<%@ Language=VBScript %> <% CheckState() CheckSub() %> Understanding core brand equity: guidelines for in-depth elicitation of brand associations
JMRS

International Journal
of Market Research

15 Northburgh Street, London, EC1V 0JR, UK
Tel: +44-20-7490 4911, Fax: +44-20-7490 0608

   

Understanding Core Brand Equity: Guidelines For In-Depth Elicitation of Brand Associations

Magne Supphellen
Norwegian School of Economics and Business Administration

‘There is no way to understand core brand equity without in–depth responses from qualitative research’.

John Pawle, Unilever (1999, p. 24)

INTRODUCTION

Relevant and accurate research on brand associations is imperative to brand managers (e.g. Kapferer 1997; Keller 1998). In fact, one of the major characteristics of the new era of branding is a sharpened focus on customer brand related experience and the effects of such experiences on consumer memories. This interest logically follows from definitions of branding. Branding is typically defined as the enterprise of creating added value in the minds of consumers, that is, building perceived values beyond the observable physical value of the product, and thus differentiating the product (Aaker 1991, 1996; Macrae 1996; Keller 1998; De Chernatony & McDonald 1998; Kapferer 1997). Correspondingly, the major purpose of branding is to create customer–based brand equity, defined as the presence of strong, unique and favourable brand associations causing differential effects on customer responses to the marketing of the brand (Keller 1993, 1998).

The consequences of superficial knowledge of brand associations can be serious. When managers fail to grasp the full breadth and depth of the associations people have for their brands, their understanding of customer brand perceptions and the way brands are positioned relative to competitors in the mind of customers will be biased. Communication strategies will be less effective and managers may fail to discover a negative development in the positioning of the brand. Furthermore, important opportunities for brand extensions and brand alliances are missed because managers are not aware of favourable associations that are relevant to brands in other product categories. National brands threatened by private label brands are particularly in need of in depth insight into brand associations because their major competitive advantage is exactly the superior perceptions that they have built over the years (Quelch & Harding 1996; Kapferer 1997). Thus, national brand managers need a very clear and deep understanding of the nature of their superior associations in order to effectively tailor the marketing mix and stay ahead of private label brands.

Gaining in depth insight into brand associations is, however, a very difficult and challenging objective. A large proportion of consumer brand perception is acquired under low involvement conditions (see Heath 1999) and is therefore not subjected to conscious processing by the right hemisphere of the brain. Consequently, many associations are unconsciously stored in a non–verbal mode (Zaltman 1997). How can such non–verbal and pre–conscious associations be elicited?

Some market researchers with a sociological orientation seem to have taken the position that, due to 'the limits of asking', we should divert our attention from respondent verbalisations and instead observe them in relevant situations and then infer what they are thinking from behavioural patterns, gestures and facial expressions (e.g. Mariampolski 1999). This view on qualitative consumer research is not tenable for two reasons. First, such a view fails to recognise the different roles of psychological and sociological methods in market research. Whereas ethnographic methods are certainly useful in exploring brand–related behaviours and contexts, the consumer himself should be acknowledged as the primary source of information about consumer memory. Though there are limits of question asking, there are more limits and a greater potential for misinterpretation in letting researchers make inferences about the thoughts and emotions of consumers. Second, research within psychology and market research points to important means of overcoming the limits of question asking. Therefore, instead of disregarding the voice of consumers in the elicitation of brand associations, we should focus on how to ask better questions, or rather on how to help consumers express their brand associations.

The purpose of this paper is to develop a set of practical guidelines for in depth elicitation of brand associations. In order to understand the problems involved, the characteristics of brand associations are first briefly reviewed from the perspective of cognitive psychology. Major challenges are then defined and seventeen practical guidelines for selection of techniques are developed in order to meet those challenges.

BACKGROUND: THE NATURE OF BRAND ASSOCIATIONS

An exhaustive review of the nature of brand associations is beyond the scope of this paper. The focus here is on aspects of memory associations that are particularly relevant to qualitative elicitation of brand associations.

Extant theories of long term memory apply a network perspective (Anderson 1983; Fiske & Taylor 1995). Associations are viewed as nodes linked together in large memory networks. Of particular importance here are theories of how memory associations, and thus brand associations, are represented in such networks, and the nature and variety of representation modes.

