Content uploaded by Byron Sharp
Author content
All content in this area was uploaded by Byron Sharp on Jun 16, 2014
Content may be subject to copyright.
Conceptualizing and measuring
brand salience
Jenni Romaniuk and Byron Sharp
University of South Australia, Australia
Abstract. Historically, brand salience has been considered synonymous with the brand
being ‘top of mind’ (mentioned first) when the product category is used to cue retrieval
from memory. In this article we argue that this conceptualization (and associated
measure) is too narrow. We show that there is value in distinguishing salience from the
concepts of awareness and attitude by conceptualizing brand salience as the brand’s
propensity to be noticed or come to mind in buying situations. Brand salience reflects
the quantity and quality of the network of memory structures buyers’ hold about the
brands. This article offers guidelines to facilitate research on the role of brand salience
in brand choice and buyer behaviour that are an important progression from the
evaluation (attitude) focus of contemporary marketing theory. Key Words
•
advertis-
ing
•
brand image
•
brand salience
•
buyer behaviour
•
memory
Introduction
Drawing mainly from social psychology, the concept of salience is most com-
monly associated with the ability of an item to ‘stand out’ from its environment
or background (Guido, 1998). Its origins in brand research are often attributed to
a discussion on attitudes by Krech and Crutchfield (1948) who state ‘Saliency
refers to the fact that not all of man’s beliefs stand out with equal prominence in
his cognitive field. He may be more aware of certain of his beliefs than others, they
may enter his thoughts more readily’ (cited in Myers and Alpert, 1977: 163). This
‘prominence’, from a cognitive perspective, is typically translated to accessibility
from long-term memory (Guido, 1998). Put simply, some things are noticed
more easily, and some things are easier to retrieve from long-term memory, so
they have a higher propensity to enter working memory.
While the salience of attributes or beliefs has received attention as a moderating
effect between brand beliefs and buyer behaviour (Fazio and Williams, 1986;
Fishbein and Ajzen, 1975), the salience of the brand has also been considered (Alba
327
Volume 4(4): 327–342
Copyright © 2004 SAGE
www.sagepublications.com
DOI: 10.1177/1470593104047643
articles
and Chattopadhyay, 1986; Miniard et al., 1989). It is this latter ‘brand salience’,
which we define as the propensity of the brand to be thought of by buyers (i.e.
‘stand out’ from memory) in buying situations, that is the focus of this article.
In marketing literature and research, brand salience has often been neglected in
favour of concepts such as brand attitude, image and, more recently, equity. In
part this is due to an over emphasis on evaluation as part of the buying process.
But it is also due to salience being conceptualized simply as ‘top of mind’ (TOM)
awareness. TOM awareness refers to whether or not buyers recall the brand first
(strictly speaking before any other brands in the category) when asked to name
brands from a given product category (i.e. the product category is the cue to
retrieve items from memory (e.g. Alba and Chattopadhyay, 1986)). Thus it is fre-
quently used either interchangeably with, or as a subset of, brand awareness
(Keller, 1993), rather than being considered as a stand-alone concept.
Recently brand salience has assumed a greater prominence in both brand and
advertising literature. There have been calls for the consideration of building
brand salience as an outcome of effective advertising (Bullmore, 1999; Ehrenberg
et al., 2000a; Miller and Berry, 1998), as well as discussion of the role of brand
salience from a brand management perspective (Ehrenberg et al., 1997; Keller and
Davey, 2001; Romaniuk, 2002). Therefore it seems timely to re-examine what the
concept of brand salience should mean in a marketing context and the implica-
tions for measurement. In this article, we explore how brand salience is relevant
to brand buying. The aim is to draw from a wide field of research, particularly
consumer memory and behaviour, to increase our understanding of this concept
and specify the requirements for a valid (and useful) measure. This will allow
researchers to examine empirically the contribution of brand salience to the
understanding and prediction of buyer behaviour.
Background
The most popular conceptualization of brand salience is as the accessibility or
‘prominence’ of the brand in buyer memory (Ajzen and Fishbein, 1980; Alba and
Chattopadhyay, 1986; Fazio, 1990). From a human memory perspective, promi-
nence can be apparent in two ways. The first is a ‘current’ prominence whereby
the very recently retrieved brand is in working memory (e.g. when an Elvis Presley
song is playing on the radio, Elvis becomes easier to recall as a singer). The second
is the longer-run average accessibility from long-term memory (e.g. for most
people, most of the time, Elvis has more chance of being retrieved than say,
Marvin Gaye). It should be noted that we focus in this article on the latter, the
accessibility from long-term memory. While it is possible that there is interest in
the accessibility of the brand name under varying conditions (e.g. when viewing
an ad, or when talking with friends), arguably situations of most interest for
the vast majority of marketers are buying or media consumption situations.
Therefore we are concerned with the ‘prominence’ of the brand when people are
in these situations. We are concerned with the buyer memory structures about
marketing theory 4(4)
articles
328
the brand that, as commented by Keller and Davey (2001), affect how often and
easily the brand is evoked under various situations or circumstances. Brand
salience is part of brand equity (the differential effects of having the brand name
(Ailawadi et al., 2003), being the effect of heightening propensity to be thought
of/noticed in buying situations.
Throughout the marketing literature and in market research practice the
concept of brand salience has been linked strongly to a single operationalization.
This is a buyer’s ease of recalling the brand (name) as a member of the product
category (Alba and Chattopadhyay, 1986). In doing this, researchers implicitly tie
the concept of brand salience to first, being the first mentioned or ‘top of mind’
(TOM) and second, with the product category label as the specific (and typically
only) retrieval cue. This measure of link to the product category cue is a subset of
the brand recall or recognition measures of brand awareness (Rossiter and Percy,
1997) and contributes to the ambiguity between the brand awareness and salience
concepts. It assumes that the product category cue is the only mechanism for
buyers to think of the brand in buying and media consumption situations.
