Customer Loyalty and Customer Loyalty Programs
Mark D. Uncles School of Marketing University of New South Wales Sydney NSW 2052 Australia
[email protected] Grahame R. Dowling Australian Graduate School of Management University of New South Wales Sydney NSW 2052 Australia
[email protected]
Kathy Hammond London Business School Regent’s Park London NW1 4SA
[email protected]
School of Marketing Working Paper 98/6 Paper submitted to Journal of Consumer Marketing April 2002 Revised and resubmitted to Journal of Consumer Marketing August 2002 Final revision for Journal of Consumer Marketing October 2002
We thank Jack Cadeaux, Robert East, Jennifer Harris, Byron Sharp and Chris Styles for their constructive suggestions. Also, all those who commented on earlier drafts of this paper at workshops organized by the Marketing Science Institute, University of New South Wales and University of Melbourne. The assistance of an Australian Research Council (Small Grant) is acknowledged.
Author Biographies Mark D. Uncles Professor of Marketing and Head, School of Marketing, University of New South Wales. His research interests include buyer behavior, brand and store choice, and the analysis of consumer panels. His work has appeared in journals such as Marketing Science, Sloan Management Review, Journal of Retailing, International Journal of Research in Marketing, Journal of Advertising Research. He is co-author of the New Penguin Dictionary of Business (Penguin, 2003). Grahame R. Dowling Professor of Marketing, Australian Graduate School of Management, a joint school of the University of New South Wales and the University of Sydney. His current research focuses on customer loyalty programs and corporate reputations, with publications appearing in journals such as the Journal of Marketing and Sloan Management Review. His most recent book is Creating Corporate Reputations (Oxford University Press, 2001).
Kathy Hammond Assistant Professor of Marketing, London Business School. Her research covers a wide range of consumer buying studies, focusing particularly on brand loyalty and customer relationship management. She is a frequent speaker and writer on New Media issues, and has published in a number of marketing journals including Marketing Science, Journal of Advertising Research, Marketing Letters, and Journal of Interactive Marketing. She is co-author of a short book, Predictions: Media (Phoenix, 1998).
2
Customer Loyalty and Customer Loyalty Programs
Abstract
Customer loyalty presents a paradox. Many see it as primarily an attitude-based phenomenon that can be influenced significantly by Customer Relationship Management initiatives such as the increasingly popular loyalty and affinity programs. However, empirical research shows that loyalty in competitive repeat-purchase markets is shaped more by the passive acceptance of brands than by strongly-held attitudes about them. From this perspective, the demandenhancing potential of loyalty programs is more limited than might be hoped. We review three different perspectives on loyalty, and relate these to a framework for understanding customer loyalty that encompasses Customer Brand Commitment, Customer Brand Acceptance and Customer Brand Buying. This framework is used to analyze the demand-side potential of loyalty programs. We discuss where these programs might work and where they are unlikely to succeed on any large scale. A checklist for marketers is provided.
Keywords Customer relationship management; Customer loyalty; Customer loyalty programs; Affinity programs; Buyer behavior
3
Customer Loyalty and Customer Loyalty Programs
1. Introduction The past decade has seen many firms (re)adopt a customer focus - often through a formal program of Customer Relationship Management – CRM (e.g., Brown 2000; Kalakota and Robinson 1999; Peppers and Rogers 1997). Recent advances in information technology have provided the tools for marketing managers to create a new generation of CRM tactics. One such tactic that thousands of firms have considered, and which many have adopted, is to establish a customer loyalty program. Examples of these schemes can be found in Japanese retailing, US airlines and hotels, French banks, British grocery stores, German car companies, Australian telecommunications, Italian fashion stores, American universities, and many other areas. Typically these programs offer financial and relationship rewards to customers, and in some instances benefits also accrue to third-parties such as charities1 . Two aims of customer loyalty programs stand out. One is to increase sales revenues by raising purchase/usage levels, and/or increasing the range of products bought from the supplierA second aim is more defensive, by building a closer bond between the brand and current customers it is hoped to maintain the current customer base. The popularity of these programs is based on the argument that profits can be increased significantly by achieving either of these aims 2 . While loyalty programs can have many other peripheral goals – such as furthering cross-selling, creating databases, aiding trade relations, assisting brand PR, establishing alliances, etc. – we do not assess these goals in this paper. But, how effective are these programs in enhancing the number, the loyalty, and/or the sales from customers? And, are they likely to be profitable when fully costed? To answer these questions we first discuss what is meant by the term ‘customer loyalty’. A review of the literature reveals that this task is not straightforward – generally people have in mind one of three different models (Section 2).
We consider whether these models are based on
competing or complementary theories (Section 3). This provides a platform for thinking
4
about a loyalty continuum (Section 4). We show that it is crucial to define and understand customer loyalty if the demand-side benefits of loyalty programs are to be properly evaluated. Next, drawing on these conceptualizations, we review the goals, successes and failings of loyalty programs (Section 5). We show that, at one extreme are programs for niche products that presume customers are committed to ‘a favorite brand’. At the other extreme there are promotional programs that cater to the divided loyalty of their customers. In between, and widely represented across many different products and services, are loyalty programs that are best described as “for the brands people already buy”. Future prospects are discussed briefly (Section 6). The focus of this paper is on established repeat-purchase markets where there is direct competition between branded products and services. These markets include most packaged goods, personal services such as banking and travel agents, food and beverages, hotels, transport, retail, OTC pharmaceuticals, basic cosmetics, and media.
They are hugely
important in terms of the share of disposable consumer income for which they account, and they have been the focus of much research.
2. Customer Loyalty At a very general level, loyalty is something that consumers may exhibit to brands, services, stores, product categories (e.g., cigarettes), and activities (e.g., swimming). Here we use the term customer loyalty as opposed to brand loyalty; this is to emphasize that loyalty is a feature of people, rather than something inherent in brands. Unfortunately there is no universally agreed definition (Jacoby and Chestnut 1978; Dick and Basu 1994; Oliver 1999). Instead, there are three popular conceptualizations: loyalty as primarily an attitude that sometimes leads to a relationship with the brand (Model 1); loyalty mainly expressed in terms of revealed behavior (i.e., the pattern of past purchases) (Model 2); and buying moderated by the individual’s characteristics, circumstances, and/or the purchase situation (Model 3) (see Figure 1).
