The findings of a controlled field study examining the effects of background music on shopping behavior in a supermarket is examined. It is found that musical preference influenced both the amount of time and money shoppers spent in the store, although musical tempo and volume had no observable effects.

Copyright MCB UP Limited (MCB) 1996

J. Duncan Herrington: Assistant Professor of Marketing at Radford University, Virginia, USA. Louis

Introduction

Many services marketing plans focus exclusively on some combination of the traditional "marketing mix": product/service, pricing, promotional activities and distribution. Some argue that when marketing services the traditional "four Ps" do not cover all the bases adequately and that services marketers must also consider the effects of personnel, customer service, and physical facilities on customer satisfaction (Magrath, 1988). The impact of service providers' physical facilities (environments) on customer behavior and satisfaction has recently gained the attention of some services marketers (Bitner, 1992).

Within service environments customers can be exposed to numerous stimuli, all of which potentially affect how customers act, what they buy, and their satisfaction with the service experience. Of the multitude of environmental stimuli to which a shopper may be exposed in a service setting, background music has been identified as one of the most readily manipulatable and influential elements (Milliman, 1982, 1986). Research conducted over the years provides some empirical evidence to support the existence of the effect of music, especially in service environments. However, our knowledge of musical effects remains somewhat limited.

The primary purpose of this article is to report the findings of a controlled field study examining the effects of background music on shopping behavior in a traditional service environment: a supermarket. A supermarket environment was selected because most US households shop for groceries at least once a week (Progressive Grocer, 1995).

Unique to this study is a discussion of and test for the effects of musical preference on shopping behavior: a relationship yet to be explored. The research findings provide additional insight into the effects of background music on shopping behavior as well as some important considerations for the design of service environments. The reader should keep in mind that while the data used in the following analyses were collected in a retail environment (i.e. a supermarket), the results may be equally applicable to other types of service settings where customers visit or at least view the service provider's facilities (e.g. offices of health service practitioners, financial institutions, sporting arenas). This article frequently uses the term "store" to denote service environments of all types, retail or otherwise, and the term "shoppers" to denote customers within service environments.

Conceptual and empirical literature

The notion that background music can be used to influence consumer behavior is derived from the concept of atmospherics (Kotler, 1973). Given a choice, consumers prefer environments which offer a pleasant atmosphere and extend a feeling of acceptance (Martineau, 1958). In certain shopping situations the atmosphere may be more influential than the product itself in the purchase decision (Kotler, 1973) and can directly influence consumer behavior (Baker et al., 1992) and attitude/ image formation (Darden and Babin, 1994). Bitner (1992) suggested that the physical surroundings (atmosphere) of services businesses have a large impact because patrons frequently consume services within the firm's environment. Background music is generally considered to be one of several atmospheric variables that can influence evaluations, willingness to buy (Baker et al., 1992), mood, cognition, and behavior (Bruner, 1990).

A review of the literature dealing with the effects of background music on shopping behavior reveals several interesting findings. First, background music influences shoppers' evaluations of the service provider and the shopping experience. Shoppers like hearing music when they shop, and feel that stores providing background music care about their customers (Linsen, 1975). In addition, shoppers may feel that they spend less time standing in the checkout queue (Linsen, 1975) or shopping in the store (Yalch and Spangenberg, 1993).

Background music can influence certain overt in-store behavior too. Several field studies report that shopping and dining pace, and in some instances expenditures, are influenced by the volume (Smith and Curnow, 1966) and tempo (Milliman, 1982, 1986) of background music. Slower tempos and lower volumes tend to make retail patrons shop or eat at a more leisurely pace and in certain instances spend more money than faster tempos and higher volumes. Also, shopper behavior tends to differ according to the type of music played. More specifically, shoppers may spend more when the music "fits" with the product (Areni and Kim, 1993) and/or closely matches the musical tastes of shoppers (Yalch and Spangenberg, 1993).

One possible explanation for musical effects would be that individuals adjust their pace (walking, chewing, speaking), either voluntarily or involuntarily, to match the tempo of music and walk/chew/speak faster in the presence of loud music. The research conducted by Smith and Curnow (1966) and Milliman (1982) provide some measure of support for this explanation.

