedal and start to speed up, the scenery in your windshield changes very fast. Let up on the gas pedal and the scenery changes very slowly. One of the questions that we occasionally get asked by advertisers, presumably because of the previously published research, is “What is the ‘speed limit’ for editing film?” Of course, the real answer is that it depends on what you’re trying to do.
If you are trying to communicate a single, pure idea or feeling you can fix it on your horizon and with a tunnel-vision focus of attention speed toward it as fast as you like. A montage commercial is an example of this approach. If you are trying to communicate multiple ideas or sales messages, then you must slow down, so that you can look around and take in this idea or that one from the passing countryside. The speed limit of a commercial is set by the complexity of the strategic concept you are trying to communicate.
To measure the rate of information flowing through a commercial you could, as before, simply count the number of shots in the ad. However, camera shots can last a relatively long time so that as the action unfolds, the visual information present in the beginning of the shot might be perceptibly different from that in the middle or at the end of the shot. For that reason, instead of shot length, we took as our measure of information speed or visual complexity the number of pictures used in an Ameritest picture sorting deck.
Counting shots is similar in concept to counting the number of pictures in a picture sort deck. However, because changes in visual content within a shot are also counted, picture sorts represent a fine-tuning of the previous approach for analyzing the effects of commercial speed. Using this tool that is more sensitive, we replicated MacLachland’s advertising analysis, but this time using the performance metrics of two other major pre-testing systems, Millward Brown and Ameritest.
From a commercial sample taken from the Ameritest copy testing database, we examined 369 ads from a wide variety of categories, from high tech to financial services. We also analyzed 140 packaged goods commercials tested by Unilever in the Millward Brown system (which has a license to use Picture Sorts®). Each of the two systems provides an Attention Score as a measure of breakthrough power, and a Branding Score as a measure of branding, and a Motivation or Persuasion score based in post-viewing purchase intent. Neither system, however, provides a measure of Recall.
Each system measures attention and branding differently, however. Ameritest measures attention within a clutter reel format where a test commercial has to win the fight for attention against four other ads. Millward Brown derives its measure of attention from a composite of two rating statements on enjoyment and memorabilty, Ameritest measures how well a commercial is branded with a top-of-mind awareness of the brand name after the clutter reel exposure, whereas Millward Brown uses a five-point rating of the commercial’s fit with the brand.. Despite these differences in how the two systems operationalize the theoretical constructs of attention and branding, previously published research has shown that the two systems generally produce similar outcomes with regard to the kinds of advertising executions their scoring systems appear to reward.
With our new analysis of commercial speed and advertising effectiveness, we reach a somewhat different conclusion from that of the MacLachlan research. We find a significant positive—not negative—correlation between speed or visual complexity as measured by the picture counting method, and the attention scores produced by either pre-testing system, as shown in Exhibit 1. Initially, the correlation appears stronger for Ameritest, but once we control for differences in product categories later on, you will see that the strength of the relationship is similar for the two systems.
On the other hand, there is a significant negative correlation between visual complexity and both measures of branding. (Again, some of the difference between the two is due to differences in the categories represented in the sample.)
Recall is a combined effect of attention and branding. Therefore, taken together, these two findings do not contradict the earlier research but rather provide some insight into the reasons for the negative relationship between the number of shots and recall. Our conclusion is that as commercials move faster, or become more visually complex, additional care must be taken by advertisers to ensure that their ads are well-branded.
If we control for category differences and look at 120 packaged goods commercials tested in the Ameritest system and compare them to 120 commercials for similar product categories tested in the Millward Brown system, you can see that the correlations are now quite similar, as shown in Table 1. And now that major category and brand development differences are removed, you can also look at the relationship between commercial speed and motivation or persuasion.
Column 1 in the table describes the visual complexity—the speed of the ad—from the “objective” or outsider perspective of the researcher choosing the number of pictures to use in the sorting deck. Replicating the findings of the earlier research, we do find a statistically significant negative correlation between commercial speed and the motivation or persuasion scores of both testing systems.