5
(a) Loyalty as primarily an attitude that sometimes leads to a relationship with the brand (Model 1) Many researchers and consultants argue that there must be strong ‘attitudinal commitment’ to a brand for true loyalty to exist (Day 1969; Jacoby and Chustnut 1978; Foxall and Goldsmith 1994; Mellens et. al. 1996; Reichheld 1996). This is seen as taking the form of a consistently favorable set of stated beliefs towards the brand purchased. These attitudes may be measured by asking how much people say they like the brand, feel committed to it, will recommend it to others, and have positive beliefs and feelings about it – relative to competing brands (Dick and Basu, 1994). The strength of these attitudes is the key predictor of a brand’s purchase and repeat patronage. This is what Oliver (1997) has in mind when he defines customer loyalty as: “A deeply held commitment to rebuy or repatronize a preferred product/service consistently in the future, thereby causing repetitive same-brand or same brand-set purchasing despite situational influences and marketing efforts having the potential to cause switching behavior” (p. 392). In the fields of advertising and brand equity research this model receives much conceptual support (e.g., Aaker 1996; de Chernatony and McDonald 1998; Keller 1998). The approach also appeals to many practitioners in advertising and brand management because it is empathetic with the search for strategies to enhance the strength of consumers’ attitudes towards a brand. Moreover, there is some evidence to suggest it is a profitable strategy. Ahluwalia, Unnava and Burnkrant (1999) have shown that attitudinally-loyal customers are much less susceptible to negative information about the brand than non-loyal customers. Also, where loyalty to a brand is increased, the revenue-stream from loyal customers becomes more predictable and can become considerable over time – as analyses of cases such as Federal Express, Pizza Hut franchises, and Cadillac dealerships have shown (Gremler and Brown 1999). An extension of the ‘attitudes define loyalty’ perspective is to suggest that consumers form relationships with some of their brands. A good example of this perspective is provided
6
by Fournier (1998), who sees loyalty as a committed and affect-laden partnership between consumers and brands. It is a partnership that will be even stronger when supported by other members of a household or buying group, and where consumption is associated with community membership or identity. Examples in support of this argument include Skoal smokeless tobacco among some North American cowboys, loyalty to particular European soccer teams (Arnould, Price and Zinkhan 2002), the Beanie Babies craze (Morris and Martin 2000), Jeep brandfests (McAlexander, Schouten and Koenig 2002), and the classic case of Harley-Davidson bikers (Schouten and McAlexander 1995). Despite the psychological and sociological richness of the ‘attitudes drive behavior’ and ‘relationship’ approaches to understanding customer loyalty, these conceptualizations of loyalty are not without their critics (e.g., Dowling 2002).
They are thought to be less
applicable for understanding the buying of low-risk, frequently-purchased brands, or when impulse buying or variety seeking is undertaken, than for important or risky decisions (Dabholkar 1999). Also, as Oliver (1999) has noted, there is little systematic empirical research to corroborate or refute this perspective of customer loyalty. The examples above are isolated cases, often cited as illustrative of the revenue-effects that might be achieved, rather than the profit impacts that have been achieved.
(b) Loyalty mainly expressed in terms of revealed behavior (Model 2) Paradoxically, Model 2 is arguably the most controversial but the best supported by data. The controversy comes about because loyalty in this model is defined mainly with reference to the pattern of past purchases with only secondary regard to underlying consumer motivations or commitment to the brand (Ehrenberg 1988; Fader and Hardie 1996; Kahn, Kalwani and Morrison 1988; Massy, Montgomery and Morrison 1970). Researchers have gathered impressive amounts of data about these purchase patterns over many years - across dozens of product categories and for many diverse countries (Uncles et al. 1994). They have found that few consumers are ‘monogamous’ (100% loyal) or ‘promiscuous’ (no loyalty to any brand). Rather, most people are ‘polygamous’ (i.e., loyal to a portfolio of brands in a 7
product category). From this perspective, loyalty is defined as “an ongoing propensity to buy the brand, usually as one of several” (Ehrenberg and Scriven, 1999). These researchers tend to adopt a market focus as opposed to an individual focus (e.g., key performance measures are brand shares, penetration, average purchase frequencies, repeat-buying – for a defined period). Stochastic modeling techniques describe the observed patterns of customer buying. Given these descriptions, loyalty is inferred to operate in the following manner. Through trial and error, a brand that provides a satisfactory experience is chosen.
Loyalty to the brand (measured by repeat purchase) is the result of repeated
satisfaction that in turn leads to weak commitment. The consumer buys the same brand again, not because of any strongly-held prior attitude or deeply-held commitment, but because it is not worth the time and trouble to search for an alternative. If the usual brand is out of stock or unavailable for some reason, then another functionally similar (or substitutable) brand (from the portfolio) will be purchased (e.g., East 1997; Ehrenberg, Barnard and Scriven 1997; Ehrenberg, Uncles and Goodhardt 2003). There is little reason to spend much effort weighing up the alternatives when all are likely to be satisfactory. However, over repeated purchases a weak commitment to the (limited) number of brands bought in a product category can form. All these studies are grounded in considerable amounts of market research data and analysis. But, despite the weight of empirical evidence, controversy persists. Those who subscribe to the ‘attitudes drive behavior’ and ‘relationship’ approaches expressly rule-out revealed behavior as a dominant measure of loyalty. That, they argue, may merely reflect happenstance. Even combined measures of revealed behavior and satisfaction may not probe deeply enough for us to be sure there is true loyalty (Arnould, Price and Zinkhan 2002; Oliver 1999).
(c) Buying moderated by the individual’s characteristics, circumstances, and/or the purchase situation (Model 3) Proponents of Model 3, the contingency approach, argue that the best conceptualization of loyalty is to allow the relationship between attitude and behavior to be 8
moderated by contingency variables such as the individual’s current circumstances, their characteristics, and/or the purchase situation faced3 . That is, a strong attitude toward a brand may provide only a weak prediction of whether or not the brand will be bought on the next purchase occasion because any number of factors may co-determine which brand(s) are deemed to be desirable (Belk 1974, 1975; Blackwell et al. 1999; Fazio and Zanna 1981). Individual circumstances include budget effects (e.g., the desired brand is too expensive), and time pressure (e.g., the need to buy any brand in the category at the next available opportunity). Individual characteristics are reflected in the desire for variety, habit, the need to conform, the tolerance for risk, etc. Purchase situation effects include product availability, promotions/deals, the particular use occasion (e.g., gift, personal use, family use), etc. A three-factor model emerges, based on antecedents (including weak prior attitudes and characteristics of the consumer), contingency factors (including type of use occasion and the purchase situation), and consequences (up-dated attitudes, intentions and the actual purchase behavior). The difference between this contingency perspective and the attitude perspective is that the contingency variables are elevated from the status of loyalty inhibitors in Model 1 to loyalty co-determinants in Model 3. For example, in Oliver’s (1997, 1999) definition cited earlier, attributes of the individual and the purchase situation are conceptualized as ‘nuisance’ variables that inhibit the natural evolution of customer loyalty, whereas in the contingency model these variables are seen as playing a primary and inescapable role in explaining the observed patterns of purchase behavior. This is even more evident where attitudes are weakly held. Here it is repeated satisfaction and weak commitment that together with other relevant contingency variables co-determine future brand choices. [Figure 1 about here.] Figure 1 poses two questions about customer loyalty. First, do the three models suggest different courses of action for marketing managers – especially in the context of developing and using customer loyalty programs? Second, is it possible to combine these three approaches to develop a more unified concept of customer loyalty and therefore to 9
provide a more complete guide for program management? These questions are addressed in Sections 3 and 4 respectively.