However, many attempts at explaining the effects of music are based on environmental psychology theory. Mehrabian and Russell (1974) propose that individuals respond emotionally to environmental stimuli (e.g. background music) which in turn leads to "approach-avoidance" behavior (Figure 1). Approach-avoidance behavior in service environments would include emotional evaluations of the environment, evaluations of the service experience, long-term patronage decisions, attitudes toward others in the environment, spending behavior, shopping time and exploration of the environment (Donovan and Rossiter, 1982).

Based on environmental psychology theory, two noncompeting explanations of the effects of music in service environments may be extended. First, Russell and Snodgrass (1991) suggest that molar environments provide a certain level of information (see Figure 2). Faster tempos and higher volumes add to the complexity of the environment, thus increasing the level of information to be processed. Increases in information are generally related to increased arousal which in turn may moderate approach-avoidance behavior (Mehrabian and Russell, 1974). The findings reported by Smith and Curnow (1966) and Milliman (1982, 1986) are consistent with the information processing explanation.

A second noncompeting explanation stems from the relationship between pleasure and approach-avoidance. Namely, that the tempo, volume, and/or some other characteristic(s) of the background music add to or detract from the pleasantness of an environment which in turn influences approach-avoidance behavior (see Figure 1). While plausible, little empirical support has been provided for this alternate explanation.

Affective appraisal of background music

There is some evidence to suggest that the tempo, volume, mode and other structural characteristics of background music influence behavior. However, music has an affective dimension as well which may also influence behavior. A musical composition's emotional context can have an impact on mood (Bruner, 1990) which in turn may mediate differing behavior within service settings (Gardner, 1985). Perhaps even more influential than general emotional tone would be the listener's level of preference for the musical composition.

The precise nature of the relationship between musical preference and relevant service setting behavior is not currently known. It can be suggested that service patrons should respond more (less) favorably to environments when they like (dislike) the background music. However, there is no empirical evidence providing direct support for this proposition.

Despite the lack of direct evidence in support of preference effects, there exists some evidence linking the level of pleasure derived from environments to approach-avoidance behaviors. Namely, people are more likely to move toward, remain in, pay attention to, and regard favorably environments evaluated as pleasant than environments regarded as unpleasant (Mehrabian and Russell, 1974; Russell and Mehrabian, 1978). These effects also hold for retail environments (Darden and Babin, 1994).

An individual's preference for a musical composition is dependent on a number of factors. Musical preference tends to vary with the structural characteristics (Kellaris, 1992) and complexity (Burke and Gridley, 1990) of the music. In addition, musical preference tends to vary according to the listener's age (Yalch and Spangenberg, 1993), musical training (Vanderark and Ely, 1993), cultural background (Wright, 1975), and familiarity with the music (Davies, 1991). Musical preference is typically expressed in terms of specific musical genre and specific musical artists (Herrington and Capella, 1994).

Hypotheses

Empirical evidence regarding the effects of the physical characteristics of music (e.g. tempo, volume) on behavior is mixed. Studies conducted in service settings provide some support for tempo and volume effects. However, the evidence is quite limited. Therefore, approximations of the hypotheses tested by Smith and Curnow (1966) and Milliman (1982) are included here for the purposes of confirmatory testing.

H1:The time shoppers spend in a service environment will be reduced by loud music (Smith and Curnow, 1966).

H2:The amount of money shoppers spend in a service environment will be unaffected by loud music (Smith and Curnow, 1966).

H3:The tempo of background music will affect the total shopping time of shoppers (Milliman, 1982).

H4:The tempo of background music will affect the amount of money spent by shoppers (Milliman, 1982).

The term "approximation" is used because the hypotheses offered here differ from the original hypotheses as follows. First, with respect to H3, Milliman (1982) operationalized shopping pace as the amount of time, in seconds, it took a shopper to move between two designated points within the supermarket. In this study, shopping time was operationalized as the total amount of time a shopper spends in the selling area of the store, excluding time spent in the checkout queue.

The second departure concerns H4. Originally, Milliman (1982) operationalized purchase amount as "daily gross sales volume." In this study purchase amount is operationalized as the amount of money spent by individual shoppers participating in the study. This was done because the use of total daily sales volume does not allow for control of certain competing explanations. In addition, the use of individual expenditures allows for more detailed examinations as well as analyses of linear relationships between musical preference and expenditures.