However, research regarding how information in memory is accessed suggests
that this implicit assumption is not valid. Recalling brands from memory does not
just depend on a single cue (the product category) and noticing a brand is more
than simply recognising it (linking the name to the product category). We will
explore this in more detail, but first we briefly discuss the key literature about the
structure and processes of human memory.
Retrieval from memory
Since the 1980s, theories of how humans encode, store and retrieve information
have permeated marketing thought and theory development (e.g. Keller, 1987).
One of the theories most widely adopted into marketing is that of the Associative
Network Theories (ANT) of memory (e.g. Anderson and Bower, 1979).
Fundamentally, under these theories, memory consists of nodes that hold infor-
mation/concepts. If two pieces of information are ‘associated’, connections are
conceptualized as existing between them, thus making up a network of associated
information. If the information is new (e.g. seeing an ad for McDonalds and
salads for the first time) new links can be formed (i.e. ‘McDonalds’ and ‘salads’
become linked). If the consumer had prior exposure to these two concepts
together (seeing an ad for McDonalds offering salads for the second time), further
exposure is thought to reinforce the existing link (the link between ‘McDonalds’
and ‘salad’ becomes ‘stronger’). Both of these processes result in improved
accessibility of the information at a later point in time.
The links between concepts are asymmetric in strength (probability of activa-
tion/recall). So when asked to list all the things associated with McDonalds a con-
sumer might very well say ‘salads’, but be far less likely to say McDonalds if asked
to list all the things they associate with salads. Rossiter and Percy (1997: 110) point
out that some product categories are bought largely by recall (services), in which
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
329
case a strong product category need to brand name link is desirable, whereas
others are bought by recognition (groceries) in which case a strong brand name
to category need link is perhaps more desirable.
It is important to note that while we refer to retrieval from memory, we do not
confine the relevance solely to memory-based purchase situations (in line with
Lynch and Srull, 1982). Even when all or most brands are present at the point of
choice (e.g. at a supermarket) consumers still retrieve information from memory
and only notice specific brands, as evidenced by the short amount of time taken
to choose brands despite wide ranges on offer (Dickson and Sawyer, 1990).
Therefore the issue of retrieval of brand name information from memory is still
relevant in that context. Recently, Chandon and Wansink (2002) found that
heightening salience of a product (through pictures and pantry shelf position)
resulted in increased consumption incidence for high-convenience products.
They postulate that this higher salience resulted in the product being noticed
more frequently and therefore considered as an option to consume.
When a customer is exposed to the brand in a specific context, links in memory
between the brand name and specific concepts can be created or reinforced. This
network of information linked to the brand name constitutes what has been
referred to as the brand’s image (Keller, 1993) or more recently brand knowledge
(Keller, 2003). These linked concepts can be retrieved when the brand name is
used as the retrieval cue (e.g. thinking of Persil [brand name acting as cue] con-
jures up thoughts of in tablet form (attribute)), and/or cues to retrieve the brand
name when stimulated in a buying situation (e.g. want something refreshing
(attribute acting as cue) brings to mind Coca-cola (brand name)) (Holden, 1993;
Holden and Lutz, 1992). This second point is frequently neglected in marketing,
with most attention on the brand to attribute link. While it has been acknow-
ledged that the requirement of buyers to consider/think of the brand is a crucial
first step to purchase (e.g. Rossiter and Percy, 1987), it has usually been taken for
granted that the link to the product category is the cue that activates this process
(some noticeable exceptions to this being Holden and Lutz, 1992; Nedungadi,
1990). This means that work in building brand consideration has focused on
increasing awareness between the product category cue and the brand name. Once
this is established, brand awareness is typically taken as a given (‘everyone has
heard of Coca cola’ mentality). As a consequence, brand awareness is rarely con-
sidered to be a key marketing objective for established brands (Macdonald and
Sharp, 2003). However, the appropriate objective should be to ensure that the
brand continues to be noticed and/or thought of as an option to buy, not brand
awareness per say. Just because the brand is known or recognized as a member of
the category does not mean it will come to mind in buying situations, as retrieval
is dependent on the cue and the accessibility of the linked information (Alba and
Chattopadhyay, 1985; Holden and Lutz, 1992; Tulving and Pearlstone, 1966).
The commonly accepted process for retrieving information from long-term
memory is that a stimulus/cue activates a stored memory (Anderson and Bower,
1979). This ‘activation’ travels to memories linked to that initial cue. These con-
cepts can be retrieved and therefore used in working memory and/or can act as
marketing theory 4(4)
articles
330
retrieval cues to access further information (e.g. Collins and Loftus, 1975).
Consequently, the cues employed are a key influence of what is retrieved from
memory as only information linked to those cues is available. Any brand name
association can potentially act as a cue for accessing the brand name. Whether it
does or not, however, depends on the buying situation.
The cues that have received the majority of the attention in the literature are
those representing product categories (e.g. ‘offers home loans’, ‘soft-drinks’). A
relationship that links the brand, product category labels and stated brand
choice/preference, has been claimed by multiple researchers since the 1960s (e.g.
Axelrod, 1968; Cohen, 1966; Nedungadi, 1990). The emphasis on the product
category cue probably stems from early consumer behaviour models (e.g. Howard
and Sheth, 1969). In these models, the product category cue is used to access
memory to create the list of options (also referred to as the evoked/consideration
set) the buyer then considers/evaluates prior to choice. Other concepts or attrib-
utes are then posited to affect choice from within this set of considered brands. In
these models the consideration set for a buyer was thought to be a fixed entity,
with the same brands retrieved at different points in time and in different situa-
tions. Variation in brand choice across time/situations was thought to be due to
variation in the relative importance of attributes (Howard and Sheth, 1969; Miller
and Ginter, 1979). For example, when in a hurry, ‘quick service’ is more impor-
tant, while on a hot day ‘refreshing’ is more important.