We usually think of brand associations as verbal descriptions of the brand. However, psychological theories tell us that this is only one of several representation modes (e.g. Anderson 1983). In fact, there are good reasons to believe that a majority of associations are not verbal, but visual with no corresponding verbal descriptions (Zaltman 1997). This is because two thirds of all stimuli that reach the brain are visual (Kosslyn et al. 1990). While awake, we constantly make visual observations of the environment. Only a small fraction of these impressions – the ones deliberately attended to and subjected to cognitive elaboration – will have a verbal description attached to them when entering memory. The recognition that memory is largely visual has important implications for elicitation (to be discussed below).

In addition to verbal and visual representations, brand associations are probably also stored in terms of sensory impressions, that is, cognitive blueprints of physiological experiences of taste, smell, sound, etc. (see Marks 1996). For example, when closing our eyes and concentrating we are able to sense the taste of a cup of tea, the smell of a rose, or the melody of a favourite sound track. These impressions are not verbal descriptions but 'raw' blueprints of real physiological experiences. Some important sensory associations will probably be verbalised when individuals think or speak about the brand, but the great majority of associations will remain non–verbal. Thus, we may expect that a considerable proportion of associations for a given brand are non–verbal sensory experiences stemming from interactions with the product. In fact, sensory associations have been seen to increase in importance as more brands than ever focus on the sensory experiences of pleasure, excitement and fun (see Holbrook & Hirschman 1982; Schmitt 1999).

Brand associations can also be represented in memory as emotional impressions (see Fiske & Taylor 1995). Brand emotions are neural, non–verbal appraisals of brand–related experiences. For example, the taste of Coca Cola or the observation of a young handsome man drinking it will evoke certain emotional reactions which are stored in memory, together with taste associations and associations of brand users. Extant theories of memory state that emotion and reason are not to be considered opposites (see de Sousa 1987; Isen 1993). On the contrary, the two mingle in human decision–making (Kahnemann 1994). In low involvement conditions, where individuals do not spend a lot of time and effort considering alternatives, emotional brand associations are often the dominating determinants of choice.

So far, we have briefly reviewed four modes of representation of brand associations: verbal, visual, sensory and emotional. An important related characteristic of brand associations is that most of them are unconscious (Plutchik 1993). This contention is consonant with the recognition that only a minor proportion of the impressions that reach the brain is subjected to deliberate reasoning. Large numbers of visual, sensory and emotional impressions are not consciously attended to, but are still stored in the associative network together with verbal associations.

Another important characteristic of memory associations, with important measurement implications, is that associations (regardless of representation mode) tend to be stored in terms of metaphors (Zaltman 1997). A metaphor is a description of a thing in terms of a different thing (see Lakoff & Johnson 1980 for a broader discussion of metaphors). For example, when a consumer has tried a new Porsche that he considers buying, he may describe the car as 'a wildcat'. Such a metaphor says a lot about his experience because we all have rich associations to cats and wild animals; they are fast and powerful, they roam, they are elegant, and they look impressive. Usage of metaphors is an effective way of communicating because metaphors capitalise on existing knowledge and thus demands less cognitive capacity. Likewise, impressions of brands are largely represented in memory in terms of metaphors because this is an effective way to understand and store impressions about brands.

The final aspect of memory associations attended to here is that, within the larger associative network, they tend to be lumped together in clusters (Anderson 1983; Barsalou 1983, 1985). Some associations are more strongly linked to each other because they have been involved in the same cognitive processes, e.g. evaluating the brand in a particular situation, defending the brand to a colleague, or making a choice between the focal brand and other brands. Some associations will be part of many clusters, but some will be found only in particular clusters corresponding to specific brand–related situations. In a similar vein, Tulving’s perspective on memory implies that important associations referring to the context of brand experiences (e.g. places of consumption) are stored in separate episodic memories (Tulving 1972, 1983). One important implication of these aspects of memory networks is that some associations could be missed if relevant situations are not discovered. Thus, when eliciting brand associations we also need to elicit situations in which individuals have had experiences with the brand.

BASIC PROBLEMS OF ELICITATION

The brief review of characteristics of brand associations points out two main measurement problems:

  1. the problem of access, and

  2. the problem of verbalisation.
    Additionally, the context of elicitation interviews in general highlights a third problem:

  3. the problem of censoring.

The following discussion of these points leads to three general principles for elicitation of brand associations. Finally, 17 practical guidelines for the selection and use of techniques are derived from these principles.