The review of the literature suggests that a musical composition may have at least three primary dimensions: a physical dimension represented by the physical characteristics of the music, an emotional tone which is likely to be related to the physical characteristics of the composition (Bruner, 1990), and a preferential dimension. It is hypothesized that the preferential dimension, the degree to which the shopper likes or dislikes the background music, may also affect shopping behavior. In an effort to test this proposition, the following hypotheses are offered.

H5:Preference for the background music will affect the amount of time shoppers spend in the service environment.

H6:Preference for the background music will affect the amount of money shoppers spend in the service environment.

Research method

Sample

Data were collected from the stream of shoppers entering the supermarket between the hours of 1 p.m. and 8 p.m., Monday through Thursday, over a three-week period at a single location. Complete data were obtained from 140 shoppers. A subset (n = 89) of the total sample was identified and distinguished by level of awareness of the background music.

The distinction between the total and sub-samples is important because only certain data were available from each. The amount of purchase and shopping time were obtained from all 140 individuals surveyed. Preference for the background music, because it was dependent on whether the participant was aware of the background music, was available from only 89 of the 140 participants. Each hypothesis was tested using data from the largest sample level available.

Sample procedure and description

Prior to entering the selling area of the service environment shoppers were asked to participate in a study conducted as part of a university project. Potential participants were offered a five dollar gift certificate in return for their participation. The surveyed sample represents a diverse group of 140 adults from a metropolitan area in Southeastern USA.

The sample consisted of 112 (80 percent) females and 28 (20 percent) males. The age of individuals in the sample ranged from 18 to 80, with an average of 42.54 years. Of the participants, 67 percent were married, and 68 percent shopped alone. The average size of household among respondents was 2.82 and the average level of education of shoppers was 15.65 years. Because of the exploratory nature of the study, rigorous procedures for assessing sample representativeness were not employed. However, the supermarket's general manager indicated that the sample was typical of the store's clientele.

After agreeing to participate in the study and prior to entering the service environment, each participant was instructed to complete the first questionnaire containing items designed to measure antecedent mood state, perceived time pressure and several demographic characteristics. After completing the first questionnaire, participants were asked to return to the researcher's booth on departing the checkout queue to complete another brief questionnaire. After the initial instructions were relayed to the participant the researcher unobtrusively recorded the participant's time of entry into the selling space of the service environment.

While attending to his or her shopping task, each participant was subjected to one of seven different musical factor level combinations: slow-tempo/low-volume, slow-tempo/loud-volume, fast-tempo/low-volume, fast-tempo/loud-volume, business background music/low-volume, business background music/loud-volume, or a no-music control level. The background music used in the study began playing before the beginning of data collection and continued to play until the last participant had departed.

The researcher next recorded the time at which the participant entered the checkout queue. The time lapse between entering the selling space of the service environment and entering the checkout queue, measured in whole seconds, served as the measure of shopping time and was processed as a dependent variable. Only the time the shoppers spent within the selling area was included in the measure of shopping time. After the participant exited the checkout queue, he or she was instructed to produce his or her cash register receipt from which the researcher recorded the purchase amount. At this time participants were instructed to complete the second questionnaire containing several items designed to measure musical preference.

Tempo and volume

In previous research on the effects of musical tempo on shopping time (Milliman, 1982, 1986), fast- and slow-tempo music treatments were operationalized by selecting different musical compositions which were originally recorded at either fast (90 or more beats per minute, or BPM) or slow (60-65 BPM) tempos. As a result, different musical compositions were used for each treatment group. By using different musical selections, the responses to the background music may have been confounded with the other characteristics of the music (e.g. musical preference, musical style, key, mode, harmonies and timbre).

The specific tempo levels used in this study were similar to those used in Milliman (1982, 1986): slow-tempo (60-65 or fewer BPM), fast-tempo (90 or more BPM). A business background music condition (mean = 104) was included so that the supermarket could assess the effects of its normal musical programming on shopping behavior.

