Brand extensions: A manager’s perspective Received (in revised form): 28th June, 2005
EDWIN J. NIJSSEN holds a PhD from Tilburg University and is a professor of marketing at the Nijmegen School of Management, Radboud University Nijmegen, The Netherlands. His interests focus on relationship marketing, strategic and international marketing issues, brand management and new product development. He was a visiting professor at Michigan State University and New York University and his work has been published in the Journal of the Academy of Marketing Science, International Journal of Research in Marketing, Journal of Product Innovation Management and Long Range Planning, among others.
CLARA AGUSTIN is an assistant professor in the Department of Economy and Business, Universitat Pompeu Fabra, Spain. She holds a masters degree in psychology and is in the process of finishing her PhD. Her work has been published in the Journal of Marketing Research.
Abstract Brand extension research has focused on consumers’ evaluations of potential brand extensions. Little attention has been paid to the mental models used by brand extension decision makers, that is, their interpretation of market preferences and successful brand extension strategies. Using conjoint analyses a study was made of what marketing managers consider to be the ideal brand extension. The results show that their mental model for successful brand extension is driven by variables such as the consumers’ perceived fit between parent brand and extension, the positioning of the brand, and the level of added value of the extension product. Other variables, like the breadth of the product line of the parent brand and the number of brands in the company’s portfolio, also play a significant but marginal role. The results suggest that managers use a non-compensatory model of decision making.
INTRODUCTION
Edwin J. Nijssen Nijmegen School of Management, Radboud University Nijmegen, PO Box 9108, 6500HK Nijmegen, The Netherlands Tel: ⫹31 (0)24 361 1868 Fax: ⫹31 (0)24 361 1933 E-mail:
[email protected]
In 2003 alone, some 27,000 new food and household products were introduced to the US market.1 While the number of brand extensions cannot be derived from these data, the use of existing brand names on new products in different categories has become a major means to maximise the value of a firm’s intellectual property. Marketing managers use their company’s brand names to increase the rate of acceptance of new products by both retailers and consumers, while keeping advertising and promotion costs low.2,3 Thus brands have become an important launching platform for new products.4,5
Earlier brand extension research has aimed to explain the underlying mechanism of consumers’ evaluation and purchase intention regarding brand extensions. With a few exceptions6–8 this research has predominantly focused on consumers’ evaluations.9 Drawing mainly from psychological theories, it has provided an understanding of the key variables that consumers use for evaluating brand extensions.10 Recently, researchers have been more concerned with the external validity of their models, and have included more extension attribute information.11–13 Despite the large body of work on consumers’ evaluations of brand extensions, little academic attention
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
33
NIJSSEN AND AGUSTIN
has been given to what marketing managers think about the subject. This paper argues that one important and neglected link between customers’ evaluations of brand extensions and strategies for brand extension are the mental models that key decision makers use to interpret different options for extension. What do brand extension decision makers consider to be the ideal brand extension? And, more importantly, which factors do they generally include in the equation? This lack of attention in the literature to the brand extension decision makers’ point of view seems strange because their interpretation of customer and competitor behaviour, together with personal preferences for extension pathways, are likely to influence the success of brand extension strategies. The literature mentions many factors that influence the success of brand extensions. Next to consumer response, retailer acceptance, brand portfolio and risk considerations, among others, play a role.14 Many multinationals are currently in the process of restructuring their brand portfolios using line and brand extensions as paths to reach more integrated brand architectures.15 This process is clearly influenced by the nature, ie breadth and scope, of their brands and retailer power.16 The objective of this study is to understand the extent to which marketing managers’ mental model for making brand extension decisions coheres with the strategic literature on consumers and brand extension. It will help to discover the degree to which marketing managers use the same variables that have been used in the literature over the past 15 years. 34
Assuming that managers include both customer- as well as managementrelated variables, the research can help to uncover whether, and to what extent, these types of variables compete for attention or are weighted against each other. Two research questions were formulated to guide this research: — Which variables do marketing managers take into account while evaluating brand extensions for their brands? — What is the relative importance of these variables in the actual brand extension decision? This paper first develops a model of managers’ brand extension decision making drawing on interviews with a small number of marketing managers. A model is set forward and research hypotheses relating to the various elements of the model are proposed. The research methodology for the empirical study is then discussed, followed by the findings of a conjoint experiment involving 45 marketing managers. Managerial implications are examined, and directions for future research suggested.
LITERATURE REVIEW Ever since the seminal article of Aaker and Keller,17 brand extension research has predominantly focused on consumers’ evaluations, and has developed into a significant body of literature. Results show that perceived brand quality18,19 and the degree of fit between the extension and the parent brand20,21 are key drivers of brand extension success in consumer markets. A brand’s portfolio and extension his-
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
tory also affect consumers’ evaluations of a brand extension.22,23 A successful brand extension enhances consumers’ evaluations of a proposed subsequent extension,24 while a brand’s portfolio quality variance negatively moderates the positive relationship between the number of products affiliated with a brand and consumers’ confidence of brand extension success.25 Recent research has focused on increasing the external validity of consumer research experiments by increasing the amount of attribute information and introducing control variables.26–28 The results show, for instance, that ad repetition and ad content have the power to increase brand extension acceptance,29–31 and that price increases have a stronger negative effect on perceived value and purchase intention of a brand extension for similar rather than dissimilar extensions.32 Furthermore, the negative influence of dissimilarity between extension and brand on brand extension evaluation has been found to dissolve when also including or calculating target consumers’ innovation disposition.33 These and other results have provided important guidance to managers on how to build brand equity in general, and how to select successful brand extensions in particular. An integrated management model is still lacking, however. As an initial step towards developing such a model, in-depth interviews were conducted.