3. Competing or Complementary Theories of Customer Loyalty? Depending on the model one adopts, the implications for practice can be significantly different. For example, advocates of the attitude approach (Model 1) aim to increase sales by enhancing beliefs about the brand and strengthening the emotional commitment of customers to their brand. Moving customers up a ‘loyalty ladder’ through image-based or persuasive advertising and personal service (recovery) programs are frequently used tactics (Brown 2000; White and Schneider 1998).
Loyalty programs are also designed to strengthen
commitment and create velvet handcuffs to bond the customer to the brand. This way of thinking has become commonplace in communications, branding and CRM textbooks. Alternatively, advocates of the behavioral focus (Model 2) suggest that most consumers have split-loyalty portfolios of habitually-bought brands. Here it is assumed that consumers tend to view advertising and other forms of marketing communication more as publicity that sustains awareness and offers reinforcement, rather than as highly persuasive information that fundamentally changes their attitudes and/or levels of commitment (Ehrenberg, Barnard and Scriven 1998). While these customers may participate in loyalty programs, they are also thought to be less influenced by these programs than the advocates of Model 1 assume (Dowling and Uncles 1997). Managers who adopt this approach try to maintain their share of category sales by matching competitor initiatives and avoiding supply shortages, and achieve growth via increased market penetration (by, for example, securing wider distribution). Under these circumstances, a loyalty program might be launched for mainly defensive purposes, in a bid to match competitors or as a publicity generating gesture, but with no expectation of dramatic changes in customer attitudes and behavior. Advocates of the contingency approach (Model 3) adopt a slightly different approach. They emphasize what might seem to be prosaic factors – such as avoiding stock-outs,
10
extending opening hours, offering the appropriate assortment mix (to cater for various usage situations and variety seeking), having 24-hour call centers, providing on-line access, etc. They also often use price promotions, deals and special offers to attract the customers of competitor brands (e.g., as with gasoline retailers). Here the potential for loyalty programs to impact demand is very limited. Indeed, the product or service provider is likely to gain greater loyalty by responding directly to the contingent factors, and an image-building program may run counter to such a goal. Nevertheless, loyalty programs have been launched by companies who operate in markets with very little product/service differentiation – many of these can be seen as continuous promotional programs (Palmer and Beggs 1997). For management, the choice of theory becomes important when brands competing in a category are functionally similar and marketing budgets are not big enough to fund the tactics implied by all three models.
Even where budgets are large – allowing for the
simultaneous expansion of the sales base, advertising to encourage more positive beliefs about the brand, and tactical promotions – the need for strategic focus may preclude one or two of these options. For instance, as noted above, the launch of a loyalty program may run counter to the creation of a price-competitive image (particularly if it is perceived as an unnecessary expense that inhibits price-cuts from being passed on to customers). In the next section the conceptual implications of these different approaches to customer loyalty are explored.
4. Conceptual Implications of the Different Approaches to Customer Loyalty In Figure 2 we use the three models of loyalty to introduce the notion of a loyalty continuum. The anchor points are Customer Brand Commitment (CBC) and Customer Brand Buying (CBB), with Customer Brand Acceptance (CBA) occupying the densely populated middle ground. All these loyalty patterns profile customers, not brands per se; that is, consumers are distributed across the curves with respect to their loyalty to a brand. For example, most customers may accept a number of airlines, while a few customers may be
11
committed to one or two airlines, and some others may buy purely on the price/route combination. These people’s air travel schedules may result in them having quite a few brands in their portfolio. Nevertheless, the nature of the market in which customers buy and brands compete will govern what is normally observed – thus, in highly competitive repeatpurchase markets acceptance is to be expected more often than the other models. We elaborate below. [Figure 2 about here.]
(a) Customer Brand Acceptance (CBA) The concept of Customer Brand Acceptance (CBA) is the base case of customer loyalty in competitive repeat-purchase markets. It draws heavily on Model 2, but also brings together some elements of Models 1 and 3. The contribution of Model 2 is that customers exhibit loyalty to a number of brands because there is little reason to develop exclusive attitudinal loyalty to any one of the brands purchased. A prime reason for this is that a proliferation of brands in most markets has destroyed one of the key reasons for exclusive loyalty, namely brand distinctiveness.
Weilbacher (1993) and Ehrenberg, Barnard and
Scriven (1997) argue that in many product categories, both the functional and the perceived differences among competing brands are small, so it is not surprising that customers perceive few critical and meaningful differences across competing brands. For many of these brands the advertising messages and loyalty programs are fundamentally similar too (compare the similar car hire adverts in travel magazines or the near-identical benefits of alternative airline frequent-flier programs). Figure 3 summarizes the concept of Customer Brand Acceptance in terms of the familiar five-stage model of consumer choice. Need arousal is included as a trigger to the purchase process – but this operates mainly on product category decisions, not brand-based ones. For instance, because of a desire to stay sober the need is for low-alcohol beer, but not necessarily for any particular brand of low-alcohol beer. Since this is a model of ongoing customer brand acceptance for frequently-purchased products, the (external) information 12
search and evaluation stages are assumed to have been completed after the initial one or two purchases in the category, and so are not explicitly included in the diagram. Choice among the functionally equivalent alternatives will reflect the accessibility, availability and conspicuousness of a brand at the point of purchase. Most likely, this will be seen as a set of acceptable brands that are ordered as first favorite, second favorite, third favorite, and so forth (Hammond 1997)4 . Typically, the relative likelihood of buying each brand will endure over successive purchase cycles, assuming the brands remain functionally adequate and accessible. Satisfaction with past purchases, and any consequential habit formation, explain most of a person’s ongoing propensity to buy one or a number of acceptable brands. Unexpected purchase situation circumstances (e.g., an existing brand being on sale) may influence the actual brand chosen on a specific purchase occasion (drawing on Model 3). The introduction of new brands or the reformulation of current brands may alter the purchase propensities, although the aggregate impact on short to medium-term brand loyalty is likely to be marginal. This is not to suggest that attitudes will not form towards these brands over time (Model 1), but they will be of secondary importance to the functional adequacy of the brand. Indeed, for the markets which are the focus here, research shows that these beliefs may simply be a playback of the message content of the brand’s advertising or publicity – that is, simple learning (Barwise and Ehrenberg 1985; Castleberry et. al. 1994). This can be seen in the very similar attitudes reported for descriptive attribute beliefs (e.g., “Volvos are safe”, “United Airlines is friendly”, “Woolworths offers fresh food”) by both brand users and nonusers (Dall’Olmo Riley, et al. 1997; Hoek, et al. 2000). Furthermore, the discussion of CBA suggests that having a favorable set of beliefs about one brand does not preclude having an equally favorable set of beliefs about other functionally similar brands in the category – and their almost identical loyalty programs (as is the case with many airline and retailer programs). For customers who are buying on a routine and mundane basis, it is not necessarily important to have a set of strong value-laden beliefs toward the brands that are purchased, as long as these brands are believed to “do the job”. 13
Research suggests that to the extent that a customer does express a consistently favorable attitude about a brand, it is more likely to be based on frequent satisfied use than on valueladen beliefs (Dall’Olmo Riley, et al. 1997). [Figure 3 about here.]