In an effort to reduce the potential for confounding, this study used the same musical compositions for both the slow-tempo and fast-tempo levels, thereby controlling for differential responses based on musical style and composition. The musical compositions used were instrumental and were all reproduced in major modes, with two exceptions. In order to use the same musical compositions, each was reproduced at different tempos while holding all other physical characteristics of the music (e.g. pitch, key, harmonies, timbre) constant between the two factor-levels. This was accomplished by using digitized musical sequences which were reproduced and recorded at slow and fast tempos through the use of an electronic synthesizer. This method of recording allowed the music to be reproduced at both fast and slow tempos while holding all other characteristics of the musical compositions constant between the two tempo levels, thus reducing the potential for confounding (Bruner, 1990). The prerecorded background music - slow and fast-tempo treatment levels - was broadcast over the supermarket's public address system using an auto-reverse tape-deck. The business background music treatment level was administered by playing a business background music channel over the supermarket's public address system.

The volume levels of the background music used in this study were subject to restrictions set by the supermarket manager. The supermarket manager subjectively selected two volumes (soft and loud) which he felt would be the minimum and maximum volume levels acceptable to customers. To ensure consistency, a decibel meter was used to set sound levels prior to each data collection session.

Measurement of musical preference

The musical preference items (Table I) were designed to measure the participant's subjective evaluation of the degree to which he or she liked or preferred the background music present during the shopping experience. The musical preference scale items consisted of nine statements with seven-point Likert-type response formats (7 = extremely accurate and 1 = extremely inaccurate). Three of the items were negatively coded and reverse scored prior to analysis.

Factor analysis of the nine items measuring preference for the background music show that all nine of the items load together on a distinct factor (Table I) with one item, "I found the background music to be annoying," also loading significantly on a second factor. Given that the stray loading was both negative and substantially lower for the secondary factor, all nine items were combined to obtain a summated scale score for preference for the background music. The estimated coefficient alpha for the nine items representing preference for the background music is 0.83.

Extraneous influences

In an effort to reduce the possibility of rival explanations, three additional variables known to influence shopping time and expenditures were measured so that their variation could be eliminated statistically. These variables included size of household (Dawson et al., 1990), antecedent mood state (Gardner, 1985), and time pressure (Park et al., 1989). The size of household consisted of a single item regarding the number of individuals living in the respondent's home.

The items measuring antecedent mood state (Table I) were obtained from Watson et al. (1988). The PANAS scales consist of 20 adjectives with a five-point response format (1 = very slightly or not at all and 5 = extremely). The 20 descriptive adjectives measure two distinct dimensions of mood: positive affect and negative affect. The factor analysis of the complete set of scale items shows all ten of the positive affect items loading together on a single factor (Table I). All ten of the negative affect measures load together on a separate factor. The reliability score estimates are 0.88 and 0.90 for the positive affect and negative affect measures respectively.

The seven items measuring perceived time pressure (Table I) provided for a five-point Likert-type response (1 = very inaccurate and 5 = very accurate). The measures of perceived time pressure represent the participant's subjective comparisons of the anticipated amount of time he or she needed to complete the shopping task compared to the amount of time he or she had available due to extraneous time constraints.

The seven items measuring perceived time pressure load together on a distinct factor (Table I). One of the time pressure measures, "There are other things I need to be doing right now," loads significantly on two factors. However, this item loads higher on the time pressure factor than on the other factor (negative affect). Therefore, all seven items were used in constructing the summated scale score for perceived time pressure (reliability = 0.84).

Analysis and results

Hypotheses tests

The primary aim of the analysis was to determine the extent to which background music influences the amount of time and money shoppers spend in a service environment. Given that two related dependent variables were explored, the data were analyzed using multivariate analysis of covariance (MANCOVA). The experimental factors were the tempo and volume of the background music. No significant interaction effects were detected. Shoppers' preference for the background music was processed as a covariate along with positive affect, negative affect, time pressure and family size.

The MANCOVA for the differences in shopping time and purchase amount (Table II and III) indicates no differences among the background music factor-levels for either time or money spent (Wilks' Lambda = 0.898, F = 1.17, p = 0.31). Therefore, H2, because of its negative wording, is supported, but H1, H3 and H4 are not. The tempo and volume of the background music did not influence the shopping time or expenditures of shoppers. Of the control variables (covariates), only family size contributed significantly(t = 4.87, p < 0.01): and then only for the amount of money shoppers spent.

H5 and H6 propose that shopping time and purchase amount respectively are influenced by shoppers' preference for the background music. The MANCOVA (n = 89) for differences in shopping time and purchase amount (Tables IV and V) indicate that preference for the background music did influence behavior. The test of significance for the influence of preference for the background music on shopping time (t = 2.22, p = 0.029) provides support for H5. Similarly, the test of significance for the effects of preference on purchase amount (t = 1.76, p = 0.083) provides marginal support for H6. The signs for the regression coefficients obtained when shopping time and purchase amount are regressed on preference for the background music (263.62 and 7.68 respectively) are both positive. It would appear that musical preference can have a positive influence on the amount of time and money shoppers spend in service environments.