Access

A majority of brand associations are unconscious. Hence, if only the easily retrieved associations are elicited, managers are inclined to get a biased picture of their brands. A major challenge in the elicitation of brand associations is therefore to obtain access to the less conscious associations. One objection against the qualitative elicitation of brand associations that I often encounter among managers buying such research is that the results typically confirm what they already know and seldom lead to new insights. This could reflect a lack of ability of the researchers to surpass the conscious primary associations, which are very easily elicited (and thus usually well known), and to delve further into the less conscious majority of associations in which the potential for new insight is hidden. To succeed in eliciting less conscious associations we need a variety of techniques. The one word association test (such as 'what do you associate with Mercedes?') which is the most common technique for elicitation of brand associations, is not sufficient.

Verbalisation

The problem of verbalisation relates to the problem of access. Unconscious associations are also mainly non0verbal. They are raw impressions represented in visual, sensory or emotional modes and have not been subject to active cognitive reasoning. This contention is consistent with the general agreement that most human communication is non–verbal (e.g. Patterson 1991). In social encounters, individuals derive meaning from many other sources than verbal language such as eye contact or body language. Accordingly, non–verbal visual and sensory impressions give meaning to brands in addition to verbal concepts derived from linguistic reasoning. However, non–verbal associations are not readily communicable. Thus, we run the risk that these associations are not reported even when we gain access to them in memory.

Censoring

If we succeed in eliciting hidden, unconscious associations, and respondents manage to communicate them, we may still lose these associations because they are consciously or subconsciously held back by respondents. Interviews are not everyday events. To the contrary, they are extraordinary episodes that entail an explicit focus on the self of respondents (Haire 1950). When there is a focus on the self concept of individuals, psychological mechanisms of impression management and self deception are automatically activated (Sackeim 1988; Leary & Kowalski 1990). These processes help to maintain the self concept of individuals and to develop their identities. The result of impression management and self deception mechanisms in interviews is that respondents engage in monitoring their own responses and, to some extent, report associations that comply with their desired identities. Long interviews provide ample opportunity to replace sensitive information with less sensitive answers and to adjust associations to make answers more consistent with desired images.

Three general principles

The problems of access, verbalisation and censoring clarify that obtaining a valid and representative picture of associations for a given brand is an extremely difficult objective. However, knowledge of the nature of these difficulties makes it possible to address the problems systematically and to design elicitation studies that provide rich data of the most important aspects of brand perceptions. First, three general principles of elicitation are suggested.

Principle 1: Focus group interviews are not sufficient, long personal interviews are needed

Focus groups are particularly suitable when the research purpose is to study social reactions to brand stimuli or to obtain creative responses to new concepts or products (see Greenbaum 1993). Traditional focus groups are, however, not the best alternative for in-depth elicitation of brand associations. Hybrid focus groups and group support systems (GSS, see Sweeney et al. 1997) offer important improvements over traditional focus groups. In particular, the use of GSS, which implies that participants use sophisticated software on individual workstations, should alleviate censoring problems related to traditional focus group interviewing. Nevertheless, long personal interviews are preferred because of their superior potential to delve deeply into the memories of respondents by means of long, personal and individually adapted probing (Zaltman 1997; Malhotra 1999).

Principle 2: Select a portfolio of techniques that are complementary with respect to the three problems of access, verbalisation and censoring

One technique alone is not likely to be sufficient in overcoming all elicitation problems. Instead, a carefully chosen portfolio of techniques should be selected in which some techniques are especially adequate for evoking hidden associations, others are directed at overcoming the problems of verbalisation, and some are designed to mitigate censoring problems. This means that a variety of techniques need to be considered.

Principle 3: Validate responses

When customers are asked directly about brand associations, we can be fairly sure that we get relevant thoughts about the brand. However, as previously underscored, we probably get only the tip of the iceberg. When we make use of other and more indirect measures of brand associations, such as visual and projective techniques (to be described below), we increase the probability of detecting hidden associations, but at the same time increase the risk of evoking irrelevant associations. Hence, we need to validate elicited associations. This point cannot be overstated. In the 1950s and 1960s, there was a stream of qualitative research in marketing (often referred to as motivation research) in which several new techniques were developed or adapted from other disciplines such as clinical psychology, sociology and anthropology. However, in the early 1970s these methods were heavily criticised for their lack of scientific merit, and American market researchers in particular abandoned many techniques. It is commonly believed that this was due to the strong influence of positivistic philosophies of science (which imply an inordinate focus on quantification).