If you look further down the first column, you can see a set of rating statements that are commonly used to explain the report card performance metrics of these two systems. The relationship between the picture counts in the sorting decks and these diagnostics provides an insight into why the above correlations occur.
Complexity Promotes Invlovement
As you can see, there is a strong positive correlation between visual complexity, that’s the number of “picture-bits” in the ad, and how involving, interesting, and unique commercials are rated; similarly, there is a negative correlation between visual complexity and ratings of boring and ordinary. This fits with the positive correlation you see with the attention scores. The human eye is delighted by unusual forms, colors and movement—kaleidoscopes engage our attention.
On the other hand, there is a negative correlation between visually complexity and how important the message is perceived to be or how relatable the situation shown in the ad is. Though interestingly there is no significant correlation with confusion ratings (though we frequently see problems with the flow of audience attention where consumers are not aware that they are “confused” about the real intended message of a commercial). But these diagnostics help explain the negative correlation with motivation or persuasion scores. Fast-talking salesmen are less likely to persuade.
Now we can begin to understand why advertisers and their agencies persist in developing visually complex advertising. In an age of increasing media clutter, breaking through all that noise, getting your foot through the door of the mind is of paramount importance. The first, though not the only, job of advertising is to get noticed. And here we see that viewers will reward with their attention ads that are visually complex, involving, interesting or unique and ignore ads that are too simple or too slow if they are boring and ordinary.
The creative trick, of course, is to strike the right balance between getting attention and being well-branded and motivating.
Of course, the visual complexity of a piece of film can be defined in more than one way. Simply counting the number of picture-bits of visual information in the film misses the role of rhythm and tempo and timing that come into play in organizing the audience’s experience of a piece of film. All these things affect how audiences process advertising images in order to integrate them into brand concepts.
We have defined commercial “speed” from an objective point of view of the information flowing through the ad—which is an “outsider” perspective. What happens if we look at the data from the “insider” point of view, which is after the audience has processed that information—in other words, how the audience watched the ad.
The simplest measure of processing is the percentage of images the audience actually remembers seeing in the ad—a binary sort of remember/don’t remember. This measure of processing is captured in the flow of attention average shown in the second column of Table 1.
As you can see, the average level of recall in the flow of attention doesn’t tell us very much. While there is a modest correlation to branding, there are no significant correlations to any of the other advertising metrics. The positive correlation between branding and the average level of processing suggests that better branded commercials are those where all of the information in the ad, including the brand identifiers, is well integrated into memory so that recall is higher overall
Let’s look at the key parameter that is used to describe the shape of the attention curve. Typically, flow of attention curves look like a sine-wave, displaying a rhythmic structure, with attention rising and falling and rising from one moment in the ad to the next. This suggests that as the audience is watching the images in the film, the search engine of the eye is organizing the information from the ad into meaningful chunks or units of information.
The images that stand-out above their neighbors are the focus of audience attention. They are in the foreground, front and center, of what the audience is looking at—or rather, searching for—in the film, while the other images around them are in the background of audience attention. We call these privileged moments “peaks.”
In a recent ESOMAR paper reviewing the current state of the art of cognitive neuroscience for advertising researchers, Page and Raymond (2006) describe a similar concept, the “attentional blink”, which they also attribute to the processes of selective perception.
“It is as if the brain system ‘catches’ what it wants from the image stream then blinks a metaphorical ‘eye’ (selective attention), failing to ‘see’ anything during the metaphorical blink…research in Raymond’s lab has shown that the information recalled from video ads depends directly on the duration and complexity of successive scenes. By lengthening some scenes or shortening others by as little as two-tenths of a second, recall of critical brand content in an ad can be drastically improved.”