(b) Customer Brand Commitment (CBC) The first exception to CBA concerns those consumers who value psychological and social value more than function. This is easiest to see when these consumers are buying highidentity products (luxury goods, expensive cosmetics, etc.) and thinking of life-choices (education, sporting allegiances, etc.). Here there may be a brand component that drives choice and commitment for a significant number of customers, especially the initial adoption of some distinctive brands such as the Apple Macintosh, the Sony Walkman and HarleyDavidson motorbikes. We label this Customer Brand Commitment (CBC). In this situation, attitudes, values and social norms are seen as having a major influence, and the consumer can develop a relationship with the brand – in keeping with Model 1. Because these relationships are defined in the consumer’s head, they may help to differentiate one brand from another and they may support a price premium for that brand (Kapferer 1999). It is presumed consumers have a consistently favorable set of stated beliefs towards the brand purchased. However, none of this is guaranteed – especially when the focus is on frequentlybought brands. First, even for cases where the level of consumer involvement is high, differentiation among brands may be relatively low (such as with most airlines and hotel chains) – resulting in the type of behavior best described by consumer brand acceptance. For example, frequent flyers tend to use a number of different airlines; research on international travelers indicates that these people are typically members of multiple frequent-flyer programs and therefore show multi-brand loyalty to both the airlines and their programs (OAG 1998). It is mainly the infrequent flyers who are loyal to a single frequent-flyer program, but invariably, these are the less profitable customers. In most markets, the socio-
14
psychological elements of competing brands may in fact offer limited scope for creating meaningful differentiation. Second, when a brand is designed to have a distinct and unique personality, it does not mean customers will recognize and value this. Likewise, a manager may want to create a meaningful relationship between the brand and the customer, but customers do not necessarily desire this or reciprocate (Fournier, Dobscha and Mick 1998; Hart et al. 1999; Horne and Worthington 1999). When the type of loyalty is defined by the customer, it means that the same brand may be the object of commitment for one person but merely acceptable to another. Third, even where a relationship develops, it may not be the only one in a particular product category. For example, Fournier and Yao (1997) quote instances of customers having ‘compartmentalized friendships’ with different brands of coffee - perhaps Starbucks in the morning and Folgers in the afternoon. Moreover, with CBC, while the non-functional sources of value may be strong, they will not eliminate the need for the brand to “do the job”. HarleyDavidson, one of the strongest personality-relationship brands, was forced to instigate a quality improvement program to save the brand from Japanese competition.
(c ) Customer Brand Buying (CBB) The second exception to CBA concerns those consumers who exhibit very low levels of loyalty. Their choices are shaped by considerations of immediate availability, price, promotions, etc., and – at most – weak attitudes (e.g., users of an on-line travel agency may express liking for it because it obtains for them best price airfares). The concept of CBB is closely allied to Model 3, where contingencies are the co-determinants of choice, and not simply nuisance factors. In summary, our contention is that CBC and CBB are the exceptions rather than the rule in most repeat-purchase markets. One way to see this is as a sampling problem (Figure 2). Consider the example of car rental: if we were to draw from a large sample of the population, most customers of Avis or Hertz would be characterized by CBA, and only a few 15
by CBC (committed to my Hertz) or CBB (renting from literally any car hire firm that happened to be discounted at the time of purchase). Some researchers however have used highly selective sampling to highlight the exceptions and thus convey a very different impression of the relative importance of CBA, CBC and CBB in repeat-purchase markets (a point previously noted by Uncles and Laurent 1997). A distinction must be drawn between the loyalty of some customers to some brands, and the loyalty of most customers to most brands. Our review of customer loyalty provides the necessary basis from which to evaluate the aims and potential commercial effectiveness of loyalty programs – at least in terms of the customer-related (demand-side) issues.
5. Implications for the Management of Customer Loyalty Programs What gives poignancy to the concept of customer loyalty is the supposed justification it gives for managers to spend millions of dollars on CRM programs and the costly customer databases that support these. Customer loyalty programs are a current manifestation of this trend. Proponents tend to focus on the psychological bonding that eventuates from membership (a customer benefit), and the enhanced customer insights that can be gained from analyzing the program database (a firm benefit) (Brown 2000; Pearson 1996). Critics argue that the loyalty – both attitudinal and behavioral – for most customers is quite passive and resembles habit rather than serious commitment. Also, they argue that these programs are expensive to set up and maintain and that there is little or no evidence that any changes in behavior justify the expenditure (Dowling and Uncles 1997). These are strong claims and counter-claims, and to a large extent they rest on the different models of customer loyalty we have outlined above. Supporters of loyalty programs have in mind Model 1, where the program is seen to reinforce CBC-type outcomes. Or they envisage a combination of Models 3 and 1, where consumers with no loyalty (CBB-types) are converted into single-brand loyal CBC-types
16
because of the customer benefits of the program. Critics favor the multi-brand dividedloyalty model (Model 2), and assume most customers are CBA-types who are not strongly swayed by the program. In evaluating the aims and demand-side success of loyalty programs, we take account of these somewhat contradictory positions. We examine the issues from the perspective of (a) individual customers, (b) markets, and (c) touch on the contribution to profits of such schemes5.
(a) Loyalty Programs from an Individual’s Perspective Where the focus is on individual customers, loyalty programs can be seen as vehicles to increase single-brand loyalty, decrease price sensitivity, induce greater consumer resistance to counter offers or counter arguments (from advertising or sales-people), dampen the desire to consider alternative brands, encourage word-of-mouth support and endorsement, attract a larger pool of customers, and/or increase the amount of product bought. For instance, Bolton, Kannan and Bramlett (2000) found that members of the loyalty program of a financial services company were generally less sensitive than other customers to perceptions of lower service quality from their company and any price disadvantage relative to competitors. In keeping with this focus, the rhetoric of many consultants suggests that the aim of a loyalty program should be to create a bigger group of single-brand loyal customers (consistent with Model 1). But the review of customer loyalty presented in the previous section and the empirical evidence cited, suggests that this is both undesirable for most customers and unachievable for most firms (the implication of Figure 2). Indeed, if customers are already single-brand loyal, the scheme will only be sales effective if it can get these people to buy more of the brand. This is not easily achieved when so many single-brand loyal buyers are light users – in these circumstances they must be persuaded to buy more of the brand and more of the category or other categories offered by the same firm (Ehrenberg, Uncles and Goodhardt 2002). Something that is exceedingly hard if customers are not particularly motivated by the brand, category or program.