However, empirical research has shown that in buying situations, customers
can (and do) use cues other than product categories to think of potential brands
to buy (Holden, 1993; Nedungadi, 1990). The cues used by buyers come from
both internal influences and the external environment. These are highly variable
factors and as such, the cues used in the brand choice process can vary between
buyers and for the same buyer over time. The impact of cues on retrieval is largely
subconscious and thus often unnoticed by buyers. We often have no conscious
appreciation of why we happen to be humming a particular song, or why we
suddenly thought of Cadbury chocolate. Even more importantly, we seldom
notice ourselves not noticing or not retrieving things. Rarely do we give thought
to why we visited a particular store to look for a new shirt and yet did not consider
a number of other stores we knew sold shirts. This is perhaps another reason why
the area of brand salience has been relatively neglected in the past by marketing
theorists. We have been distracted by what can happen to the information once
it is in buyer memory so we have neglected to understand the mechanisms for
getting the information there initially.
Cues are situation specific and how they impact can be circumstance specific.
Buyers can use combinations of cues at the one occasion (e.g. ‘Italian biscuits’, or
‘Italian biscuits for eating with an espresso’). Cues need not concern functional
aspects or benefits of the brand or its product category. Nor are cues necessarily
actual words or even easily able to be described by words. Buyers can also narrow
the list of retrievable items to a subcategory such that something easily retrieved
at the category level, is no longer accessible (Nedungadi and Hutchinson, 1985).
Therefore cues need not directly elicit the brand name but may instead simply
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
331
enhance the likelihood of a particular brand, or set of brands, being elicited
should another cue be encountered. Being at the beach in itself may not make
someone think of ice cream (sometimes, but probably mostly not). But if that
person on the beach encounters a cue like ‘a treat’ then they may be far more
likely to think of ice cream in that situation than they would if they encountered
that cue when in a shopping mall. Different cues frequently result in different
‘sets’ of options being retrieved from memory (Holden and Lutz, 1992;
Ratneshwar and Shocker, 1991).
Variability of cues is one factor that influences retrieval. Other reasons are
linked to the characteristics of buyer memory. First, there is the limited capacity
in working memory (e.g. Miller, 1956), which means there is only so much
people can think of at any one time. So while they might be able to name 20
brands of chocolate bars when they are probed, rarely in ‘real life’ will they think
of more than a few at any one time. Second, the importance buyers place on brand
choice is typically low (Heath, 1999) and therefore there is little motivation to go
beyond the easily accessible on any one occasion. These factors combine with the
influence of other brands (‘competitors’ linked to the relevant cues) to influence
(and vary) retrieval of any specific item. This makes retrieval from memory a
highly variable and unpredictable outcome at any one occasion.
This individual level variability is well documented for respondents’ statements
about their attitudes, perceptions and beliefs (e.g. Mills and Nelson, 1977; O’Brien
and Ford, 1988; Schuman and Presser, 1981). For instance, despite the total
proportion of respondents who associate a brand with a particular attribute
remaining the same over two interview waves, typically only around 50 per cent
of those who associate a brand with an attribute at Time A will do so again at Time
B (Castleberry et al., 1994; Dall’Olmo Riley et al., 1997; Sharp et al., 2002). This
variability is cue independent in that it is present in both directions; that is, from
attribute-to-brand and from brand-to-attribute (Sharp and Romaniuk, 2002).
That the individual level repeat rate of attribute responses is far from 100 per
cent shows that retrieval is variable. However, it has been recently discovered that
the repeat rate (RR) is systematically linked to the proportion of respondents who
associated the brand with the attribute on the first interview (roughly of the
relationship RR = initial response level + 20). This suggests that retrieval is prob-
abilistic, and possibly reflecting an underlying propensity that is different for each
individual (Sharp, 2002), in other words, their level of brand salience. This notion
of probabilistic retrieval is further supported by the modelling of the relationship
between initial response levels and repeat rates using a derivation of the NBD-
Dirichlet model (Rungie and Dall’Olmo Riley, 1999) – a theory which describes the
probability distributions of brand choice (Ehrenberg et al., 2000b; Goodhardt et
al., 1984). Therefore, the probabilistic nature of brand choice seen in buying
behaviour is also evident in buyer’s retrieval of brands. Indeed, this probabilistic
retrieval may, at least partially, explain the probabilistic nature of repeat buying.
In summary, the cues that customers use in buying situations are many and
varied and extend beyond the product category cue. The accessibility of the brand
for each of these cues, and their various combinations, is also variable due to
marketing theory 4(4)
articles
332
memory limitations (both ability and motive) and the presence of other
brands/options competing for retrieval. This suggests that a brand’s salience is not
reliably based on the strength of the single link to cue of the product category
name (as is traditionally supposed), but is better reflected in the propensity for the
brand to be retrieved/mentioned across a range of cues likely to be ‘used’ by
buyers in buying situations. This broadens brand salience from a conceptualiza-
tion based on the strength of association with a single cue/attribute to being based
on the presence of links across a range of attributes. This is a ‘prominence’ based
on the ‘quantity’ of links as well as the more traditional ‘quality’ of a single link.
Salience versus attitude
Brand attitude is a lasting evaluation (Solomon, 1992) that can act as a mental
reminder to act. It is well established that attitudes have only a weak influence on
future behaviour (Kraus, 1995; Wright and Klÿn, 1998). This is because they are
often not recalled (as shown in Dall’Olmo Riley, 1995; Dall’Olmo Riley et al.,
1997, 1999) and typically when they are, they are only weakly motivational.