Flow of Attention™ peaks are cognitive blinks. The limited bandwidth of human information processing capacity is reflected in the time it takes for a viewer to organize the information from an ad. This is a significant factor shaping audience responses to speedy, visually complex commercials. Page and Raymond state:
“Attentional ‘blinks’ have huge implications for creative development. These can most easily be grasped for ‘video’ advertisements, where the workspace has to create and retain multiple representations very quickly, to enable us to understand the ad. Care needs to be taken to ensure that the workspace can create a representation for the idea or concept, or crucially, brand, before the next element of creative comes along—as it will either shove the first one out of the workspace before it can be fully assembled and used, or the process of creating the first representation will stop the next one even being registered.”
The importance of rhythmic structure in communications has also been studied by social scientists who have commented on the importance of rhythm and beat in language to help us organize information in a conversation to avoid cognitive overload by focusing on predictable, strategic moments to pay attention. In spoken language, for instance, the words you stress are the important ones, and you automatically speed up or slow down on the words in between so that major points of stress form a regular chain of beats. In face-to-face conversation, the beat is when speakers often convey key information or bring up new topics.
The film writer Robert McKee describes the “beat” as the “fundamental unit of film structure”—though his definition is given from an actor’s perspective and pertains to the verbal and non-verbal give and take between actors on the screen. Our definition is orthogonal to McKee’s, perpendicular to the screen. From a visual storytelling perspective, the peak moments in a flow of attention curve can be thought of as the beat of the co-creative dance that takes place between the director and the audience—the director can lead audience attention by his rhythm and pacing of the visual information in the film, but the audience must follow.
The Flow of Attention graph is a tool for visualizing the beat of film. The fit with the theoretical ideas described above and the empirical shape of our actual flow of attention curves is quite evident. Peak moments stand out in the arc of film processing. Operation-ally, a “peak” should be understood as a relative term, not a statement about the absolute level of recall of an image. A peak is defined locally, as an image that is higher than the other images in its “neighborhood,” compared to the images before and after.
Peak moments are those moments in the ad where assembly of the brand idea takes place, before the audience’s “got-it!” blink. An average commercial contains between four and five peaks. But some effective commercials might contain only one peak, as in the climactic moment of a reveal-type ad, or even none, as in montage. The third column in Table 1 shows the relationship between the number of peak moments of attention and advertising performance.
As you can see, a higher frequency of peaks moments in a commercial are associated with more involving, interesting, and unique executions and hence higher attention scores in either pre-testing system. But importantly there is no statistically significant trade-off in terms of motivating a consumer with an important message. Apparently, when we look at information complexity from the standpoint of how it is organized in audience perceptions—i.e. with narratively structured peak experiences or moments of focused attention—the trade-off between the attention-getting power of an ad and it’s persuasiveness is less of an issue.
Thus, looking at the problem of fast paced editing or information “speed” from either perspective—an outsider’s perspective of the objective content of the film or an insider’s perspective of how much visual information the audience actually focuses on, we draw similar conclusions.
This picture sorts analysis confirms what advertising creatives already know intuitively. The legitimate objective of much contemporary advertising is not to “teach”—which it what day-after-recall measures—but rather is to pack meaningful, memorable experiences into as short a space as possible. Otherwise, in this fast moving world, consumers will not reward advertisers who waste their time with their valuable attention.
MacLachlan, J. and Logan, M., “Camera Shot Length in TV Commercials and Their Memorability and Persuasiveness,” Journal of Advertising Research, March/April 1993
McKee, R., Story: Substance, Structure, Style and the Principles of Screenwriting, New York: Harper Collins, 1997
Raymond, J.E., “Perceptual Links and Attentional Blinks,” The Limits of Attention: temporal constraints in human information processing. (Ed.) K. L. Shapiro, Oxford UP, Oxford, 2003
Spottiswoode, R., A Grammar of the Film, Berkeley: University of California Press, 1967
Young, C. “Visual Connectedness and Persuasion,” Journal of Advertising Research March/April 1992