17
Significantly, in Bolton et al.’s study, 43% of respondents did not use their ‘loyaltybuilding’ credit card during the one-year study period and a further 36% used their card on fewer than six occasions – the program could not be described as particularly motivating. Similarly, Wright and Sparks (1999) found that as many as a fifth of retail loyalty cardholders did not make any use of their card over a three-month period. In general, enhancing the bond between customers and their brands and expecting that this will automatically stimulate more demand for the product category is not a sustainable outcome. Why? Because most people generally only buy what they need. At the other extreme, it is possible that a loyalty program could be offered to people who do not buy the target brand (but do buy from the category). If the program is sufficiently appealing, then it might entice customers to switch brands (creating a new group of singlebrand loyals, as in Model 1) or to include the brand in their repertoire of acceptable brands (making them polygamous loyals, as in Model 2). However, in this case the substantial appeal of the program can actually be its weakness. First, this course of action is likely to induce a quick counter-response from managers whose brands start to loose share. Second, when the program is attractive, customers may come to build a relationship with the program rather than the brand.
Then a large part of the brand’s equity becomes dependent on
something that might have little directly to do with the brand, as well as something that is vulnerable to competitive responses. In between these extremes, recall that over a number of purchases, most consumers buy more than one brand in the category. For these consumers there is some scope for bringing about a re-allocation of purchasing without demanding any fundamental shifts in attitudes or behavior toward the brands bought or the product category.
However, if
consumers have good reasons for being multi-brand loyal, then it is unrealistic for brand managers to expect them suddenly to become single-brand loyal. Empirically, consumers appear not to want to watch one television station, eat at one restaurant, patronize one hotel chain, drink one brand of wine, get all their business news from one magazine, go to one holiday destination, buy one brand of petrol, only attend one theatre, always shop at the same 18
bookstore, etc. Hence, it is a major challenge for brand managers to convince enough people to reduce their repertoire of brands such that the propensity to buy their particular brand increases enough to cover the full costs of the program. In these circumstances, the best way for customers to reallocate some of their category purchasing to a particular brand is for a program to address the underlying reasons for polygamy. Thus, program members might be given greater access to the brand, offered more variety, or helped to consolidate their purchases with fewer business providers/brands. The launch of the One World Alliance in international airlines can be seen as an example of this strategy, although the existence of competing alliances – notably the Star Alliance – shows the difficulty of devising a unique and well-differentiated program in highly competitive, repeat-purchase markets. It also shows how difficult it is to separate purely functional and economic benefits (more routes and flexible schedules in the case of airlines) from membership benefits (Driver 1999, Goh and Uncles 2003). An important component of many loyalty programs is the scope for cross-selling, in an attempt to increase share-of-wallet, rather than market share (Peppers and Rogers 1997). Loyalty-program members are encouraged to buy products they would not normally have bought from that provider. In essence, the loyalty program is seen as a brand extension aid. For example, through their respective programs, United tries to interest customers in car hire and hotels, while Tesco attempts to expose its Clubcard members to high-margin wines, financial services and electrical goods, as well as lower margin groceries. One issue is that many of the cross-selling opportunities are themselves in highly competitive markets – hotels, car hire, restaurants, financial services, etc. – and often these markets support other loyalty programs. The implication here is that only a truly exceptional program will change the purchasing behavior of customers to significantly increase sales revenues.
19
(b) Loyalty Programs from a Market Perspective At an aggregate level, repeat-purchase markets typically have a well-defined structure – namely, most brands exhibit a Double Jeopardy (DJ) effect whereby small brands have fewer buyers who buy them less often than big brands (Ehrenberg, Goodhardt and Barwise 1990). Whatever their market shares, it is to be expected that, for all brands, there will be some CBB and CBC buyers, and a majority of CBA buyers. This market structure gives rise to three strategies for enhancing the observed level of repeat-purchase or loyalty of a brand. A possible fourth strategy is considered too. The first strategy is to try to grow the size of the brand. This can be achieved by making the brand acceptable to a larger number of potential customers – in keeping with Figure 3 and the focus on CBA. Tactically, this means exposure at the point of purchase, offering greater perceived value, gaining wider distribution, suggesting more usage occasions, etc. The second strategy is to create a niche brand by aiming to keep the numbers of buyers relatively low but at the same time increasing the average amount bought by these buyers. This could be achieved by reducing the distribution coverage of the brand and using the money saved to better support/promote the brand to current customers. This strategy implies a higher proportion of behaviorally-loyal and committed buyers (CBCs) for the level of market share than predicted by the DJ effect. In its early years, the Body Shop was a successful niche brand. The third strategy is for a big brand to become a ‘super-loyalty brand’. These are brands that exhibit signs of strong commitment and that have higher than expected (using the DJ model) repeat-purchase (Fader and Schmittlein 1993) (i.e., an above-average number of CBCs at a high level of market share). During the early 1990s, icon-status Nike appeared to be such a super loyalty brand. A fourth strategy implied by the DJ effect is to exploit the desire of customers for change-of-pace. Here the penetration is higher and the repeat-purchase rate lower, than predicted by the DJ effect (Kahn et al. 1988). Some imported and premium beer brands fall 20
into this category, though the typical beer brand of this type is simply small. This is primarily a penetration effect and cannot be seen as loyalty building unless an organization offers a portfolio of these brands (in which case the portfolio can again be expected to conform to the DJ pattern). Successful instances of strategies two and three are uncommon (by definition, they are deviations from the norm). They already have above-average shares of loyal (CBC) customers, and therefore a loyalty program would have to be unusually effective to raise loyalty levels further (although it may help to maintain the deviation). The first strategy is more common – it implies offering better value than competitors and growing the size of the brand. The easiest, and most cost-effective role for a loyalty program here is to improve the accessibility, availability and conspicuousness of a brand (e.g., its top-of-mind awareness or salience). This can be achieved initially by advertising and publicizing the program; this provides something newsworthy to say about the brand. Thereafter, periodic communication with members can take place. If this is successful the program will help the brand to grow and repeat purchase will be a natural outcome – although subject to the DJ constraint. The consumer may even re-evaluate the brand, moving from some weakly-held attitudes to more strongly held ones. However, to expect that the program will be sufficiently powerful to create single-brand commitment from initial divided loyalty for enough people to cover its full costs is a considerable challenge. An example illustrates the point. Sharp and Sharp (1997) used consumer panel data and stochastic modeling to establish normal patterns of consumer repeat-purchasing as a benchmark and then looked for departures from these predictions as evidence of the impact of the Australasian FlyBuys loyalty program on creating excess loyalty. Two conclusions from this study are noteworthy. First, the authors state that they “do not observe the consistent finding of FlyBuys brands showing higher levels of average purchase frequency given their individual levels of penetration” (p. 479). Second, they find that: “Of the six loyalty program brands, only two showed substantial repeat-purchase loyalty deviations and both of these showed this deviation for non-members of the loyalty program as well as members” (p. 485). 21
Given that FlyBuys was a particularly large-scale and bold attempt to use a loyalty program to re-engineer patterns of repeat purchase, these results are not encouraging. From a market perspective, a major implication is to see whether loyalty programs have the potential to help grow the size (and thus sales revenue) of a typical brand – when used in combination with other marketing programs. This is precisely the approach adopted by the UK retailer Tesco. The Tesco Clubcard scheme is regarded as a financial success, especially in terms of cross-selling and up-selling, although this success is also closely linked to high-profile advertising, product range extension, and strategic developments in the management of Tesco (Broadbent 2000, East and Hogg 1997).