Attitudes’ influence on buyer behaviour depends on first, the attitude being
remembered (e.g. ‘did I like this wine when I had it last?’) and second, how
motivationally strong the attitude is (e.g. ‘I really liked this wine’). We don’t
always have attitudes to draw upon to guide our behaviour, and many times all of
the options are effectively equally liked or disliked (they will all ‘do’). Conse-
quently the impact of attitudes on buying behaviour is generally weak, especially
in terms of influencing brand choice. In contrast, salience differentials have a huge
impact on individuals’ personal brand repertoires, and at aggregate level they do
much to explain the market share differences between brands.
We see the influence of brand salience as being quite different to that of brand
attitude. Attitudes are about evaluating the brand (do you think it is a good
brand?), whereas salience is largely about having a chance of being thought of (are
you likely to notice or retrieve the brand in that buying situation?).
Indeed, brand attitudes are, we think, often largely a function of salience, in that
we have a tendency to like the familiar (Harrison, 1968; Zajonc, 1968). If a brand is
more salient (under our conceptualization), then it will also be more likely to be
positively evaluated. Just as attitude research over the past few decades has steadily
moved in the direction of pointing up situational (non-attitudinal) aspects (Foxall,
2002) we downplay the evaluative role of stored/created past evaluations and
emphasize the role of the extent of the network of brand information.
Brand salience: selection from considered options
While we do see a distinction between salience and attitude, we do not confine the
role of salience to influencing the brand’s ability to be considered (and thereby
indirectly affecting brand choice, see Nedungadi, 1990). We also see a further role
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
333
for brand salience during the selection from the options that the buyer is con-
sidering. That is, brand salience also has a positive influence on that brand being
selected from that set of considered options, when more than one brand presents
itself to the customer (Macdonald and Sharp, 2000).
When the brand is retrieved, along with the brand name there is also evoked
some sense of whether this is a brand that the customer knows will do what they
require at that point in time. This provides a sense of assurance. Consumers have
something of a tendency to ‘assume the worst’ in the case of missing information
about brand attributes (Johnson and Levin, 1985). Therefore the more ‘gaps’ in
knowledge, or uncertainties about the brand, the lower the sense of assurance
(relative to other brands) and so the lower the probability of selection.
This assurance combines with customers’ inherent tendency to ‘satisfice rather
than maximise’ when buying brands (Simon, 1957, cited in Johnson and Puto,
1987). The lack of motivation to devote substantial cognitive processing efforts to
a brand purchase means that rather than carefully evaluate brands in order to
make the ‘best’ choice at any one point in time, buyers typically buy what they can
be reasonably confident will ‘do the job’. They do not attempt to evaluate all the
alternatives on the market, nor even all the ones they conceivably know of. Nor
do they usually work hard to choose the very best from the selected group of
alternatives they consider. Instead, buyers show a degree of loyalty in that they
happily adopt restricted repertoires of brands (Ehrenberg et al., 2004). This is
evident even in ‘involving’ or high-value choices (such as choosing a mortgage or
a car). Even those buyers who actively ‘research the market’ typically only evalu-
ate a tiny fraction of the available offers (Alba and Chattopadhyay, 1986; Hauser
and Wernerfelt, 1990; Howard and Sheth, 1969; Mitra, 1995; Narayana and
Markin, 1975; Nedungadi, 1990). Indeed ‘consideration’ sets of one are far from
uncommon.
Summary: conceptualizing brand salience
Brand salience is conceptualized as the propensity of the brand to be thought of
in buying situations. This is reflected in the quantity (how many) and the quality
(how fresh and relevant) of the network of brand information in memory, or the
brand’s ‘share of mind’. This conceptualization is based on buying situations
being complex multi-cue environments, which means that buyers are affected by
a range of cues, beyond the product category, when eliciting options (typically but
not necessarily brands (Holden and Lutz, 1992)) for buying. This variability in
cues is one explanation as to why there is variability in the retrieval of brands as
an option to be considered – not all brands are linked to all cues for all buyers.
However, the lack of motivation/ability to retrieve everything from memory and
the presence of competing brands that, through links to the same cues, also have
a probability of being thought of, means that even when the same cue is used
brand retrieval is not guaranteed. The more cues to which the brand is linked the
more likely it will be linked to the cues the customer is likely to come across in
marketing theory 4(4)
articles
334
buying situations, and hence the brand can be considered to have a greater
propensity to be thought of as an option to buy.
In addition, when the brand is retrieved, the customer also gets some sense of
how much they know about the brand. This provides a sense of assurance that
the brand will be appropriate for the situation. Thus, brand salience provides a
further benefit by then making the brand, once thought of, more likely to be
chosen. This benefit is, however, secondary to the influence of the brand being
thought of in the first instance, as without that initial step, positive evaluations
will have no effect on buyer behaviour (e.g. Holden and Lutz, 1992; Rossiter and
Percy, 1987).
This concept of brand salience is distinct from the concept of brand attitude,
where the focus is on evaluation of the brand rather than the quantity and quality
of the memory structures. It provides a different perspective and explanation for
the fact that buyers buy different brands in different circumstances and situations
(and even in the same situations).
In the following sections we discuss the framework for measuring brand
salience, how to interpret the results and some expected relationships with other
constructs to aid in future validity testing.
Framework for measuring brand salience
We now draw on the prior discussion to develop the key aspects that need to be
incorporated into a brand salience measure. We provide the framework for
researchers to measure brand salience, and then test the relationship between
brand salience and buyer behaviour. We suggest three key factors a measure of
brand salience should exhibit, which are as follows:
1. Contain a representative range of attributes/cues used to ‘think of’ brands;
2. Measure recall/noticing relative to competitors rather than for a single brand
independently;
3. Focus on whether the brand is thought of rather than seeking to determine
how favourably the brand is judged.