Managerial implications and recommendations

Before reaching any binding conclusions regarding the effects of music in service environments several limitations should be noted. First, the results, while somewhat generalizable to supermarkets, may not be as applicable to other service environments. Differences among patrons and service selection processes/situations may dictate different conclusions and recommendations. In addition, while the sample obtained for this study was somewhat representative of the store's clientele it may not be representative of all supermarket shoppers.

Second, data were collected only between the hours of 1 p.m. and 8 p.m., Monday through Thursday. A national survey of supermarket shoppers shows that just over half (56 percent) of all supermarket shopping is done in the afternoon and evenings and that 30 percent of shoppers under the age of 55 shop Monday through Thursday while 59 percent of shoppers 55 years and over visit the grocery store Monday through Thursday (Progressive Grocer, 1995). Therefore, the sample may not be representative of all shopper types: namely those who are compelled or prefer to shop Friday through Sunday.

Finally, the tempos and volumes used in the experiment were somewhat limited in range. Because of this, true variance due to tempo and volume may have gone undetected. Future research may wish to use a broader range of tempos and volumes as well as exploring the relationship between tempo and volume and musical preference.

Despite certain possible limitations, several recommendations can be offered. Overall, the results of this study indicate the source of the effects of music at the point-of-purchase to be shoppers' preference for the background music rather than tempo or volume. The tempo and volume of the background music did not significantly influence the shopping time or purchase amount of the sample of shoppers. In addition, the amount of time and money spent in the supermarket by shoppers exposed to the business music - the music normally played in the supermarket - was not distinguishably different from that of the shoppers exposed to the experimental music.

Instead, it was observed that the amount of time and money spent in the supermarket was positively related to shoppers' level of preference for the background music. More specifically, shopping time and expenditures were observed to increase with the level of preference for the background music, regardless of tempo and/or volume. These findings suggest that efforts to influence shopping behavior via tempo and/or volume, a strategy supported by the findings of previous music research, may not be completely effective.

However, service providers can potentially maximize both exploration and expenditures by making sure that the musical compositions contained in the background music match, as closely as possible, the tastes and preferences of shoppers as well as the nature of the service(s) provided.

Matching background music with shoppers' musical tastes can be accomplished in a variety of ways. First, shopper surveys can be used to determine the most favored types and styles of music as well as the specific compositions and artists shoppers enjoy most. Retailers can then use this information to develop the background music format(s) most appropriate for their stores. Preference for the various categories of music (i.e. popular contemporary, rock-and-roll, country and western, classical) tends to vary by age, income, education and ethnic background (Herrington and Capella, 1994). Second, business music services such as Muzak, AEI, DMX, Music Choice and 3M provide a wide variety of music formats, each of which is designed to appeal to a particular segment of listeners. Service providers subscribing to a business music service may be able to obtain valuable information regarding the musical preferences of their target market(s) from the provider.

Some service providers find that patron characteristics vary somewhat by time of day, day of week, and/or by department. Older patrons in the mornings and midday may be replaced in the afternoon by a much younger, school-aged crowd. The evening may give way to middle-aged clientele. The practice of dayparting and staging (playing different music at different times of day and/or in different areas of the store) requires the availability of several different musical formats. The key to selecting the most appropriate business music service is variety. The availability of a wide variety of musical formats would ensure that service providers could obtain an approximate match between their background music and the preferences of their customers.

Music may be used to help create a distinctive image and position in the market. There are numerous examples of service providers who use music to help create atmospheres that are consistent with their service offerings. Restaurants offering international menus (e.g. Mexican, Chinese, Italian, French) often play music of the country of origin. Other restaurants specializing in a particular type of food can also use music in a similar fashion. For instance, some Bar-B-Que restaurants play either rhythm and blues or country and western, depending on the desired theme, to create an ethnic atmosphere. Clothing stores specializing in European fashions often play European contemporary music to reinforce their image.