However, one might argue that lack of ability or willingness of motivation researchers to test qualitative responses quantitatively was also a major reason why motivation research failed to become a recognised part of scientific market research. Instead of opposing quantitative testing (though this was sometimes warranted), motivation researchers should have specified the type of qualitative responses amenable to quantitative testing and suggested appropriate measurement procedures. Although theoretical market researchers abandoned the new techniques, they were still used by applied market researchers both in Europe and the US. The issue of validation is, however, often not properly attended to.

Four main challenges of elicitation are derived from the review above, as shown at the top of Figure 1. First, we need inventive methods to obtain access and activate unconscious associations. Then, other methods should help respondents express both verbal and non–verbal responses. However, this is in vain if associations are actually held back (censored). Consequently, some techniques should be able to reveal censored information. Finally, elicitation results must be thoroughly validated. In the following, practical guidelines are developed for dealing with each of these elicitation challenges.

PRACTICAL GUIDELINES

Few theory–based guidelines exist in the published literature for the selection of techniques. In this paper, relevant characteristics of memory associations have been reviewed. This theoretical input and the limited body of empirical investigations of elicitation techniques derive the basis for the development of such guidelines. The presentation is organised as follows: first, the three measurement challenges are addressed successively, and relevant techniques for meeting each challenge are presented (see Figure 1). Next, the order in which techniques should be used, sampling issues, and interviewing flexibility are discussed.

The challenges of access and verbalisation

Because many of the same techniques that help researchers gain access to hidden association also mitigate problems of verbalisation (see Figure 1), these two challenges are dealt with jointly.

(1) Include at least one visual technique

Several visual techniques are used by research companies or are known from the literature. With one of the most popular techniques, often termed the moodboard technique, respondents are instructed to select pictures from magazines or newspapers that represent what they think or feel about the brand. Respondents describe the pictures, and associations are derived from content analysis of the descriptions. With another variant of this technique, participants are asked to take pictures with a camera in order to generate the stimuli (e.g. Zaltman 1997). Visual techniques of this kind are useful for two reasons. First, such techniques are based on the use of metaphors, and metaphors are known to be effective in evoking hidden unconscious knowledge (Glucksberg 1995). Thus, such techniques can be useful for eliciting unconscious sensory and emotional associations about brands. Second, visual techniques are not dependent on verbal language in activating associations. When respondents look at pictures representing their associations, it becomes easier to find the right words.

(2) Include at least one object–projective technique

Projective techniques in market research are adapted from psycho–analytical methods. These techniques are designed to cause respondents to project meaning onto some object (object-projective techniques) or personal figure (person–projective techniques). The distinction between object and person is important because the two types of projective technique have different strengths (person–projective techniques, which imply that thoughts are projected onto other persons, are discussed below in guideline 10). With object-projective techniques (OPT) respondents are instructed to describe the brand in terms of a car, an animal, a fabric, a vegetable, a celebrity, etc. (see Durgee & Stuart 1987). Such stimulus objects are well known and have rich associative networks. Thus, participants have ample opportunity for creating metaphors for their associations. This way, unconscious associations may surface. OPT may supplement visual techniques in obtaining access to hidden associations. One of the limitations of visual techniques is the pictorial stimuli used (e.g. magazines). It is impossible to provide an exhaustive set of stimuli because the variety of associations is not known. Important types of associations can then be missed if there are no pictures that make a link to them. OPT often lead to additional insights when objects are selected other than those found in the pictorial stimuli of visual techniques.

(3) Probe for secondary associations

The first associations reported are the most central conscious and verbal associations in memory. They are usually well known and often refer to product class, price, quality, or overall attitudes toward the brand. One simple way of delving deeper into the associative network is to use these primary associations as stimulus words for subsequent probing of secondary associations (e.g. 'what do you associate with quality?'). In turn, some important secondary associations can be used as cues in a third test, and so on. This snowballing technique is especially suited for gaining in depth insight into the multifaceted nature of primary associations, for example what it means for a specific brand in a specific product category to be perceived as a high quality brand.

(4) Probe for relevant situations

Brand associations are organised in memory clusters according to situations in which individuals have experienced the brand or drawn on knowledge about the brand. Consequently, we need to identify those situations (see also Restall & Gordon 1993, p. 63). This is usually not a very difficult objective. It is sufficient to use direct questions about the situations in which participants have encountered the brand or have been thinking of the brand (for example, in commercials, in use by oneself or others or in discussion with others, in purchase situations). When relevant situations have been identified, free association techniques with follow up probing of secondary associations can be used for each situation (such as 'what do you think or feel about the brand in this situation?'). Alternatively, or additionally, a situation–projective technique can be used. With this technique, respondents are subjected to drawings of situations involving the brand (e.g. a man driving a Mercedes) and instructed to describe the situations and the thoughts and feelings of the people involved (see Rook 1988).