We now discuss each of these.
Contain a representative range of attributes used as cues
Previous salience measures have focused on links to the product category labels.
As discussed however, there is evidence that consumers use all sorts of different
attributes when retrieving brands. Buyers can use multiple cues both at a single
point in time and over different purchase occasions. Types of attributes that may
be relevant include buying/consumption situations (e.g. take to a barbecue, give
as a gift); benefits (e.g. something refreshing, low fat); or functional qualities
(e.g. has a lid I can screw back on, is red). For a list of categories of possible cues
see Holden (1993); Holden and Lutz (1992). In designing a list of attributes for a
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
335
salience measure, a diverse range should be incorporated to capture the quantity
(as well as quality) of the network in memory. Empirical testing should also be
directed to uncover if there are any attributes that are more or less appropriate in
a measure of brand salience. For example should ‘descriptive’ non-evaluative
attributes, given that they follow atypical usage response patterns (as per Barwise
and Ehrenberg, 1985), be included?
Measure relative to competitors rather than a single brand in isolation
Including competitors ensures that the measure is closer to replicating the actual
buyer experience. Retrieval from memory is a competitive activity, with linked
items in memory competing with each other when each cue is stimulated. A
measure that asks respondents to assess association between the brand and the
attribute individually will not capture the competitive interference effects. This
suggestion is supported by prior evidence that comparative scales (relative to
competitors) perform better than absolute scales (a single brand in isolation) in
studies looking to predict future purchase (Axelrod, 1968; Clancy and Garsen,
1970).
Focus on retrieval rather than how favourably respondents judge the brand
Most brand measurement instruments focus solely on evaluations of aspects of
the brand (e.g. good value, excellent service). While we are not saying these
attributes should be totally excluded from a salience measure, we suggest that the
majority of the attributes should be those that are used to think of brands/options.
These may often not be evaluations of specific brands, but concern more the
internal and external situation that the buyer is in.
This also has implications for the nature of the scale for measurement. Items
from memory are thought to be stored as concepts and generally either retrieved
or not. This suggests brand salience should be measured via a ‘yes/no’ measure
rather than a ‘degrees of association’ measure (such as rating scales). Scales (par-
ticularly) also imply a deliberation about the merits of the brand to develop
an evaluation, which is much more cognitive processing than generally seems to
happen. Thus it is more appropriate to use a measure that gives buyers attributes
(i.e. potential cues) and identifies links with brands than to give customers an
attribute and ask them to rate the brand on, say, a five-point scale.
There are some areas of measurement that are debatable as to their importance.
The first is whether the measure be brand-to-attribute or attribute-to-brand?
Given that we are focusing on the retrieval of the brand and competitor links, we
would suggest that attribute-based measures, where the attribute is given and
brands are linked would be more appropriate. However, this is an area for empiri-
cal testing to see if there is a substantive difference in the results provided by
different cue directions. The second is whether brand names should be prompted
or unprompted. There has been research that suggests that prompted/recall and
unprompted/recognition retrieval of the brand name are just different ways
marketing theory 4(4)
articles
336
of measuring the same underlying construct in memory (Laurent et al., 1995).
Others have argued, however, that these are very different mechanisms (Rossiter
and Percy, 1997). Therefore we recommend this also be included as a facet in
empirical testing of measures of brand salience.
Interpretation of results
This point concerns the output of measurement, or what specific information we
are trying to ascertain. A measure of brand salience should capture the extent to
which a customer knows and thinks about a brand. Given the variability of likely
cues and variability in responses to each of these cues over time (as per Castleberry
et al., 1994; Dall’Olmo Riley et al., 1997; Sharp et al., 2002), this indication of
pervasiveness in memory (or size) is more important, and indeed probably more
reliable, than the information pertaining to specific links.
Therefore analysis should focus on the overall results for each customer, such
as the proportion of instances a customer mentions the brand across a repre-
sentation of likely cues/attributes. Traditionally, researchers are concerned with
the proportion of customers that associate a brand with one specific attribute, the
product category label. While advocating an overall measure, we still recommend
a decompositional approach, which builds up from the attributes to create the
overall pictures of memory structures. While it would be good to have a single
measure that could capture this, the complexity of both memory and the inter-
action of cues and the lack of awareness of the buyer of these processes mean we
think it unlikely that a single survey question can capture brand salience.
These guidelines should form the basis for a valid measure of brand salience.
One clear implication is that, as we have argued, a single item measure using only
the product category cue should be an inadequate measure.
Future research agenda
This article suggests a new direction for research into the relationship between
brand perceptions and buyer behaviour. Clarifying this brand salience concept
and how it should be measured is important if we are serious about testing its con-
tribution to knowledge about why buyers buy the brands they do. However this is
only the start. The next step is to identify measures that fit the guidelines we have
established here and establish that the measures are valid. This article is deliber-
ately conceptual in nature to avoid tying this new conceptualization of brand
salience to a single measure and thereby repeating the mistakes of the past, where
the measure and the concept become one. Our objective is to provide the basis
such that multiple researchers can empirically test the role of brand salience in
buyer behaviour.
To aid in the identification of brand salience measures, we now state where we
would expect a salience measure to have a higher or lower correlation. This is to
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
337
allow researchers to undertake convergent and discriminant validity tests on any
proposed measure of brand salience. Based on our conceptualization of brand
salience, we would expect a brand salience measure to have:
1. A higher correlation with measures of brand familiarity than with brand atti-
tude as it is related to knowledge about the brand rather than evaluations of the
brand;
2. A higher correlation with the longer term brand repertoire than the brands
used at any one point in time as it should tap into the long term memory
structures. So brand salience should also show some ability to predict future
repertoire composition;
3. More stability over time at individual level than individual attribute
responses.