MODEL AND RESEARCH HYPOTHESES Based on in-depth interviews with four senior marketing managers of four national brand manufacturers of super-
market products, the key management dimensions of brand extension decisions were identified. The four companies were leaders in their field and had introduced many new brand extensions over the last decade. The managers all had more than ten years of experience with regard to brand extension decisions. For the interviews, steps suggested by Vriens and Wittink34 were adopted for conducting adaptive conjoint analyses. Each interview took approximately 4 hours. First, respondents were asked to talk about their experience with brand extensions and to provide examples. Secondly, they were asked to identify the key dimensions for evaluating brand extensions, drawing on their experience. Thirdly, they were asked to comment on the variables for successful brand extension identified from the consumer and management literature. A card-sorting task was also used to help the respondents identify and cluster variables. Finally, the choices of acceptable and unacceptable attribute levels of the variables were discussed. The data from the qualitative phase were then analysed. First, two researchers identified, independently, common themes in the interview data. They compared results and decided on a common classification scheme using three sets of variables, ie (1) new product or extension characteristics; (2) parent brand characteristics; and (3) brand portfolio characteristics. Next, a detailed classification of variables took place. Differences were discussed and resolved. The final results were returned to the four managers for review and feedback. They confirmed that the framework fitted their individual perspectives and had face validity. Although other variables might
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
35
NIJSSEN AND AGUSTIN
Parent brand characteristics
Brand extension decision
• Brand image • Brand breadth
Brand portfolio characteristics • Number of brands
Brand extension characteristics • Extension fit • Extension added value
Figure 1
Model of managers’ evaluations of brand extensions
exist, the model seems to capture the core variables of managers’ brand extension decision making. The framework is shown in Figure 1. Several research hypotheses relating to the impact of these variables on managers’ evaluation of brand extensions were then formulated. These are next discussed in more detail, referring back to the interviews with the experts.
New product or extension characteristics: Fit All managers mentioned the target audience’s perception of the fit between a parent brand and an extension as a requirement for brand extension success. They were aware of different forms of ‘fit’ and levels of similarity and dissimilarity between the brand and the extension. The level of fit was considered particularly important for image spill over to occur. The managers were most concerned with avoiding poor fit. Poor fit would result not only in a lack of acceptance of the new product, but 36
could also potentially damage the parent brand’s image and sales. These results are consistent with the extant brand extension research literature.35–37 Although the respondents considered a good fit important, two of the marketing managers noted that moderate fit was sometimes to be preferred. They referred to a lack of recognition of a brand extension when fit was too obvious. This is consistent with the findings of Meyers-Levi et al.,38 which show that extensions that require serious mental processing result in more positive extension evaluation, and with Yentis and Bond’s39 warning that obvious extensions may lack awareness. Therefore: H1: Managers prefer extensions with a good or moderate fit with the parent brand over extensions with a poor fit.
New product or extension characteristics: Added value In accordance with Swaminathan’s40 observation that managers should ‘focus
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
on ensuring that a brand extension will be able to sustain on its own in the long term’ (p. 441), the experts in the current study argued that a brand extension needs a certain amount of added value to gain competitive advantage in the marketplace and thus perform well. They argued that the value can come from real physical and/or emotional benefits (eg ingredients versus unique packaging). A competitive level of added value will prevent consumers from abandoning the extension after an initial trial, may enhance parent brand sales and decrease the risk of potential negative consequences for the parent brand when loyal customers defect after a disappointing experience with the new product.41 All managers also mentioned retailer power while discussing the level of added value of the brand extension. Retailers have become more critical regarding new product launches because their shelf space is limited and because there is more focus on direct product profitability.42,43 They often demand evidence of a brand extension’s sales potential in the form of positive market test results. Thus: H2: Managers prefer extensions with a high added value over those with moderate value, and a moderate- over a lowvalue alternative.
Parent brand characteristics: Brand positioning The managers interviewed argued that brand image was more important than product quality in their industries. Moreover, poor-quality supermarket products hardly exist any more. The managers also agreed that high-image
brands are easier to extend than low-image ones. High-image brands signal more manufacturer competence and trust than their low-end counterparts. Prior research findings confirm that brand rather than product quality is key for creating successful brand extensions,44 and that prestigious brands extend more easily.45 Thus: H3: Managers prefer extensions involving a parent brand with an excellent or good image over one with a moderate positioning.