(5) Address sensory associations directly

In order to capture important sensory associations, each of the senses should be addressed in succession (for example, 'which auditory impressions do you associate with the brand?'). This logic is consistent with the 'brand fingerprint' procedure of Restall & Gordon (1993). In addition to evoking product–related associations of appearance, sound, taste, smell, or feel, this technique may elicit important metaphors that activate unconscious associations. For example, in a study of sensory associations for a Scandinavian insurance company, one respondent described 'the smell of company X,' as similar to fertiliser. This metaphor was used because of the ability of fertilisers to make things grow, but also due to its bad scent. The bad scent metaphor in turn elicited important negative emotional associations about the personal interaction with company representatives, such as arrogance, self importance and lack of care.

(6) Use real stimuli when practically possible

Many unconscious sensory associations are hard to evoke even when they are addressed directly and respondents are given all the time they need to report associations. The use of real stimuli may elicit extra associations. For example, before asking users of a coffee brand about associations, we may let them have a taste of the coffee. The result could be very rich descriptions of taste experiences, emotions, and social experiences involving the brand. The advantage of using real stimuli is that primary emotional and sensory associations are activated by the experience itself. Therefore, less conscious secondary associations are more easily activated. The use of real product experience as stimuli is particularly pertinent to experiential products and brands focusing on sensory benefits.

However, other kinds of real stimuli are also useful, for example the brand elements. Brand elements are those physical devices that serve to identify the brand, such as brand names, logos, symbols, brand characters, slogans, jingles and packages (Keller 1998). The use of these elements as stimulus cues may clarify the role of each element in brand perceptions.

(7) Use established scales for emotional and personality associations

Emotions are very hard to access because they are mostly unconscious and non-verbal. General feelings for brands are often reported in elicitation sessions (for example, 'I like this brand' or 'this brand makes me feel good'), but the variety of underlying emotions determining overall reactions is often missed. Some emotions are identified by visual and projective techniques. However, because emotions are such important determinants of brand attitudes and behaviour, we need a more direct and systematic approach to elicit the full breadth of emotional associations. Direct questions about emotions are not recommended. People have numerous problems finding words for emotions and many feel awkward about discussing their personal feelings (Jourard 1964). One possible solution to this could be the use of pre–defined lists of emotions. For example, the Burke & Edell (1987) scale of emotions contains a large number of verbal descriptions such as 'pleased,' 'relaxed,' 'excited,' 'elated' and 'inspired'. When such scales are used for elicitation purposes, respondents are asked to read through the list of emotions and mark the ones relevant to their own emotional reactions to the brand. The strength of this technique is that respondents are spared the demanding task of finding words for non–verbal emotions. Moreover, responses are made on anonymous response sheets and not directly to an interviewer. In a similar vein, established scales of brand personality dimensions can be used to elicit relevant associations about typical brand users. Jennifer Aaker has developed an extensive scale for measurement of brand personality (Aaker 1997). The list includes five dimensions: sincerity, excitement, competence, sophistication and ruggedness. Parallel to the Burke & Edell scale, this scale can be used to identify relevant personality associations for specific brands by letting respondents tick off subsets of personality associations from the list which are associated with the brand in question.

(8) Instruct respondents to take their time and create acceptance for pauses

Irrespective of technique, it is imperative that respondents are allowed sufficient time to think and find words for non-verbal associations. In any interview, respondents will automatically and subconsciously be influenced by learned principles of conversation (see Molenaar & Smit 1996). One of these principles is that conversations should run fluently, without long pauses. Hence, respondents are likely to feel awkward when they cannot readily find words for non-verbal associations. On this account researchers should instruct respondents carefully and encourage them to pause during their responses.

Mitigating censoring effects

(9) Assure confidential treatment of responses

In order to feel free enough to probe deeply into their memories, respondents need to be confident that their responses will not be revealed to any third party or used for promotional purposes. Hence, thorough instructions are recommended in which respondents are assured that all responses will be held confidential. To avoid censoring of associations, however, assurances of confidentiality alone are insufficient.

(10) Use person–projective techniques

Confidentiality assurances do not remove the need of respondents to display a favourable image toward themselves and the interviewer. Because of this, person–projective techniques (PPT) are needed in order to identify censored associations. When using PPT, respondents are instructed to report associations on behalf of some person or figure belonging to the same group as the respondent (for example, 'most teenagers'). Thus, the focus of attention is removed from the respondent, the need for management of responses should be alleviated and respondents are freer to report sensitive associations (Fisher 1993). In addition to third person questions about brand associations, drawings of brand users in different situations can be useful (see Rook 1988).