(c)
Loyalty Programs and Profitability Firms employing loyalty programs should expect them to be profitable. On the cost
side of the profit equation, accurate estimates are difficult to obtain – even within corporations. One reason for this is that marketing programs in general, and loyalty programs in particular, seldom are fully costed. There are establishment costs (often including new advertising and promotional activity), enrollment costs, IT hardware, database creation and maintenance costs, servicing costs, management costs, editorial and production costs of loyalty magazines, the direct costs of rewards, and the opportunity costs of spending money on a loyalty program instead of on other marketing initiatives (e.g., new product development). A formula for factoring-in some of these costs is provided by Niraj, Gupta and Narasimhan (2001). Many different types of information are available about the sales effects of loyalty programs. But interpreting this information is difficult: often there is too much of some types of sales information and too little of other types; the evidence from sales is contradictory; and much of the data is gathered from poorly designed studies. Turning to the first of these problems, a purported benefit of loyalty programs is that they provide vast amounts of data that allow both a better insight into customer behavior and greater efficiency in targeted marketing. This information provides the analytical basis for 22
Model 1 and CBC outcomes.
Typically, information is obtained on demographics and
lifestyle at the time of joining the program; subsequently, product purchasing and responses to targeted marketing initiatives are documented for each purchase occasion. In practice the danger is that rather too much of this information is acquired. A scheme with millions of active members (such as those run by national grocery chains, multinational automobile companies, banks and airlines) gathers more EPOS data than it can usefully analyze or use for targeting purposes. Secondly, few of these programs collect data about the complete customer experience or the portfolio of brands bought (i.e., little information on either decision-making or total category expenditure). Nor do they have much to say about the total market and competitor marketing activity (e.g., non-customers are ignored). Yet, our discussion in the first part of the paper shows that this information is essential for anything other than a superficial and possibly inaccurate understanding of customer loyalty. Specifically, a thorough and accurate understanding of Models 2 and 3 requires data that is rarely available from loyalty-program databases. Here, the problem is that too little of the right kind of data is collected. A third area of concern is that data come from two sources – these often provide contradictory evidence. One source is the companies that have introduced these schemes. Not surprisingly, many suggest that their schemes are successful (publicly at least). The effects are reported as one or more of the following: increased sales of the target brand, higher levels of cross-selling, fewer customer defections, and more satisfied customers (e.g., Rayner 1998 reports on a number of UK schemes). However, researchers are beginning to question the accuracy of these effects (e.g., Reinartz and Kumar 2000, 2002). Notwithstanding the various initiatives that have been tried over the years, the empirical regularities of purchase incidence noted earlier (namely, DJ effects) have been robust to attempts by loyalty schemes to change them. One reason for this is that if a scheme looks as though it might be successful in increasing levels of accessibility, availability and conspicuousness, or in adding to the perceived value of the brand, it is quickly copied by competitors. The classic example of such
23
imitation is the airline frequent-flier programs – there are now no major airlines without such a scheme. When widespread copying happens, any benefit gained is likely to be ephemeral. A fourth area of concern is that evaluations on the sales effectiveness of loyalty programs are often based on a poor quasi-experimental design. When quantitative measures of effectiveness are developed they typically compare post-program levels of sales, customer retention, customer satisfaction, etc. with pre-program measures. Thus, there is often no control group (that is, no group subjected to the same new service regime, but without the ‘benefit’ of a loyalty program). Hence, program effects are confounded with the effects of other marketing initiatives. In Rayner’s (1998) review she reports that all the programs she describes were introduced as part of more wide-ranging marketing initiatives. For example, the oft-cited success of Tesco’s loyalty scheme is difficult to determine because it was introduced as part of a much broader program of new business development and store acquisition (East and Hogg 1997). This is not to criticize what was done; indeed, the paradox is that good business practice is based on an integrated approach to marketing that will most likely give rise to confounded measures of success.
Unfortunately, echoing Cook and
Campbell’s (1979) warning, we should be wary of making causal inferences from studies with weak experimental designs. A final potential problem is the choice of benchmark. The typical benchmark consists of conditions prevailing before the program was introduced. A much tougher test of the effectiveness of the loyalty program would be to compare the post-program results with what may have been achieved had the full costs of the program been used in another way – such as establishing a policy of everyday low prices, a new product introduction, more direct forms of brand extension, an increase in advertising spend, or improvements in the channels of distribution. The literature on good decision-making suggests that this type of comparison is likely to produce better management outcomes than evaluations based on one alternative versus the status quo (Hammond, Keeney and Raiffa 1999). The major managerial implication from looking at the potential profitability of loyalty programs is that, in most cases the jury is still out – but the early signs are not encouraging. 24
The jury is still out because much of the evidence relied on to support customer loyalty programs is not scientifically valid. The early signs are not encouraging because two of the better scientific studies, namely those of Sharp and Sharp (1997) and Reinartz and Kumar (2000) do not support the widespread use of customer loyalty programs. In fact, the Reinartz and Kumar study suggests a very weak association between customer profitability and longlife (loyal) customers. Thus managers should be cautious of claims extolling the costeffectiveness of these schemes.