Parent brand characteristics: Brand breadth Managers also took a brand’s extension history into account while evaluating a potential brand extension. The managers said that they preferred to extend brands with excellent extension track records over extending brands with little or no extension experience. Less-positive records made them more cautious. This is consistent with the literature, ie consumer studies.46–48 The pressure to find new avenues of growth has, however, forced managers into also extending brands with no or little extension experience, and sometimes into extending brands opportunistically, focusing on short- rather than longterm brand performance. Consistent with this, a theme in the interviews was potential brand dilution versus the lack of a proven path of extension of mono-brands. This suggests a pattern of decreasing followed by increasing extension risk. Thus: H4: Managers prefer extensions involving a brand that covers several different products over a single-product or many-product brand.
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
37
NIJSSEN AND AGUSTIN
Brand portfolio characteristics: Number of brands Brand portfolio management involves taking care of future cash flows by creating, developing and eliminating brands. This includes introducing new products, but also acquiring and selling off brands. For instance, recently the Dutch–British multinational Unilever reduced its brand portfolio from 1,600 to 400 brands and pruned product lines in order to be better able to compete in the global marketplace. Such change operations involve brand transformation strategies and include brand and line extensions.49 Referring to this development, several managers discussed the relationship between the number of brands and extension policies. In the case of a single or mono-brand, an unsuccessful extension might seriously jeopardise a business unit’s or company’s existence. In the case of a portfolio that consists of many brands, the proliferation may make it difficult to decide which brands to develop, and make it difficult to keep things within the brand manager’s span of control.50 Thus managers will prefer extending brand portfolios with a moderate number of brands because these combine moderate extension risk with moderate span of control problems. H5: Managers prefer brand extensions of brand portfolios that hold several rather than many or just one brand.
METHODOLOGY Sample The model and hypotheses were tested using a conjoint experiment. Conjoint analysis is a multivariate technique 38
based on random utility theory, which can be used to analyse the trade-offs that people make when making decisions. The technique focuses on calculating respondents’ utility functions and part-worths, ie the relative importance they give to each decision attribute.51,52 The sampling frame existed of 86 marketing managers of manufacturers of fast-moving consumer goods (food and non-food) listed in the ‘who’s who?’ guide of the Dutch marketing society, NIMA. The managers were contacted by telephone. Forty-five managers agreed to cooperate and met the criterion of having recently been involved in one or more brand extension decisions (response rate 52.3 per cent). Appointments for face-to-face meetings were planned and conducted. Table 1 reports some sample characteristics of the managers that cooperated. Although the majority involved marketing managers, several managers classified themselves as product or brand managers. The behavioural theory underlying conjoint analysis and decision strategies will be briefly discussed next.
Random utility theory According to Louviere,53 conjoint analysis builds on: (a) Lancaster’s consumer theory, which suggests that economic goods are best conceptualised as bundles of attributes; (b) behavioural decision theory in psychology, particularly the area of information processing and preference formation for certain categories of goods; and (c) Thurstone’s random utility theory, which explains choice behaviour based on a systematic and random component. Managers form preferences for
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
Table 1
Sample composition
Job description
%
Experience (average number of extensions)
Number of brands currently involved with
Marketing manager Product manager Other, eg brand manager
74 14 12
1.9 3.1 11.3
1.9 2.2 2.7
products or brand extensions based on beliefs about product/extension characteristics. These preferences are the basis for their choice decisions. Random utility theory suggests that managers will try to choose those alternatives, ie brand extensions, that they like the best, subject to constraints such as risk and cost. Once extension product beliefs are formed, they can compare the extensions holistically relative to their evaluation rules or ‘utility functions’. To model the decision effectively, a ‘researcher has [to] identify as exhaustive a set of decision variables as possible, [and] s/he must specify how these explanatory variables combine to influence choice’ (p. 228).54 The systemic component of each attribute can be specified in several ways. Based on allowing or not allowing for attribute comparisons, that is permitting or not permitting trade-offs, two cognitive comparison strategies can be distinguished.55,56 First, a compensatory model includes a systematic evaluation of all advantages and disadvantages. Under this strategy the decision maker allows more of one attribute to offset or compensate for less of another attribute of an alternative. Secondly, a non-compensatory model suggests that decisions are taken comparing only some attributes of each alternative. Here the decision maker focuses on some core attributes and
does not allow for a trade-off between attributes in general. There are several implications of using a compensatory or non-compensatory decision strategy. First, evaluation and decision time are shorter for non-compensatory than for compensatory decision making because of the lower number of comparisons involved. A non-compensatory model thus is more efficient, although its effectiveness depends on the quality of the set of decision criteria used. Secondly, the choice of a compensatory or non-compensatory decision strategy is dependent on the level of complexity of a problem. Payne57 found that subjects used a compensatory model if a decision problem was simple, and that they switched to non-compensatory decision making when complexity increased. Consistent with Payne’s results, Johnson et al.58 showed that a high number of attributes (>3) and negative correlation between the attributes caused subjects to shift from compensatory to noncompensatory choice models.