Validation

(11) Validate minority associations on a subset of the majority

The ability of respondents to activate and verbalise associations varies considerably. Consequently, it is very common to obtain rich responses from some respondents and rather limited responses from others. For example, in the study of the Scandinavian insurance company referred to previously, one respondent, a musician, provided extremely rich responses. His associations led to several important insights about company perceptions. In elicitation interviews it is not uncommon that between one and three out of fifteen or twenty participants provide the information that leads to new insights. To validate the extra associations reported by this minority, it is useful to return to a sub–sample of the majority to test the relevance of these associations. Very often majority participants confirm that the extra associations are also relevant to their own conception of the brand.

(12) Criteria of salience and frequency should not be used uncritically

We are often interested in finding the most important brand associations because we have to focus on a subset of them in communication and in follow up surveys. In order to come up with a smaller number of associations, it is common practice to use salience scores and frequency of mentioning as selection criteria (see Keller 1998, chapter 8). Salience is often measured by recording the order in which associations are reported (thus, salient associations will have low average scores). Such criteria should not be used alone. First, as pointed out above, some respondents are better at finding words for their associations than others. Hence, if the frequency criterion is used strictly, important associations that many people hold but only a few manage to communicate could be missed (measurement of importance is discussed in the next section). Additionally, using the order in which associations are reported is dubious as an indicator of their centrality in memory. Important non–verbal associations that are hard to communicate could be prominent in memory, but reported late in an interview session because it took time to find the words. Moreover, respondents may hold back important sensitive associations until they feel that they have displayed an acceptable image and then report those associations. For this reason the selection of associations to be included in further studies or strategic planning processes should be made on a broader basis than just salience and frequency scores. For example, brand managers should be involved and the potential of associations to differentiate the brand and to build added value should be considered.

(13) Use a follow up survey to determine relationships between associations, strength of associations, favourability and uniqueness of associations

Qualitative techniques are adequate for identifying relevant associations. In order to determine relationships between associations, to measure the strength, importance and uniqueness of associations, quantitative measurements are needed. Measurement details on how to obtain valid measures of this kind are beyond the focus of this paper (see Keller 1998, chapter 8, for a discussion of quantitative measurements of various dimensions of brand associations). However, it should be noted that the importance of associations should not be measured by direct questioning because consumers often do not have good insight into how they actually make brand choices (see Jaccard & Sheng 1984). Instead, researchers should rely on qualitative inferences and statistical procedures (correlations with or regressions against measures of total evaluations, such as brand attitude and purchase intention). The validity of beliefs from person–projective questioning can be tested in a similar manner. When such beliefs are included in quantitative follow up studies, together with responses from direct questions, it is possible to compare the extent with which the two kinds of responses are related to overall measures of brand attitude or behavioural intention. Additionally, responses from person–projective techniques can be validated on previous research (secondary sources) and observations of target consumers (Hall & Rist 1999).

Sample issues

(14) Elicit associations from different types of customers and from the advertising people

The sample of customers should contain 'heavy users,' 'average users,' 'light users' and 'non–users'. Heavy users have more elaborate networks of associations than light users. The extra favourable associations held by heavy users are particularly interesting because they can be used in communications that strengthen brand perceptions among average and light users. The voice of non–users is important because knowledge of barriers toward brand usage may lead to discoveries of how to attract new customers. In addition, creative staff involved in the campaign should be interviewed using some of the techniques presented here. Since the perceptions of the creative staff will strongly influence which associations are built into the brand, brand managers need to ensure that all the creatives that work on the brand have the same basic understanding of it.

(15) Divide the sample into two and include both users and non–users

Each respondent should not be subjected to all techniques. First, we run the risk of respondent fatigue. Second, some of the techniques recommended above should not be used in conjunction with other techniques. For example, the moodboard technique (or another visual technique) and object–projective techniques (OPT) should be used for different respondents. When the same respondents are subjected to both, they are likely to repeat their answers to the first technique when responding to the second. For this reason, when both visual techniques and OPT are used, samples should be divided into two groups of 8–15 individuals each, depending on the heterogeneity of the sample.