6. Whither Loyalty Programs? We have discussed how loyalty programs might have an impact on customer loyalty in established repeat-purchase markets where there are directly competing branded products and services. In these markets Customer Brand Acceptance (CBA) is believed to describe the loyalty of most customers to most of the brands they buy. Customer Brand Commitment (CBC) and Customer Brand Buying (CBB) by some customers in some categories exists, but these are not necessarily widespread. The notion of brand acceptance draws heavily on behavioral definitions of loyalty, while allowing for weak attitude formation and the influence of major contingencies. It is within this context that most firms should assess their loyalty programs. Taking account of all considerations in this paper, the checklist in Table 1 is designed to help managers when considering the strategic and operational implications of starting or evaluating a loyalty program. [Table 1 about here.] Our review suggests that the demand-side success of many of these programs has been over-claimed by their advocates. This conclusion is based on two main observations: (a) established patterns of repeat-purchase behavior appear to be robust to the attempts of even large, well-financed programs to change them (e.g., major retail schemes or the airline frequent-flier programs), and (b) many high-profile programs are either quickly copied or induce a direct counter-response from competitors, thus nullifying much of their potential
25
impact (which, of course, is often the case with marketing and communications initiatives in established markets). Notwithstanding our cautious assessment, the fact remains that many loyalty programs are in operation and more are being introduced.
We conclude by briefly
considering why is there so much momentum behind these programs. First, it is possible to see loyalty programs as vehicles for maintaining customer loyalty (i.e., for keeping the brand in the customer’s repertoire) or for maintaining brand share (where the program works in combination with other valued enhancements, including product and service improvements). Here, rather than trying to induce single-brand loyalty from customers who previously have exhibited divided-brand loyalty, a more realistic aim is to build on existing levels of customer brand acceptance. If customers feel the need for affinity, or desire an explicit reward for their loyalty, they will join the programs of the brands they buy. The critical issue then is for the program to reinforce the value proposition of the parent brand – enhancing brand equity, not just building loyalty-program equity. The critical task for the program manager is to design a cost-effective scheme to achieve this aim. Second, another role for loyalty programs can be to improve levels of accessibility and market conspicuousness for a brand.
This can manifest itself as a more credible
proposition to retailers in order to secure more shelf-space and benefit from ‘retail push’. In other cases it may provide more opportunities to talk with customers and, perhaps, more opportunity to sell brand extensions to customers. In either case, the aim of the program is to get the brand into the customer’s set of acceptable brands. This, however, is not a substitute for the inherent functional, psychological and economic value designed into the brand, but rather it simply makes the brand easier to consider. If for some people the program provides additional emotional value, then this is a bonus. A third major factor is the me-too pressure to follow others who have embarked on this path.
Moreover, once these programs have been introduced, managers seem very
reluctant to cancel them – even if their claimed benefits are not being realized. For example, there are persistent rumors that many airlines would like to end their frequent-flier programs 26
if they could find an acceptable way to do this. So far, that goal has proved elusive, although the need to respond to deep discounting by companies such as Southwest and Virgin may force the hand of some operators. Just as there may have been first-mover advantages in creating a loyalty program, there also might be first-mover drawbacks from snatching away much heralded customer benefits.
However, there are instances of card-based loyalty
programs having been dropped (e.g., schemes operated by the DIY group Do-it-All and the grocery store Safeway in Britain, also Ford USA withdrew its credit card reward program). Despite all the problems surrounding loyalty programs, they are likely to be in the marketer’s toolkit for a long time yet, which is a powerful reason for carefully thinking through the issues discussed here.
27
Endnotes 1
Loyalty programs are schemes offering delayed, accumulating economic benefits to consumers who buy the brand. Usually this takes the form of points that can be exchanged for gifts, free product, or aspirational rewards such as air miles. Airline frequent-flier programs have been a prototype for many of the schemes. Affinity programs are a specific type of loyalty program. They are designed to enhance the emotional bond between customer and brand. Mechanisms are set up to enhance two-way communication in order for the customer to get to know the brand (or company that stands behind it) better, and for the company to learn more about the customer. No direct economic benefit is offered to the customer. Examples include telephone help lines, club memberships, alumni associations, newsletters, Web site ‘chat’ groups, etc. Hybrids also exist. For instance, where the focus is on enhancing the emotional bond between customer and brand, and a third-party (e.g., a charity) receives a financial benefit. Or the establishment of a club where consumers pay for membership, in return for access to special events and offers. This latter format is prevalent in countries like Germany where privacy and trading laws prohibit incentive-based schemes, e.g., Volkswagon Club, Swatch the Club, Mercedes Mastercard (Butscher 2002). 2
See for example, Reichheld (1996). While this generalization is often made by consultants, it takes no account of the company’s specific circumstances, particularly its target market, marketing strategy and cost structures (Niraj, Gupta and Narasimham 2001; Shaw 1998). Indeed, there is some evidence to suggest declining profitability from long-term customers in the context of catalogue buying in the US (Reinartz and Kumar 2000, 2002). 3
In this context, use of the word ‘loyalty’ is debatable. Some prefer to use the term ‘spurious loyalty’, in that any pattern of buying here is likely to result from the recurrence of contingent factors (e.g. Mellens, Dekimpe and Steenkamp 1996). It is pointed out that if the contingent factors are removed, buying may change. Nevertheless, we argue that Model 3 is the basis of brand acceptance and weak loyalty, neither of which should be regarded as spurious. 4
The main exception – where exclusive buying is observed – is among consumers who are light buyers of the product category and therefore of any brand in the category. Among these buyers, monogamous ‘loyalty’ may merely reflect a very limited number of purchase occasions (e.g., the infrequent holiday traveler versus the international business executive). Exclusive loyalty is also a function of the length of the observation period – in a short period most people will appear to be exclusively loyal because they have had so few opportunities to buy. 5
The bias is towards a consideration of customer issues, reflecting the customer-focus logic of marketing. Therefore, our comments about the contribution of loyalty schemes to profits focus largely on consumer-related demand-side issues. It would be useful for others to consider other perspectives. For example, some of these programs could be viewed as a form of indirect price cut that is desired by one segment of customers but is ultimately paid for by all customers. Programs that offer air miles as their reward could be viewed as the outcome of airlines wanting to sell excess capacity at a price greater than marginal cost. There are also broader issues of business policy and marketing strategy that need to be addressed. For instance, from the perspective of one business partner in a loyalty program, the success of the endeavor may depend on an ability to negotiate a particularly attractive deal with the other business partners – irrespective of whether the program has much impact on customer loyalty. We regard this as a very important, but separate, issue.