Experiment The conjoint experiment itself involved three stages. First, managers were asked to list all the extensions that they had been involved in and describe the most salient ones in more detail. This helped to bring to the fore the
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
39
NIJSSEN AND AGUSTIN
Table 2 Overview of independent variables and their levels
Variable
Extension fit
Levels
1. Good 2. Moderate 3. Poor
Extension added value 1. Little value 2. Moderate value 3. High value
Brand image
Brand breadth
1.Very high 2. High
1. Single product 2. Several different products 3. Many different products
3. Moderate
different factors affecting these sorts of decisions. Secondly, instructions were given as to the nature of the conjoint task. Each manager was told that they had just changed jobs and were, in the new job, being confronted with a new brand extension decision. A set of alternatives had been prepared and it was that manager’s task to select the best extension based on personal experience. Finally, respondents were asked to evaluate the conjoint task. The objective was to check possible fatigue.
Measurement The extension alternatives presented to each manager were based on the five attributes included in the model (extension fit, extension added value, parent brand image, parent brand breadth and number of brands in company portfolio), and used three levels per attribute (see Table 2). An optimal design was used rather than a factorial design in order to keep the number of alternatives for evaluation low (the conjoint task). Such a design is orthogonal and balanced, ie has no relationship between levels and attributes, and each level appears exactly the same number of times. Using this approach, 27 alternative brand extension profile cards were created. This number is well within the limit of the 40
Number of brands in portfolio 1. Single brand 2. Several brands 3. Many brands
maximum of 32 cards for subject ordering.59 Consistent with the full-profile method, each card contained a complete description of scores across all five attributes. To help the respondents rank order the cards, they were first instructed to rate each alternative using a seven-point scale (1 ⫽ very low preference for extension; 7 ⫽ very high preference). This resulted in seven or less subsets of cards. Subsequently the respondent was asked to order the cards of each subset further, arranging them from least preferred to most preferred brand extension. Finally, the respondent was allowed to reconsider the ranking of individual cards in the overall ranking. The time the respondent used for finishing the task and measured task evaluation were measured using two test-fatigue questions (1 ⫽ very uninteresting; 7 ⫽ very interesting). The task took between 20 and 30 minutes per respondent. All respondents considered the task moderately to very interesting (mean ⫽ 5.2 and standard deviation ⫽ 0.80). Based on the distribution of the fatigue variable, three cases with relatively low involvement were eliminated from the analysis.
Analysis The data were analysed in two stages using SPSS software. First, before
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
Table 3 Correlation matrix of the study constructs
Fit of the brand extension Fit of the extension brand Brand image
1.000
Brand image –0.096 (0.547) 1.000
Added value of the extension Brand breadth
Added value of the extension –0.390 (0.011) 0.132 (0.405) 1.000
Brand breadth 0.056 (0.726) 0.062 (0.697) 0.353 (0.022) 1.000
Number of brands in company portfolio
analysing the data, the correlation coefficients for all the constructs in the study were examined for potential interrelationships among the variables. The correlation matrix for all constructs is shown in Table 3. The highest correlation was between ‘fit between brand and extension’ and ‘added value of the extension product’ (r ⫽ –0.39, p < 0.01). This limited correlation suggested that multi-colinearity would not be a problem. In a second stage the conjoint analysis was performed. The validity of the observations was estimated using Pearson’s R (R ⫽ 0.99) and Kendall’s Tau (K ⫽ 0.90), and using the preference rank order number of the different brand extension alternatives as the dependent variable. Using a level of acceptance criterion for R2 of 0.65 for internal constancy, as suggested by Auty,60 two more cases were excluded from the analysis.