While potentially useful in helping to define and/or reinforce a particular image, music can just as easily offend patrons. For instance, while Indian raga music is useful in creating a distinctive image for a restaurant specializing in Indian cuisine, many diners of western descent may find this type of music annoying. As a result, the music may actually detract from the service experience. Therefore, service environment designers must take great care when using music to create, define, or reinforce the image of the service provider.

One alternative would be to use a mix of ethnic compositions and more popular tunes known to please patrons. Another alternative would be to search through all the various genres of music within the specific ethnic musical strain for one genre which is well liked by target customers. For instance, there are several different types of Mexican music (e.g. contemporary, mariachi, ranchera/tejano as well as versions of jazz, rock and roll, and folk) each of which appeal to a different segment of the Mexican population. A Mexican restaurant wishing to attract a younger clientele may wish to play Mexican rock and roll, a mixture of US and Mexican rock and roll, or some other genre suitable for younger listeners.

Service providers must also be aware of the potential effects of background music on service personnel. Numerous studies have been conducted on the effects of music on employee performance. In summary, these findings suggest that music can enhance or detract from performance and mood. More specifically, workers tend to perform better or at least feel better about their jobs when they like the music being played. One exception is when the task being performed is rather complex. In this situation music of any type may detract from performance.

Music that appeals to patrons may or may not appeal to employees. For instance, soft contemporary and country and western songs appeal to many older customers. However, younger service personnel may find the music boring and/or unpleasant, which may impact on service delivery. Whenever possible, service providers should try to select music acceptable to both patrons and personnel.

Finally, familiarity seems to be a primary determinant of preference (Davies, 1991). However, after a certain number of repetitions, patrons may tire of a song and become annoyed on hearing it. Therefore, when selecting musical compositions service providers should use music that is familiar yet still fresh to patrons.

Implications for future research

That the findings of this study seem to be at odds with previous research reporting volume (Smith and Curnow, 1966) and tempo effects (Milliman, 1982, 1986) may be troubling to some readers. However, the combined results should be viewed as complementary in that each study provides some evidence to support the phenomenon of musical effects.

The only real difference between the findings of the study reported in this article and the findings of previous studies lies in the hypothesized source(s) of the effects of music. This analysis simply provides some evidence to suggest that musical preference, a variable not previously examined, may offer a more valid explanation for the effects of music than the tempo and/or volume of background music. In fact, it can be suggested that previously observed effects may have been the result of preference rather than tempo or volume.

To illustrate, consider the experimental procedures of previous supermarket studies. In order to manipulate the tempo of background music, Milliman (1982, 1986) used two completely different sets of musical compositions: one for the slow- and a different one for the fast-tempo treatment level. The results of this study indicate that shoppers evaluate certain musical compositions more favorably than others. Consequently, the results obtained in previous research may have been more a function of the differences in shoppers' musical tastes and/or affective evaluations of the musical compositions than the structural characteristics (i.e. tempo) of the background music. It should be noted, however, that an individual's preference for a musical selection may be based in part on the tempo of the music (Kellaris, 1992).

It remains tenable that the tempo and/or volume of background music could influence shopping behavior when delivered at certain levels. The three-tempo factor-levels (slow, fast, and business music) used in this study ranged from an average of 60 to an average of 104 BPM: a more extensive range than used in previous research. Perhaps true tempo effects exist beyond this tempo range.

Given the vast number of factors influencing shopping behavior, it is unlikely that a service provider's choice of background music will, in isolation, determine ultimate success or failure. However, the results of this study and those preceding it provide enough evidence to suggest that background music can influence shopping behavior. Regrettably, not enough is currently known about the effects of music and the musical preferences of consumers to offer anything beyond general advice. Additional research is much needed.

References

1. Areni, C.S. and Kim, D. 1993, "The influence of background music on shopping behavior: classical versus top-forty music in a wine store", in McAlister, L. and Rothschild, M. (Eds), Advances in Consumer Research, Vol. 20, pp. 336-40.

2. Baker, J., Levy, M. and Grewal, D. 1992, "An experimental approach to making retail store environment decisions", Journal of Retailing, Vol. 68, pp. 445-60.

3. Bitner, M.J. 1992, "Servicescapes: the impact of physical surroundings on customers and employees", Journal of Marketing, Vol. 56, April, pp. 57-71.