A key test of a brand salience measure is its ability to predict future change in
behaviour. We would expect that using a measure that is closer to the thought
processes would yield a measure that is more predictive than Top of Mind brand
awareness, attitudinal, image or equity measures have been. However, this needs
to be tested empirically.
Further research also needs to examine to what extent these salience measures
are sensitive to marketing activity. This will provide some insight into what effect
marketers, with their marketing activities, can reasonably expect to have on
buyers’ propensity to consider/buy.
There is much to be learnt about salience such as, for example, what makes an
item become more salient in memory. It is possible that much of the establish-
ment and reinforcement of salience occurs through unconscious mechanisms. For
instance, the frequency to which someone is exposed to a concept is thought to
have an unconscious effect on future salience (Hasher and Zacks, 1984). Also,
little is known about what characteristics make something more likely to be
noticed at the point of encoding the brand name or associated attribute (as per
Craik and Tulving, 1975). The exploration of these (and related) aspects from a
brand salience perspective should lead to further insights in how to influence
brand salience and therefore buying behaviour.
Should the testing of the causal impact of brand salience continue to prove
fruitful, there are important implications for advertising strategy. It would give
more weight to the marketing importance of getting the brand name noticed,
rather than communicating a persuasive message (Miller and Berry, 1998). A
salience perspective to branding would also require a reformulation of advertising
pre-testing instruments and advertising effectiveness measures. However, before
such steps are taken, as Axelrod (1968) suggests, extensive empirical testing
of validity, predictive accuracy and sensitivity needs to be undertaken. We
encourage marketing researchers to take on this challenging new area.
marketing theory 4(4)
articles
338
References
Ailawadi, K.L., Lehmann, D.R. and Scott A.N. (2003) ‘Revenue Premium as an
Outcome Measure of Brand Equity’, Journal of Marketing 67: 1–17.
Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behav-
iour. Englewood Cliffs, NJ: Prentice-Hall.
Alba, J.W. and Chattopadhyay, A. (1985) ‘Effects of Context and Part-Category Cues on
Recall of Competing Brands’, Journal of Marketing Research 22(August): 340–9.
Alba, J.W. and Chattopadhyay, A. (1986) ‘Salience Effects in Brand Recall’, Journal of
Marketing Research 23(November): 363–9.
Anderson, J.R. and Bower, G.H. (1979) Human Associative Memory. Hillsdale, NJ:
Lawrence Erlbaum.
Axelrod, J.N. (1968) ‘Attitude Measures That Predict Purchase’, Journal of Advertising
Research 8(1): 3–17.
Barwise, T.P. and Ehrenberg, A.S.C. (1985) ‘Consumer Beliefs and Brand Usage’,
Journal of the Market Research Society 27(2): 81–93.
Bullmore, J. (1999) ‘Advertising and Its Audience: A Game of Two Halves’, Inter-
national Journal of Advertising 18(3): 275–90.
Castleberry, S.B., Barnard, N.R., Barwise, T.P., Ehrenberg, A.S.C. and Dall’Olmo Riley,
F. (1994) ‘Individual Attitude Variations over Time’, Journal of Marketing Manage-
ment 10(1–3):153–62.
Chandon, P. and Wansink, B. (2002) ‘When Are Stockpiled Products Consumed
Faster? A Convenience-Salience Framework of Postpurchase Consumption Incidence
and Quantity’, Journal of Marketing Research 34(August): 321–35.
Clancy, K.J. and Garsen, R. (1970) ‘Why Some Scales Predict Better’, Journal of
Advertising Research 10(5): 33–8.
Cohen, L. (1966) ‘The Level of Consciousness: A Dynamic Approach to the Recall
Technique’, Journal of Marketing Research 3(May): 142–8.
Collins, A.M. and Loftus, E.F. (1975) ‘A Spreading Activation Theory of Semantic
Processing’, Psychological Review 82(6): 407–28.
Craik, F.I.M. and Tulving, E. (1975) ‘Depth of Processing and the Retention of Words
in Episodic Memory’, Journal of Experimental Psychology: General 104(3): 268–94.
Dall’Olmo Riley, F. (1995) ‘Changing Consumer Attitudes in Steady Markets’, Unpub-
lished thesis, Department of Philosophy, University of London.
Dall’Olmo Riley, F., Ehrenberg, A.S.C., Castleberry, S.B., Barwise, T.P. and Barnard,
N.R. (1997) ‘The Variability of Attitudinal Repeat-Rates’, International Journal of
Research in Marketing 14(5): 437–50.
Dall’Olmo Riley, F., Rink, L. and Harris, P. (1999) ‘Patterns of Attitudes and Behaviour
in Fragmented Markets’, Journal of Empirical Generalisations in Marketing Science 4:
62–91 (Available at: http://www.EmpGens.com).
Dickson, P.R. and Sawyer, A.G. (1990) ‘The Price Knowledge and Search of Super-
market Shoppers’, Journal of Marketing 54(July): 42–53.
Ehrenberg, A.S.C., Barnard, N. and Scriven, J. (1997) ‘Differentiation or Salience’,
Journal of Advertising Research 37(November–December): 7–14.
Ehrenberg, A.S.C., Barnard, N., Kennedy, R. and Bloom, H. (2000a) ‘Brand Advertising
as Creative Publicity’, Journal of Advertising Research 42(July/August): 7–18.
Ehrenberg, A.S.C., Barnard, N.R. and Sharp, B. (2000b) ‘Decision Models or Descrip-
tive Models?’, International Journal of Research in Marketing 17(2–3): 147–58.