RESULTS Table 4 shows the results of the conjoint analysis. Both overall utilities and path worths are reported. Path worths express the importance of a particular attribute and are comparable
Number of brands in company portfolio –0.053 (0.740) –0.038 (0.810) –0.247 (0.114) 0.073 (0.646) 1.000
to beta weights in a regression equation.61 The model has an adequate overall fit (Pearson’s R > 0.995 and Kendall’s Tau, K > 0.947). The most important variables for explaining the managers’ most preferred brand extension are ‘fit of the brand extension’ and ‘added value of the extension’, ie both brand extension characteristics. The combined contribution of these factors is 76 per cent, with partial path worth effects of 47.72 and 28.17 per cent, respectively. The remaining 24 per cent is made up for the larger part, ie 18 per cent, by the two parent brand characteristics, ‘parent brand image’ (11.44 per cent) and ‘parent brand breadth’ (6.79 per cent). ‘Number of brands in company portfolio’ is the least important factor in the model, with a path worth of 5.88 per cent. Although significant, the last two effects are only marginal. Figures 2a–e show the detailed results for each variable. As shown in Figure 2a, ‘fit of the brand extension’ has a positive and linear relationship with ‘preference for brand extension’. Managers prefer a good to a moderate (t ⫽ 12.1, p < 0.00) and a moderate
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
41
NIJSSEN AND AGUSTIN
Table 4 Results of the conjoint analysis Averaged importance
47.72
28.17
11.44
Utility
6.2513 0.7884 –7.0397
–4.0106 0.4762 3.5344
0.9735 0.5026 –1.4762
Factor Extension fit –––– –––– Extension added value –– –– Brand image – –
Good Moderate Poor
Low Moderate High
Very high High Moderate
Brand breadth 6.79
–0.4286 –0.3095 0.7381
5.88
–0.1931 –0.0873 0.2804
Number of brand in portfolio Single brand Many brands Several brands
14.0000
Constant
Pearson's R = 0.995 Kendall's Tau = 0.947
Significance = 0.0000 Significance = 0.0000
over a poor fit (t ⫽ 14.42, p < 0.00). These findings partially support H1. ‘Added value of the brand extension’ also affected brand extension preference in a linear, positive way (see Figure 2c). Managers preferred extensions with high added value over those with only moderate (t ⫽ 5.97, p < 0.00) or little added value (t ⫽ 8.91, p < 0.00). The difference between moderate and little value (t ⫽ 10.38, p < 0.06) was also significant. This confirmed H2. ‘Parent brand image’ shows a positive relationship with brand extension ranking (see Figure 2b). Managers clearly prefer extending a prestigious 42
Single product Several products Many products
brand to a brand with a high or moderate image. The differences between prestigious-moderate and highmoderate images were both significant at t ⫽ 4.03, (p < 0.00) and t ⫽ 4.01 (p < 0.00), respectively. The difference between prestigious and high image had a borderline value (t ⫽ 1.92, p < 0.06). There thus is support for H3 . Contrary to expectation it was found that the managers preferred extending brands that carry many rather than one or several products (t ⫽ 4.35, p < 0.00 and t ⫽ 2.94, p < 0.01 respectively, see Figure 2d). The option of extending a brand with
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
Fit of the brand extension
5 3 1
0.78
–1 good
moderate
bad
–3 –5 –7
–7.04
–9 Figure 2a Fit of the brand extension
Brand image 1.5 1 0.5 0 –0.5
very high
high
moderate
–1 –1.5 –2 Figure 2b
Brand image
several products was evaluated similarly to a mono-brand (t ⫽ 0.151, p < 0.87). Thus H4 cannot be accepted. It suggests that managers in the sample were less concerned with brand dilution and used extension history as their main heuristic for evaluation of the fitness of a brand for further brand extension. ‘Brand portfolio’ only had a minor effect on brand extension preference (see Figure 2e). Although the effect was in the anticipated direction, its relative importance was smallest of all the attributes. Managers preferred extending a brand that was part of a portfolio of several brands rather than
one containing many or just one brand (t ⫽ 2.02, p < 0.05 and t ⫽ 2.19, p < 0.03, respectively). No difference was found between the alternatives of single versus many brands (t ⫽ 1.15, p < 0.26). These results clearly support H5 .
DISCUSSION The purpose of this study was to examine the mental model that marketing managers use for evaluating brand extensions. The most important decision variables of their mental model focused on brand extension and parent brand characteristics. Brand
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
43
NIJSSEN AND AGUSTIN
Added value of the brand extension 4 2 0 –2
high
some
any
–4 –6 Figure 2c Added value of the brand extension
Brand breadth 0.8 0.6 0.4 0.2 0 several
many
–0.2
one
–0.4 –0.6 Figure 2d
Brand breadth
No. of brands in company portfolio 0.4 0.3 0.2 0.1 0 –0.1
many
several
one
–0.2 –0.3 Figure 2e
Number of brands in company portfolio
portfolio played a significant but marginal role. The two brand extension characteristics, ‘fit of the brand extension’ 44
and ‘added value of the extension’, were the two strongest predictors of managers’ preference for a potential extension. These are the key variables
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
of the decision maker’s mental model. The importance of ‘fit of the brand extension’ refers to managers’ recognition of similarity as a prerequisite for consumers’ perception of a spill over of parent brand image and trust. This result is consistent with the results found on the role of product fit in consumers’ evaluations of brand extensions.62,63 The preference for maximal fit demonstrates that managers are unaware of (or not sensitive to) the danger that obvious brand extensions are not recognised as new products of the brand and thus are useless for revitalising the brand.64,65 The added value of the extension also played an important role in predicting managers’ brand extension preference. High value was preferred over little or moderate value, and moderate over little value. This is consistent with the finding of Alpert et al.66 that retailers will only adopt pioneer and follower brands that have added value over their competitors. It also is in accordance with Swaminathan’s67 conclusion that brand extensions need sustainable advantage from a consumer vantage point. Consumers will only consider repurchasing an extension after a trial when an adequate level of added value is present. According to brand extension decision makers, product fit is twice as important as brand extension added value for brand extension success. This finding suggests that managers are more concerned about generating trials, and thus initial adoption, than added value, although they do also recognise the need for unique selling points.68,69 While parent brand image affected the managers’ preference of brand extensions to a large extent, the influence of
parent brand breadth was limited. The finding concerning the influence of image is consistent with previous findings in consumer studies; the global attribute of brand quality plays a key role in the formation of a cognitive structure of a brand extension and helps to generate trials.70,71 The limited influence of brand breadth may be explained by managers’ focus on the short-term rather than long-term effects of an extension. In order to explore this possible explanation further, additional analyses were performed comparing the marketing managers with the product managers in the sample. The results unveiled a somewhat higher preference for extending broad brands for product managers when compared with marketing managers (p < 0.05), suggesting a lower concern about brand dilution for product managers compared to marketing managers. Limited sample size and a lack of a detailed job description for each manager hindered more detailed analyses and conclusions based on the current data. The effect of the number of brands in the company portfolio on managers’ preference for brand extensions was in the anticipated direction. Just as for brand breadth, however, its effect in the actual brand extension evaluation was relatively small. Examining the correlation matrix more closely (see Table 4), several negative correlations were identified, for example, between brand breadth and value. Together with the high number of attributes, this suggests that respondents may have shifted from a compensatory towards using a non-compensatory decision model.72 The results seem to indicate that managers might rely on a poliheuristic decision-making strategy,73 a two-stage process wherein
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
45
NIJSSEN AND AGUSTIN
the decision maker first employs a non-compensatory decision rule to eliminate variables (preliminary study) and then employs a compensatory decision procedure in an attempt to minimise risks and maximise benefits.