The order of techniques

(16) Start with thorough instructions and visual techniques

Naturally, any elicitation interview begins with thorough instructions. As previously mentioned, instructions should ensure the confidentiality of responses, stimulate respondents to probe deeply into their memories, and encourage them to take pauses. After instruction, visual techniques should be used before verbal techniques. This is because verbalisation may impair memory for visual associations and associations that are difficult to put into words (Schooler & Engstler-Schooler 1990). Thus, if respondents are instructed to report associations verbally, and those associations are stored in a visual mode, respondents will not only have difficulty communicating those associations, but the same visual associations may be harder to elicit by means of other techniques later in the interview because the attempted verbalisations overshadow the original visual memory.

Individual differences and interviewer flexibility

(17) Adapt to individual differences in response styles and response attitudes

Participants in elicitation interviews have different personalities, different response abilities and different attitudes toward research. Hence, their cognitive, affective and behavioural reactions to the various techniques will differ. For example, neuro–linguistic programming (NLP) (Bandler & Grinder 1979; Gordon & Langmaid 1988) suggest that some people may prefer visual communication whereas others prefer oral communication. Hence, researchers should not expect every technique to work perfectly on every respondent.

The present guidelines suggest a number of different techniques to be used in elicitation interviews. Researchers should be flexible and exploit the specific techniques that work for each respondent and be aware of any signals of confusion, tension, ambiguity or response fatigue. Again, some familiarity with NLP principles may be helpful. For example, respondents turning away with arms folded across the body may signal a tense and closed attitude toward the interviewer. In such situations, the researcher should turn to other techniques or take a break in the interview session in order to resolve the tension. Adapting to individual differences is very demanding and is only learned through practical interviewing experience.

CONCLUSION

Brand management is a battle of the mind, and battles cannot be won without thorough knowledge of one’s own strengths and weaknesses or without a clear picture of the battlefield. In short, this is why in depth measurement of brand associations is so important. Brand managers need a deep understanding of the sources of brand equity in order to define optimal brand strategies, design effective communication, and to make sense of the competitive environment. To gain deep insight, they are dependent on qualitative research that delves deeply into the unconscious and non–verbal associations that people have for their brands.

This paper provides brand managers and researchers with a set of practical guidelines on how to achieve this objective. By using the guidelines and the theories of brand associations reviewed here, managers may improve their own research or become more demanding buyers of research in this important area.

REFERENCES

Aaker, D.A. (1991) Managing Brand Equity. New York: The Free Press.

Aaker, D.A. (1996) Building Strong Brands. New York: The Free Press.

Aaker, J. (1997) Dimensions of brand personality. Journal of Marketing Research, 34, pp. 347–356.

Anderson, J.R. (1983) The Architecture of Cognition. Cambridge: Harvard University Press.

Bandler, R. & Grinder, J. (1979) Frogs into Princes. New York: Real People Press.

Barsalou, L.W. (1983) Ad hoc categories. Memory and Cognition, 11, pp. 211–227.

Barsalou, L.W. (1985) Ideals, central tendency, and frequency of instantiation as determinants of graded structure. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, pp. 629–654.

Burke, M.C. & Edell, J.A. (1987) The impact of feelings on ad–based affect and cognitions. Journal of Marketing Research, 26, pp. 69–83.

De Chernatony, L. & McDonald, M. (1998) Creating Powerful Brands in Consumer, Service and Industrial Markets. Oxford: Butterworth–Heinemann.

de Sousa, R. (1987) The Rationality of Emotion. Cambridge: MIT Press.

Durgee, J. & Stuart, R. (1987) Advertising symbols and brand names that best represent key product meanings. Journal of Consumer Marketing, 4, pp. 15–24.

Fisher, R.J. (1993) Social desirability bias and the validity of indirect questioning. Journal of Consumer Research, 20, pp. 303–315.

Fiske, S.T. & Taylor, S.E. (1995) Social Cognition. New York: McGraw-Hill.

Glucksberg, S. (1995) Commentary on non–literal language: processing and use. Metaphor and Symbolic Activity, 10, pp. 47–57.

Gordon, W. & Langmaid, R. (1988) Qualitative Market Research: A Practitioner´s and Buyer´s Guide. Aldershot: Gower.

Greenbaum, T. (1993) The Handbook for Focus Group Research. New York: Lexington Books.

Haire, M. (1950) Projective techniques in marketing research. Journal of Marketing, 14, pp. 649–656.

Hall, A.L. & Rist, R.C. (1999) Integrating multiple qualitative research methods (or avoiding the precariousness of a one legged stool). Psychology & Marketing, 16, pp. 291–304.