4. Bruner, G.C. 1990, "Music, mood, and marketing", Journal of Marketing, Vol. 54, October, pp. 94-104.

5. Burke, M.J. and Gridley, M.C. 1990, "Musical preferences as a function of stimulus complexity and listeners' sophistication", Perceptual and Motor Skills, Vol. 71, pp. 687-90.

6. Darden, W.R. and Babin, B.J. 1994, "Exploring the concept of affective quality: expanding the concept of retail personality", Journal of Business Research, Vol. 29, pp. 101-9.

7. Davies, J. 1991, "The musical mind", New Scientist, Vol. 129, pp. 38-41.

8. Dawson, S., Bloch, P.H. and Ridgeway, N.N. 1990, "Shopping motives, emotional states, and retail outcomes", Journal of Retailing, Vol. 66 No. 4, pp. 408-27.

9. Donovan, R.J. and Rossiter, J.R. 1982, "Store atmosphere: an environmental psychology approach", Journal of Retailing, Vol. 58, Spring, pp. 34-57.

10. Gardner, M.P. 1985, "Mood states and consumer behavior: a critical review", Journal of Consumer Research, Vol. 12, December, pp. 281-97.

11. Gorn, G.J. 1982, "The effects of music in advertising on consumer choice: a classical conditioning approach", Journal of Marketing, Vol. 46, pp. 94-101.

12. Herrington, J.D. and Capella, L.M. 1994, "Practical applications of music in service settings", Journal of Services Marketing, Vol. 8 No. 3, pp. 50-65.

13. Kellaris, J.J. 1992, "Consumer esthetics outside the lab: preliminary report on a musical field study", Advances in Consumer Research, Vol. 19, pp. 730-4.

14. Kotler, P. 1973, "Atmospherics as a marketing tool", Journal of Retailing, Vol. 49, Winter, pp. 48-64.

15. Linsen, M.A. 1975, "Like our music today Ms. shopper?" Progressive Grocer, October, p. 156.

16. Magrath, A.J. 1988, "When marketing services, 4 Ps are not enough", Business Horizons, May-June, pp. 44-50.

17. Martineau, P. 1958, "The personality of the retail store", Harvard Business Review, Vol. 36, pp. 47-55.

18. Mehrabian, A. and Russell, J.A. 1974, An Approach to Environmental Psychology, MIT Press, Cambridge, MA.

19. Milliman, R.E. 1982, "Using background music to affect the behavior of supermarket shoppers", Journal of Marketing, Vol. 46, Summer, pp. 86-91.

20. Milliman, R.E. 1986, "The influence of background music on the behavior of restaurant patrons", Journal of Consumer Research, Vol. 13, pp. 286-9.

21. Park, C., Whan, I., Easwar, S. and Smith, D.C. 1989, "The effects of situational factors on in-store grocery shopping behavior: the role of store environment and time available for shopping", Journal of Consumer Research, Vol. 15, March, pp. 422-33.

22. Progressive Grocer (1995), "62nd annual report of the grocery industry - 1994", rogressive Grocer, Mid-April.

23. Russell, J.A. and Mehrabian, A. 1978, "Approach-avoidance and affiliation as functions of the emotion-eliciting quality of an environment", Environment and Behavior, Vol. 10, pp. 355-87.

24. Russell, J.A. and Snodgrass, J. 1991, "Emotion and the environment", in Stokols, D. and Altman, I. (Eds), Handbook of Environmental Psychology, Vol. 1, Kreiger Publishing, Malabar, FL.

25. Smith, P.C. and Curnow, R. 1966, "'Arousal hypothesis' and the effects of music on purchasing behavior", Journal of Applied Psychology, Vol. 50, pp. 255-6.

26. Vanderark, S.D. and Ely, D. 1993, "Cortisal, biochemical, and galvanic skin responses to musical stimuli of different preference value by college students in biology and music", Perceptual and Motor Skills, Vol. 77, pp. 227-34.

27. Watson, D., Clark, L. and Tellegen, A. 1988, "Development and validation of brief measures of positive and negative affect: the PANAS scales", Journal of Personality and Social Psychology, Vol. 54 No. 6, pp. 1063-70.

28. Wright, D.F. 1975, "Musical meaning and its social determinants", Sociology, Vol. 9, pp. 419-35.

29. Yalch, R. and Spangenberg, E. 1990, "Effects of store music on shopper behavior", Journal of Services Marketing, Vol. 1, pp. 31-9.