CONCLUSIONS, LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH These results show that decision makers’ mental maps consist of three sets of variables, ie those related to the brand, brand extension and company brand portfolio. For the managers in the sample the brand-extension-related variables such as fit and added value were at the heart of their mental decision model. Brand breadth and the number of brands in the portfolio played a less important role. This may lead to a situation in which variables such as potential brand dilution and level of brand portfolio risk are easily neglected. Two conclusions can be drawn. First, the study shows that managers’ and consumers’ mental models partially overlap. Secondly, large companies’ appointments of brand custodians at the level of vice president and the development of clear extension charters seem appropriate and necessary measures given the results. Managers tended to focus on brands’ proven extension qualities and seemed less concerned about potential over-extension and brand dilution. The study suffers from a number of limitations. In the first place, the sample sizes of the preliminary and experimental studies were relatively small. Also, a holdout sample was not collected, which may have potentially biased the fit results of the conjoint research. Secondly, the study focused on and distinguished only one single 46
decision maker, and assumed rational decision making behaviour. Consequently, the influences of company politics and joint decision-making were eliminated. The possibility of sequential decision making was also not considered. Thirdly, this study focused on the decision regarding the best brand extension. Often, however, a brand extension may compete with the choice of introducing a new brand.74,75 Fourthly, the critique that Klink and Smith76 voiced regarding the lack of product attribute information in consumer brand extension research also applies to this study. Only general characteristics were used. Finally, this study focused on supermarket products. Managers of different industries may have different risk profiles and consequently have different dispositions towards brand extension decisions. Research findings regarding different product categories could help the ability to generalise the results. Further research is clearly needed to help better understand the problems that managers face when dealing with brand extension decisions. First, future research preferably should use a larger sample and distinguish between product, marketing and brand managers. Level of experience with branding decisions and company career or incentive structures may be included as additional control variables. Secondly, a longitudinal study tracking a set of brand extensions may help with gaining an understanding of the effect of internal organisational politics and group processes on brand extension decision making. Thirdly, more detailed/complete extension cues could be presented. For instance, the added value of the extension could be specified, distin-
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
guishing benefits to the consumer and retailer and cost benefits. Fourthly, researchers could focus on brandextension-versus-new-brand decisions rather than on brand extensions alone. This may lead to more differential effects and provide a better understanding of managers’ perspectives on strategic brand management decisions in general. Finally, this type of research also may be extended towards the retailer’s vantage point on brand extensions. Managers are a valuable source of information for better understanding the success of brand extensions.