Ehrenberg, A.S.C., Uncles, M.D. and Goodhardt, G.G. (2004) ‘Understanding Brand
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
339
Performance Measures: Using Dirichlet Benchmarks’, Journal of Business Research
(forthcoming).
Fazio, R.H. (1990) ‘A Practical Guide to the Use of Response Latency in Social
Psychological Research’, in C. Hendrick and M.S. Clark (eds) Research Methods in
Personality and Social Psychology, pp. 75–97. Thousand Oaks, CA: Sage.
Fazio, R.H. and Williams, C.J. (1986) ‘Attitude Accessibility as Moderator of the
Attitude-Perception and Attitude-Behavior Relations: An Investigation of the 1984
Presidential Election’, Journal of Personality and Social Psychology 51(3): 505–14.
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behaviour: An Intro-
duction to Theory and Research. Reading, MA: Addison-Wesley.
Foxall, G. (2002) ‘Marketing’s Attitude Problem – and How to Solve It’, Journal of
Consumer Behaviour 11(1): 19–48.
Goodhardt, G.J., Ehrenberg, A.S.C. and Chatfield, C. (1984) ‘The Dirichlet: A Compre-
hensive Model of Buying Behaviour’, Journal of the Royal Statistical Society 147(5):
621–55.
Guido, G. (1998) ‘The Dichotic Theory of Salience: A Framework for Assessing Atten-
tion and Memory’, European Advances in Consumer Research 3(3): 114–19.
Harrison, A.A. (1968) ‘Response Competition, Frequency, Exploratory, Behavior, and
Liking’, Journal of Personality and Social Psychology 9(4): 363–8.
Hasher, L. and Zacks, R.T. (1984) ‘Automatic Processing of Fundamental Information:
The Case of Frequency of Occurance’, American Psychologist 39(December): 1372–88.
Hauser, J.R. and Wernerfelt, B. (1990) ‘An Evaluation Cost Model of Consideration
Sets’, Journal of Consumer Research 16(4): 393–408.
Heath, R. (1999) ‘The Low-Involvement Processing Theory’, Admap 34(March): 14–17.
Holden, S.J.S. (1993) ‘Understanding Brand Awareness: Let Me Give You a C(L)Ue!’, in
L. McAlister and M. Rothschild (eds) Advances in Consumer Research, pp. 383–8.
Provo, UT: Association for Consumer Research.
Holden, S.J.S. and Lutz, R.J. (1992) ‘Ask Not What the Brand Can Evoke; Ask What Can
Evoke the Brand?’, in J. Sherry and B. Sternthal (eds) Advances in Consumer Research,
pp.101–7. Provo, UT: Association for Consumer Research.
Howard, J.A. and Sheth, J.N. (1969) The Theory of Buyer Behaviour. New York: Wiley.
Johnson, R.D. and Levin, I.P. (1985) ‘More Than Meets the Eye: The Effect of Missing
Information on Purchase Evaluations’, Journal of Consumer Research 12(September):
169–77.
Johnson, M.D. and Puto, C.P. (1987) ‘A Review of Consumer Judgment and Choice’,
Review of Marketing: 236–92.
Keller, K.L. (1987) ‘Memory Factors in Advertising: The Effect of Advertising Retrieval
Cues on Brand Evaluations’, Journal of Consumer Research 14(December): 316–
33.
Keller, K.L. (1993) ‘Conceptualizing, Measuring, and Managing Customer-Based Brand
Equity’, Journal of Marketing 57(January): 1–22.
Keller, K.L. (2003) ‘Brand Synthesis: The Multidimentionality of Brand Knowledge’,
Journal of Consumer Research 29(March): 595–601.
Keller, K.L. and Davey, K.K. (2001) ‘Building Customer-Based Brand Equity’, paper
presented at Advertising Research Foundation workshop, New York.
Kraus, S.J. (1995) ‘Attitudes and the Prediction of Behavior: A Meta-Analysis of the
Empirical Literature’, Personality and Social Psychology Bulletin 21(1): 58–75.
Krech, D. and Crutchfield, R.S. (1948) Theory and Problems of Social Psychology. New
York: McGraw-Hill.
marketing theory 4(4)
articles
340
Laurent, G., Kapferer, J.-N. and Roussel, F. (1995) ‘The Underlying Structure of Brand
Awareness Scores’, Marketing Science 14(3:2): G170–9.
Lynch, J.G., Jr and Srull, T.K. (1982) ‘Memory and Attentional Factors in Consumer
Choice: Concepts and Research Methods’, Journal of Consumer Research 9(June):
18–37.
Macdonald, E. and Sharp, B. (2000) ‘Brand Awareness Effects on Consumer Decision
Making for a Common, Repeat Purchase Product: A Replication’, Journal of Business
Research 48(1): 5–15.
Macdonald, E. and Sharp, B. (2003) ‘Management Perceptions of the Importance of
Brand Awareness as an Indicator of Advertising Effectiveness’, Marketing Bulletin 14:
article 2.
Miller, G.A. (1956) ‘The Magic Number Seven, Plus or Minus Two: Some Limits on
Our Capacity for Processing Information’, The Psychological Review 63(2): 81–97.
Miller, K.E. and Ginter, J.L. (1979) ‘An Investigation of Situational Variation in Brand
Choice Behavior and Attitude’, Journal of Marketing Research 16(February): 111–23.
Miller, S. and Berry, L. (1998) ‘Brand Salience Versus Brand Image: Two Theories of
Advertising Effectiveness’, Journal of Advertising Research 38(5): 77–83.
Mills, P. and Nelson, E.H. (1977) ‘Re-Interviewing in Attitude Surveys: An Experi-
mental Study’, paper presented at ESOMAR, London.