30. Yalch, R. and Spangenberg, E. 1993, "Using store music for retail zoning: a field experiment", in McAlister, L. and Rothschild, M. (Eds), Advances in Consumer Research, Vol. 20, pp. 632-6.

Executive summary and implications for managers and executives

(Provided by Marketing Consultants for MCB University Press)

Rocking them in the aisles: music in supermarkets

If, as the Bard asserted, music hath charms to soothe the savage breast, what effect will it have on shoppers? Piped music is now a part of the ambience in just about every supermarket, restaurant, shopping mall and department store. We listen to edited highlights of great classical tunes (there are times when I wish the Queen of Sheba would arrive) while waiting on the telephone and the gentle sounds of synthesized pop tunes wash over us in car parks and lifts. But does it really help? Do customers respond positively to these carefully selected melodies?

Herrington and Capella present us with part of the answer - music does make a difference in supermarkets and is seen as an established part of a supermarket experience. Moreover, they find that it is music we like that gets us going rather than music per se. Doubtless, if supermarkets played atonal classics by Schonberg or Stockhausen shoppers would leg it to some other store to be soothed by Mantovani or synthesized versions of show tunes.

So how do we know what to play? Herrington and Capella point out that restaurants often play appropriately ethnic music (jangling sitars or bangra in Asian eating places, Neapolitan love songs in pizza restaurants and Zorba's dance in Greek bistros) but, given the vast range of musical taste in the population at large, supermarkets face the challenge of selecting music that not only does not offend but makes a positive contribution to the store's ambience. The result, to the chagrin of those with sophisticated musical tastes, is what amounts to the lowest common denominator. Only the most familiar tunes are played, presenting a choice of uncompromising blandness and inoffensiveness.

For some supermarkets, keen to develop what they put out over their sound systems, ideas such as an in-store radio employing professional DJs have been introduced. The result is a replica (at least in the UK) of local commercial stations with a play list dominated by "golden oldies" interspersed with advertisements for special offers and promotions (although thankfully lacking in adverts for home improvements). However, the personal touch implicit in such an approach helps to increase the attractiveness of the music since a DJ can react to public requests and develop a rapport with the listener in a way that loop tapes of piped music cannot.

I have long believed that, while piped music adds unquestionably to the supermarket experience despite the blandness of the repertoire, stores should seek to create events in the same way as pubs and restaurants. Rather than a tape of Christmas tunes, a Salvation Army band playing Christmas carols adds extra to the shopping experience. Stores could consider live music evenings - perhaps with a jazz band one evening, a string quartet another and a barbershop choir on a third. This way music becomes a positive attraction and the store can develop links with local orchestras, schools, churches and brass bands developing its position in and contribution to a community.

As a trustee of a music charity I have often felt that if people will not come to concerts you have to take the music to the people. Links with supermarkets and department stores represent a chance to extend people's musical experience and provide the retailer with the chance to present a positive promotional message. In this way shoppers receive a more varied experience - you could even introduce a silent afternoon for those irritated by music while they shop!

Nevertheless, it remains the case that traditional (if that is the word) background music will still be played during most of the opening hours even were a store to widen the musical experience given to shoppers. This makes it important that musical tastes are considered and Herrington and Capella suggest such measures as shopper surveys to supplement the services provided by background music producers. These could guide selection and reflect varied preferences, something especially important where there are ethnic or regional differences in musical tastes. A supermarket with a large West Indian population in its catchment perhaps needs a different mix than one with a large Irish or Asian population. Similarly, the age profile of the population will guide selections - research by advertising agencies has demonstrated that there is a critical age period when popular music is most likely to hit the nostalgia button. We respond best to those popular tunes that were hits when we were 18-23 years old (even though we often hated them at the time). Similarly, the distinct shopping patterns of different age and social groups could be reflected in music variation across the day. If older people tend to shop in the morning then we need to reflect that fact in the music played. Similar consideration to when younger single people and thirtysomething couples shop could help define the most effective background music.

Retailers in general, and supermarkets in particular, have not thought about music as part of the store ambience beyond observing that it is a good thing for sales and a contributor to the overall shopping experience. I hope that Herrington and Capella's article makes them think about the use of piped music. Music in shops may be a good thing, but its effect could be made still better with a little more thought and application.