(10)
(11)
(12)
(13)
References (1) Bianco, A. (2004) ‘The vanishing mass market’, Business Week, 12th July, pp. 61–68. (2) Smith, D. C. and Park, C. W. (1992) ‘The effects of brand extensions on market share and advertising efficiency’, Journal of Marketing Research, Vol. 29, No. 3, pp. 296–313. (3) Balachander, S. and Ghose, S. (2003) ‘Reciprocal spillover effects: A strategic benefit of brand extensions’, Journal of Marketing, Vol. 67, No. 1, pp. 4–13. (4) Aaker, D. A. (1991) ‘Managing Brand Equity, Capitalizing on the Value of a Brand Name’, The Free Press, New York, NY. (5) Tauber, E. M. (1988) ‘Brand leverage: Strategy for growth in a cost-control world’, Journal of Advertising Research, Vol. 28, August/September, pp. 26–30. (6) Ambler, T. and Styles, C. (1997) ‘Brand development versus new product development: Toward a process model of extension decisions’, The Journal of Product and Brand Management, Vol. 6, No. 4, pp. 222–234. (7) Kapferer, J.-N. (1997) ‘Strategic Brand Management’, 2nd edn, Kogan Page, London, UK. (8) Sullivan, M. W. (1992) ‘Brand extensions: When to use them’, Management Science, Vol. 38, No. 6, pp. 793–805. (9) Bottomley, P. A. and Holden, S. J. S. (2001) ‘Do we really know how consumers
(14)
(15)
(16)
(17)
(18)
(19)
(20)
evaluate brand extensions? Empirical generalizations based on secondary analysis of eight studies’, Journal of Marketing Research, Vol. 38, No. 4, pp. 494–501. Czellar, S. (2003) ‘Consumer attitude toward brand extensions: An integrative model and research propositions’, International Journal of Research in Marketing, Vol. 20, No. 1, pp. 97–115. Grønhaug, K., Hem, L. and Lines, R. (2002) ‘Exploring the impact of product category risk and consumer knowledge in brand extensions’, The Journal of Brand Management, Vol. 9, No. 6, pp. 463–476. Klink, R. R. and Smith, D. C. (2001) ‘Threats to the external validity of brand extension research’, Journal of Marketing Research, Vol. 38, No. 3, August, pp. 326–335. McCarthy, M. S., Heath, T. B. and Milberg, S. J. (2001) ‘New brands versus brand extensions, attitudes versus choice: Experimental evidence for theory and practice’, Marketing Letters, Vol. 12, No. 1, pp. 75–90. Laforeˆ t, S. and Saunders, J. (1994) ‘Managing brand portfolios: How the leaders do it’, Journal of Advertising Research, Vol. 34 September/October, pp. 64–76. Douglas, S. P., Craig, C. S. and Nijssen, E. J. (2001) ‘Integrating branding strategy across markets: Building international brand architecture’, Journal of International Marketing, Vol. 9, No. 2, pp. 97–114. Alpert, F. H., Kamins, M. A. and Graham, J. L. (1992) ‘An examination of reseller buyer attitude toward order of brand entry’, Journal of Marketing, Vol. 56, No. 2, pp. 25–37. Aaker, D. A. and Keller, K. L. (1990) ‘Consumer evaluations of brand extensions’, Journal of Marketing, Vol. 54, No. 1, pp. 27–41. Park, C. W., Milberg, S. and Lawson, R. (1991) ‘Evaluation of brand extensions: The role of product feature similarity and brand concept consistency’, Journal of Consumer Research, Vol. 18, No. 2, pp. 185–193. Sheinin, D. A. and Schmitt, B. H. (1994) ‘Extending brands with new product concepts: The role of category attribute congruity, brand affect, and brand breadth’, Journal of Business Research, Vol. 31, No. 1, September, pp. 1–10. Meyers-Levi, J., Louie, T. A. and Curren, M. T. (1993) ‘How does the congruity of brand names affect evaluations of brand name extensions?’, working paper, University of Chicago, IL.
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
47
NIJSSEN AND AGUSTIN
(21) Aaker and Keller, ref. 17 above. (22) Swaminathan, V. (2003) ‘Sequential brand extensions and brand choice behavior’, Journal of Business Research, Vol. 56, No. 6, pp. 431–442. (23) Anand, B. N. and Shachar, R. (2004) ‘Brands as beacons: A new source of loyalty to multiproduct firms’, Journal of Marketing Research, Vol. 41, No. 2, pp. 135–150. (24) Keller, K. L. and Aaker, D. A. (1992) ‘The effects of sequential introduction of brand extensions’, Journal of Marketing Research, Vol. 29, No. 1, pp. 35–50. (25) Dacin, P. A. and Smith, D. C. (1994) ‘The effect of brand portfolio characteristics on consumer evaluations of brand extensions’, Journal of Marketing Research, Vol. 31, No. 2, pp. 229–242. (26) McCarthy et al., ref. 13 above. (27) Klink and Smith, ref. 12 above. (28) Keller, K. L. and Aaker, D. A. (1997) ‘Managing the corporate brand, the effect of corporate marketing activity on consumer evaluations of brand extensions’, Marketing Science Institute, working paper, report no. 97–106, May. (29) Lane, V. R. (2000) ‘The impact of ad repetition and ad content on consumer perceptions of incongruent extensions’, Journal of Marketing, Vol. 64, No. 2, April, pp. 80–91. (30) Pryor, K. and Brodie, R. J. (1998) ‘How advertising slogans can prime evaluations of brand extensions: Further empirical results’, Journal of Product and Brand Management, Vol. 