28
References Aaker, David (1996), Building Strong Brands. Boston: Free Press. Ahluwalia, Rohini, H. Rao Unnava and Robert E. Brunkrant (1999), “Towards Understanding the Value of a Loyal Customer: An Information-processing Perspective”, Cambridge, MA: Marketing Science Institute, Report No. 99-116. Arnould , Eric J., Linda L. Price and George M. Zinkhan (2002), Consumers. New York: McGraw-Hill. Barwise, Patrick and Andrew S. C. Ehrenberg (1985), “Consumer Beliefs and Brand Usage”, Journal of the Market Research Society , 27, 81-93. Belk, Russell W. (1974), “An Exploratory Assessment of Situational Effects in Buyer Behavior”, Journal of Marketing Research, 11, May, 156-163. Belk, Russell W. (1975), “Situational Variables and Consumer Behavior”, Journal of Consumer Research, 2, December, 157-164. Blackwell, S.A., S.L. Szeinbach, J.H. Barnes, D.W. Garner and V. Bush (1999), “The Antecedents of Customer Loyalty: An Empirical Investigation of the Role of Personal and Situational Aspects on Repurchase Decisions”, Journal of Services Research, 1, 4. Bolton, Ruth N., P. K. Kannan and Matthew D. Bramlet (2000), “Implications of Loyalty Program Membership and Service Experiences for Customer Retention and Value”, Journal of the Academy of Marketing Science, 28, 1, 95-108. Broadbent, Tim (2000), Advertising Works 11. Chapter 1: How Every Little Helps Was a Big Help to Tesco. Henley-on-Thames: World Advertising Research Center. Brown, Stanley A. (2000), Customer Relationship Management. Toronto: John Wiley & Sons. Butscher, Stephen (2002), Customer Clubs and Loyalty Programmes: A Practical Guide. (2nd edition). Oxford: Gower. Castleberry, Stephen B., Neil Barnard, Patrick Barwise, Andrew S. C. Ehrenberg and Francesca Dall’Olmo Riely (1994), “Individual Attitude Variations Over Time”, Journal of Marketing Management, 16, 153-162. Cook, Thomas D. and Donald T. Campbell (1979), Quasi-Experimentation. Chicago: Rand McNally. Dabholkar, Pratibha A. (1999), “Expectancy Value Models”, in P. E. Earl and S. Kemp (eds.), The Elgar Companion to Consumer Research and Economic Psychology. Cheltenham, UK: Edward Elgar, 201-208. Dall’Olmo Riely, Francesca, Andrew S. C. Ehrenberg, Stephen Castleberry, Patrick Barwise and Neil R. Barnard (1997), “The Variability of Attitudinal RepeatRates”, International Journal of Research in Marketing, 14, 5, 437-450. Day, George S. (1969), “A Two-Dimensional Concept of Brand Loyalty”, Journal of Advertising Research,9, 29-35. De Chernatony, Leslie and Malcolm McDonald (1998), Creating Powerful Brands. Oxford: Butterworth-Heinemann. Dick, Alan S. and Kunal Basu (1994), “Customer Loyalty: Toward and Integrated Conceptual Framework”, Journal of the Academy of Marketing Science, 22, 2, 99-113. Dowling, Grahame R. (2002), “Customer Relationship Management: In B2C Markets, Often Less Is More”, California Management Review, 44, 3, 87-104. Dowling, Grahame R. and Mark D. Uncles (1997), “Do Customer Loyalty Programs Really Work”, Sloan Management Review, 38, 4, 71-82. Driver, J.C. (1999), “Developments in Airline Marketing Practice”, Journal of Marketing Practice: Applied Marketing Science, 5, 5, 134-150.
29
East, Robert (1997) Consumer Behaviour: Advances and Applications in Marketing. Prentice Hall, London. East, Robert and Annik Hogg (1997), “The Anatomy of Conquest: Tesco versus Sainsbury”, Journal of Brand management, 5, 1, 53-60. Ehrenberg, Andrew S.C. (1988) Repeat-Buying: Facts, Theory and Application. New York: OUP; London: Griffin. Ehrenberg, Andrew S. C., Neil R. Barnard and John A. Scriven (1997), “Differentiation or Salience”, Journal of Advertising Research, 37, 6, 7-14. Ehrenberg, Andrew S. C., Neil R. Barnard and John A. Scriven (1998), “Advertising is Publicity not Persuasion”, R&D I Research Report. London: South Bank University. Ehrenberg, Andrew S. C., Gerald J. Goodhardt and Patric Barwise (1990), “Double Jeopardy Revisited”, Journal of Marketing, 54, 3, 82-91. Ehrenberg, Andrew S. C. and John A. Scriven (1999), “Brand Loyalty”, in P. E. Earl and S. Kemp (eds.), The Elgar Companion to Consumer Research and Economic Psychology. Cheltenham, UK: Edward Elgar, 53-63. Ehrenberg, Andrew S. C., Mark D. Uncles and Gerald Goodhardt (2003), “Understanding Brand Performance Measures: Using Dirichlet Benchmarks”, Journal of Business Research (forthcoming). Fader, Peter S. and Bruce Hardie (1996), “Modeling Consumer Choice Among SKUs”, Journal of Marketing Research, 33 (November), 442-452. Fader, Peter S. and David Schmittlein (1993), “Excess Behavioral Loyalty for HighShare Brands: Deviations from the Dirichlet Model for Repeat Purchasing”, Journal of Marketing Research, 30, 4, 478-493. Fazio, R.H. and Zanna, M.P. (1981), “Direct Experience and Attitude Behavior Consistency”, Advances in Experimental Social Psychology: 14, Academic Press, 161-202. Fournier, Susan (1998), “Consumers and Their Brands: Developing Relationship Theory in Consumer Research”, Journal of Consumer Research, 24, 4, 343-373. Fournier, Susan, Susan Dobscha and David Glen Mick (1998), “Preventing the Premature Death of Relationship Marketing”, Harvard Business Review, (January-February), pages?? Fournier, Susan and Julie L. Yao (1997), “Reviving Brand Loyalty: A Reconceptualization within the Framework of Consumer-Brand Relationships”, International Journal of Research in Marketing, 14, 5, 451-472. Foxall, Gordon R. and R. E. Goldsmith (1994), Consumer Psychology for Marketing. London: Routledge. Goh, Kevin and Mark Uncles (2003), “The Benefits of Airline Global Alliances: An Empirical Assessment of the Perceptions of Business Travelers”, Transportation Research (forthcoming). Gremer, Dwayne D. and Stephen W. Brown (1999), “The Loyalty Ripple Effect: Appreciating the Full Value of Customers”, International Journal of Service Industry Management, 10, 3, 271-293. Hammond, John S., Ralph Keeney and Howard Raiffa (1999), Smart Choices. Boston: Harvard Business School Press. Hammond, Kathy (1997), Brand Loyalty for Frequently Bought Goods. London Business School: Unpublished PhD thesis. Hart, Susan, Andrew Smith, Leigh Sparks and Nikolaos Tzokas (1999), “Are Loyalty Schemes a Manifestation of Relationship Marketing?” Journal of Marketing Management, 15, 6, 541-562. