Days after studying a recent paper by Warren Brodsky, Director of Music Science Research at Ben-Gurion University, I came across by chance a paper by scientists at the University of California who were investigating the potential for suggestibility between sniffer dogs and their handlers. Brodsky’s paper investigates the ability of consumers to detect positive relationships between a brand and music specially commissioned to advertise the brand. Well, not so much the brand itself, but a set of ‘design language’ statements established to inform all aspects of the marketing – a sort of evocative briefing document.
Brodsky’s work is important in that it aims to establish stronger criteria for creating a marketing campaign in which the musical elements actually match the overall brand message. He writes in the abstract to the paper: ‘Manufacturers, marketing agencies, and researchers of consumer studies have handled music in a haphazard fashion. Music is often captive to financial resources, political agendas, or lack of know-how; choices rarely reflect criteria attributable to the brand. Linking music to a brand or product is a liability, as consumers’ impressions can be manipulated by incongruent music, causing brand image to shift.’ All sentiments Sound Strategies would agree with, and his paper is a welcome endorsement of our aim to promote better briefing and strategic planning in relation to music and audio in advertising.
But it also demonstrates the pitfalls in adopting a reductive approach to music-brand association. One of these was thrown into relief thanks to my encounter with the dog handling research. Domestic animals are known to develop hypersensitive responses to the behaviour of their human companions. The University of California researchers found that the dog handlers’ beliefs about the presence and location of a hidden drug strongly influenced the behaviour of the dogs themselves during the test searches. The handlers were led to believe that drugs were present in the test locations, but in fact there was nothing except the occasional appetising sausage. When the supposed locations were indicated to the handlers, the dogs were much more likely to identify what the handlers thought was the correct location though there were no drugs to sniff out. With the sausages, even more so! The dogs were simply responding to tiny behavioural signals from their handlers – or to the smell of a potential lunch.
So how much inherent suggestibility was there in Brodsky’s experiment to establish whether consumers can distinguish between brand-fit music and incongruent music? The experiment went a long way to try and eliminate researcher influence, but by reducing the measurable elements to a limited choice of simple, basic musical parameters, an absence of ambiguity and a vague awareness of Western musical conventions was all that was needed to guide the test subjects towards the ‘right’ responses. They just followed the sausages.
Of course suggestibility plays a large part in the action of advertising media on brand attributes. Indeed, the whole concept of a brand attribute is based on more or less credible suggestibility, with metaphor and imagery playing a large part. However, nuance, subtlety and ambiguity – the very stuff of any artform – are clearly anathema to a process which requires simplicity, directness and maximum appeal.
Brodsky chose two current American automobile models (a Cadillac and a Chevrolet) and began by creating a music brief which matched a series of existing consumer profiles for each brand with a set of musical expression parameters derived from the work of researchers Rentfrow and Gosling (2003), who have aligned certain musical styles and characteristics with personality traits – itself a rather questionable reductionist exercise. ‘
Next, 20 music clips were commissioned from an established commercial music production studio, based on the brief generated above: 9 for the Chevrolet, 8 for the Cadillac and 3 intentionally neutral in concept. The audio clips were then scrutinised by other music professionals in order to establish a small number of ‘best fit’ tracks to match the brand design language brief. There was a reasonably high degree of agreement about the resulting 5 selected tracks.
Already one can see the increasing convergence of the process. The music is obliged to match the consumer profiles attributed to each brand, in ways which make generalised assumptions about both the consumer tastes and the kinds of music which might be appropriate. The possibility of imaginative but unpredictable solutions is all but eliminated, as is the possibility of music contributions which might enrich the brand image in unexpected ways. When the five chosen tracks were tested with groups of potential customers, a significant degree of correlation with the two automobile models was obtained, but all the quirky or unpredictable outcomes had already been eliminated, so the choices available made it highly likely that subjects would choose the ‘correct’ tracks to match the brands. Such a reductive process is almost always going to provide positive results.
Then I listened to the music clips which were used in the tests, four each for the Cadillac and the Chevrolet. Outside the confines of laboratory conditions I was immediately struck by the inappropriateness of all the tracks in relation to the brands in question. Not because the musical elements presented didn’t indeed reflect the design language, but because the overall sound – the musical end-product – screamed low production values. Cheap synthesized instrument imitations, small combos, desultory drum tracks. I heard tuneless new age, 1960s cocktail bars, US sitcom title tracks, bland corporate anthems, a cheap West Indies holiday commercial. By reducing the music to a literal collection of lowest common denominator elements, all aesthetic and thus emotional functions were effectively eliminated. The music was no longer functioning as music, but merely as an audio reflection of certain generalised emotions and expressions – basic symbolic content.
Anyone actually hearing these clips in relation to the high quality images of the automobile brands would immediately write off either the brands or the music, or both. There was simply no credible tie-in between the visual representation of the cars and the aural experience. And though the music was specially composed strictly to the design language brief, the results had almost no distinctive musical qualities: there was no richness of texture, no musical subtext or distinguishing characteristics, no ‘personality’. It sounded like cheap pastiche – indeed that’s exactly what it was.
Does this invalidate the entire experiment? Probably not, since it did establish that consumers can make a theoretical connection between certain musical symbols and associated descriptive language – ‘design language’ – aimed at encapsulating the essence of a brand. But then, good commercial composers already know that. Their job is not simply to meet a brief, but to interpret it in a way which makes musical sense, and which contributes originality and uniqueness to the brand image.