Miniard, P.W., Unnava, H.R. and Bhatla, S. (1989) ‘Inhibiting Brand Name Recall: A
Test of the Salience Hypothesis’, in T Srull (ed.) Advances in Consumer Research, pp.
264–70. Provo, UT: Association for Consumer Research.
Mitra, A. (1995) ‘Advertising and the Stability of Consideration Sets over Multiple
Purchase Occasions’, International Journal of Research in Marketing 12(1): 81–94.
Myers, J.H. and Alpert, M.I. (1977) ‘Semantic Confusion in Attitude Research: Salience
Vs. Importance Vs. Determinance’, in W. Perreault (ed.) Advances in Consumer
Research, pp.106–10. Provo, UT: Association for Consumer Research.
Narayana, C.L. and Markin, R.T. (1975) ‘Consumer Behavior and Product Perform-
ance: An Alternative Conceptualization’, Journal of Marketing 39(October): 1–6.
Nedungadi, P. (1990) ‘Recall and Consumer Consideration Sets: Influencing Choice
without Altering Brand Evaluations’, Journal of Consumer Research 17(December):
263–76.
Nedungadi, P. and Hutchinson, J.W. (1985) ‘The Prototypicality of Brands: Relation-
ships with Brand Awareness, Preference and Usage’, in E.C. Hirschman and M.B.
Holbrook (eds) Advances in Consumer Research, pp. 498–503. Provo, UT: Association
for Consumer Research.
O’Brien, S. and Ford, R. (1988) ‘Can We at Last Say Goodbye to Social Class’, Journal
of the Market Research Society 30(3): 289–332.
Ratneshwar, S. and Shocker, A.D. (1991) ‘Substitution in Use and the Role of Usage
Context in Product Category Structures’, Journal of Marketing Research 28(August):
281–95.
Romaniuk, J. (2002) ‘Want to Make Your Brand Successful?’, B&T Weekly 11 October:
12.
Rossiter, J.R. and Percy, L. (1987) Advertising and Promotion Management. Singapore:
McGraw-Hill.
Rossiter, J.R. and Percy, L. (1997) Advertising Communications and Promotion Manage-
ment. New York: McGraw-Hill.
Rungie, C. and Dall’Olmo Riley, F. (1999) ‘Patterns of Repeat-Rates for Free Response
Attribute Questions’, working paper, University of South Australia, Adelaide.
Conceptualizing and measuring brand salience
Jenni Romaniuk and Byron Sharp
341
Schuman, H. and Presser, S. (1981) Questions and Answers in Attitude Surveys. New
York and London: Academic Press.
Sharp, A. (2002) ‘Searching for Boundary Conditions for an Empirical Generalisation
Concerning the Temporal Stability of Individual’s Perceptual Responses’,
Unpublished thesis, Department of Philosophy, University of South Australia.
Sharp, A. and Romaniuk, J. (2002) ‘Brand to Attribute or Attribute to Brand – What is
the Path to Stability’, paper presented at European Marketing Academy Conference,
University of Minho, Braga, Portugal.
Sharp, B., Riebe, E., Dawes, J.G. and Danenberg, N. (2002) ‘A Marketing Economy of
Scale – Big Brands Lose Less of Their Customer Base Than Small Brands’, Marketing
Bulletin 13: research note 1.
Solomon, M.R. (1992) Consumer Behaviour: Buying, Having, and Being. Needham
Heights, MA: Allyn and Bacon.
Tulving, E. and Pearlstone, Z. (1966) ‘Availability Versus Accessibility of Information
in Memory for Words’, Journal of Verbal Learning and Verbal Behavior 5(4): 381–91.
Wright, M. and Klÿn, B. (1998) ‘Environmental Attitude – Behaviour Correlations in 21
Countries’, Journal of Empirical Generalisations in Marketing Science 3: 42–60.
Zajonc, R.B. (1968) ‘Attitudinal Effects of Mere Exposure’, Journal of Personality and
Social Psychology – Monograph Supplement 9(2:2): 1–27.
Dr Jenni Romaniuk is a Senior Research Associate in the Marketing Science Centre,
University of South Australia. She leads the Brand Salience research project which
involves a number of leading companies around the world including Procter and
Gamble, Kraft and E&J Gallo. Jenni travels extensively presenting her work on brand-
ing regularly at conferences around the world, and to corporate sponsors. Her recent
work has been published in the Journal of Marketing Management, Journal of Marketing
Communications, Journal of Targeting, Measurement and Analysis for Marketing and
other international journals.
Address: Marketing Science Centre, University of South Australia, PO Box 2471,
Adelaide, South Australia 5001.
[email: Jenni.Romaniuk@MarketingScienceCentre.com]
Professor Byron Sharp is director of the Marketing Science Centre at the University of
South Australia. He is also director of the R&D Initiative, a program of fundamental
research in collaboration with Professor Andrew Ehrenberg at London South Bank
University. The R&D Initiative is supported financially by corporations around the
world including Coca-Cola, General Motors, Turner Broadcasting, Unilever, Highland
Distillers, BT, Research International and many others. Dr Sharp’s work focuses on
empirical generalization and empirically grounded theories of buyer behaviour and
brand performance. He is on the editorial boards of five journals including Marketing
Theory, and has published in journals such as the International Journal of Research in
Marketing, Journal of Business Research, Australasian Marketing Journal, Journal of
Consumer Marketing and others. More information is available at www.ByronSharp.
com
Address: Marketing Science Centre, University of South Australia, PO Box 2471,
Adelaide, South Australia 5001
[email: Byron.Sharp@MarketingScienceCentre.com]
marketing theory 4(4)
articles
342