Heath, R. (1999) Just popping down to the shops for a packet of image statements: a new theory of how consumers perceive brands. Journal of the Market Research Society, 41, 2, pp. 153–169.

Holbrook, M.B. & Hirschman, E.C. (1982) The experiential aspects of consumption: consumer fantasies, feelings, and fun. Journal of Consumer Research, 9, pp. 132–140.

Isen, A. (1993) Positive affect and decision making. In: The Handbook of Emotion (Lewis, M. & Haviland, J.M., eds). New York: The Guilford Press, pp. 261–277.

Jaccard, J. & Sheng, D. (1984) A comparison of six methods for assessing the perceived consequences of behavioural intentions. Journal of Experimental Social Psychology, 20, pp. 1–28.

Jourard, S.M. (1964) The Transparent Self. New York: D. Van Nostrand.

Kahnemann, D. (1994) New challenges to the rationality assumption. Journal of Institutional and Theoretical Economics, 150, pp. 18–36.

Kapferer, J–N. (1997) Strategic Brand Management. London: Kogan Page.

Keller, K.L. (1993) Conceptualising, measuring, and managing customer–based brand equity. Journal of Marketing, 57, pp. 1–22.

Keller, K.L. (1998) Strategic Brand Management. Upper Saddle River: Prentice Hall.

Kosslyn, S., Segar, M.C., Pani, J. & Hillger, L.A (1990) When is imagery used? Journal of Mental Imagery, 14, pp. 131–152.

Lakoff, G. & Johnson, M. (1980) Metaphors We Live By. Chicago: University of Chicago Press.

Leary, M.R. & Kowalski, R.M. (1990) Impression management: a literature review and two component model. Psychological Bulletin, 107, pp. 34–47.

Macrae, C. (1996) The Brand Chartering Handbook. Harlow: Addison–Wesley.

Malhotra, N.K. (1999) Marketing Research: An Applied Orientation, 3rd edn. New Jersey: Prentice–Hall.

Mariampolski, H. (1999) The power of ethnography. Journal of the Market Research Society, 41, 1, pp. 75–86.

Marks, L.E. (1996) On perceptual metaphors. Metaphors and Symbolic Activity, 11, pp. 39–66.

Molenaar, N.J. & Smit, J.H. (1996) Asking and answering yes/no questions in survey interviews: a conversational approach. Quality & Quantity, 30, pp. 115–136.

Patterson, M.L. (1991) In Fundamentals of Nonverbal Behaviour (Feldman, R.S. & Rime, B., eds). Cambridge MA: Cambridge University Press.

Pawle, J. (1999) Mining the international consumer. Journal of the Market Research Society, 41, pp. 19–32.

Plutchik, R. (1993) Emotions and the vicissitudes: emotion and psychopathology. In The Handbook of Emotion (Lewis, M. & Haviland, J.M., eds). New York: The Guilford Press, pp. 53–66.

Quelch, J.A., Harding, D. (1996) Brands versus private labels: fighting to win. Harvard Business Review, Jan–Feb, pp. 99–109.

Restall, A. & Gordon, W. (1993) Brands – the missing link; understanding the emotional relationship. Marketing and Research Today, May, pp. 59–67.

Rook, D.W. (1988) Researching consumer fantasy. Research in Consumer Behaviour, 3, pp. 247–270.

Sackeim, H. A. (1988) Self deception. A synthesis. In Self Deception: An Adaptive Mechanism? (Lockhard, J. S. & Paulhus, D.L., eds). New Jersey: Prentice-Hall

Schmitt, B. (1999) Experiential marketing. Journal of Marketing Management, 15, pp. 53–67.

Schooler, J. & Engstler–Schooler, T.Y. (1990) Verbal overshadowing of visual memories: some things are better left unsaid. Cognitive Psychology, 22, pp. 36–71.

Sweeney, J.C., Soutar, G.N., Hausknecht, D.R., Dallin, R.F. & Johnson, L.W. (1997) Collecting information from groups: a comparison of two methods. Journal of the Market Research Society, 39.

Tulving, E. (1972) Episodic and semantic memory. In Organisation and Memory (Tulving, E. & Donaldson, W., eds). New York: Academic Press.

Tulving, E. (1983) Elements of Episodic Memory. Oxford: Clarendon Press.

Zaltman, G. (1997) Rethinking market research: putting people back in. Journal of Marketing Research, 34, pp. 424–437.

<

http://www.warc.com
© WARC. All rights reserved.