7, No. 6, pp. 497–508. (31) Boush, D. M. (1993) ‘How advertising slogans can prime evaluations of brand extensions’, Psychology & Marketing, Vol. 10, January–February, pp. 67–78. (32) Taylor, V. A. and Bearden, W. O. (2002) ‘The effects of price on brand extension evaluations: The moderating role of extension similarity’, Journal of the Academy of Marketing Science, Vol. 30, No. 2, pp. 131–140. (33) Klink and Smith, ref. 12 above. (34) Vriens, M. and Wittink, D. R. (1990) ‘Conjoint analyses in market research’, in ‘Year Book 1990–1991 of the Dutch Society of Market Researchers’, De Vrieseborch, Haarlem, The Netherlands. (35) Kim, C. K., Lavack, A. M. and Smith, M. (2001) ‘Consumer evaluation of vertical brand extensions and core brands’, Journal of Business Research, Vol. 52, No. 3, June, pp. 211–222. (36) Sheinin and Schmitt, ref. 19 above. (37) Aaker, D. A. (1990) ‘Brand extensions: The
48
(38) (39)
(40) (41)
(42)
(43) (44)
(45) (46) (47) (48)
(49) (50)
(51)
(52)
(53)
(54) (55)
(56) (57)
good, the bad, and the ugly’, Sloan Management Review, Vol. 31, No. 4, pp. 47–56. Meyers-Levi et al., ref. 20 above. Yentis, A. and Bond, J. (1995) ‘Andrex comes out of the closet: A case history’, Marketing and Research Today, Vol. 23, No. 2, May, pp. 104–113. Swaminathan, ref. 22 above. Swaminathan, V., Fox, R. J. and Reddy, S. K. (2001) ‘The impact of brand extension introduction on choice’, Journal of Marketing, Vol. 65, No. 3, October, pp. 1–15. Messinger, P. R. and Narasmihan, C. (1995) ‘Has power shifted in the grocery channel?’, Marketing Science, Vol. 14, No. 2, pp. 189–223. Alpert et al., ref. 16 above. Bhat, S. and Reddy, S. K. (2001) ‘The impact of parent brand attribute associations and affect on brand extension evaluation’, Journal of Business Research, Vol. 53, No. 3, pp. 111–122. Park et al., ref. 18 above. Keller and Aaker, ref. 24 above. Dacin and Smith, ref. 25 above. DelVecchio, D. (2000) ‘Moving beyond fit: The role of brand portfolio characteristics in consumer evaluations of brand reliability’, Journal of Product and Brand Management, Vol. 9, No. 7, pp. 457–471. Douglas et al., ref. 15 above. Osler, R. (2004) ‘Making the case for a new brand name’, The Journal of Brand Management, Vol. 12, No. 2, pp. 81–91. Priem, R. L. (1992) ‘An application of metric conjoint analysis for the evaluation of top managers’ individual strategic decision making processes: A research note’, Strategic Management Journal, Vol. 13, No. 8, pp. 143–151. Green, P. E. and Srinivasan, V. (1978) ‘Conjoint analysis in consumer research: Issues and outlook’, Journal of Consumer Research, Vol. 5, No. 2, September, pp. 103–123. Louviere, J. J. (1994) ‘Conjoint analysis’, in Bagozzi, R. P. (ed.) ‘Advanced Methods of Marketing Research’, Blackwell Publishers, Cambridge, MA, pp. 223–259. Ibid. Von Newmann, J. and Morgenstern, O. (1953) ‘Theory of Games and Economic Behavior’, Princeton University, Princeton, NJ. Simon, H. A. (1981) ‘The Science of the Artificial’, MIT Press, Cambridge, MA. Payne, J. W. (1976) ‘Task complexity and contingent processing in decision making:
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
BRAND EXTENSIONS: A MANAGER’S PERSPECTIVE
(58)
(59)
(60)
(61) (62) (63)
An information search and protocol analysis’, Organizational Behavior and Human Performance, Vol. 16, No. 2, pp. 366–387. Johnson, E., Meyer, R. J. and Ghose, S. (1989) ‘When choice models fail: Compensatory models in negatively correlated environments’, Journal of Marketing Research, Vol. 26, No. 3, pp. 255–270. Steenkamp, J. E. B. M. (1985) ‘The construction of sets of profiles to estimate main and interaction effects using conjoint analyses’, in ‘Year Book 1985–1986 of the Dutch Society of Market Researchers’, De Vrieseborch, Haarlem, The Netherlands. Auty, S. (1995) ‘Using conjoint analysis in industrial marketing: The role of judgment’, Industrial Marketing Management, Vol. 24, No. 3, pp. 191–206. Priem, ref. 51 above. Aaker and Keller, ref. 17 above. Sheinin and Schmitt, ref. 19 above.
(64) (65) (66) (67) (68) (69)
(70) (71) (72) (73)
(74) (75) (76)
Meyers-Levi et al., ref. 20 above. Yentis and Bond, ref. 39 above. Alpert et al., ref. 16 above. Swaminathan, ref. 22 above. Dacin and Smith, ref. 25 above. Broniarczyk, S. M. and Alba, J. W. (1994) ‘Theory versus data in prediction and correlation tasks’, Organizational Behavior and Human Decision Processes, Vol. 57, No. 1, pp. 117–139. Bhat and Reddy, ref. 44 above. Park et al., ref. 18 above. Johnson et al., ref. 58 above. Mintz, A., Geva, N., Redd, S. and Carnes, A. (1997) ‘The effect of dynamic and static choice sets on political decision making: An analysis using the decision board platform’, American Political Science Review, Vol. 91, No. 3, pp. 553–566. Tauber, ref. 5 above. McCarthy et al., ref. 13 above. Klink and Smith, ref. 12 above.
HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 13, NO. 1, 33–49 OCTOBER 2005
49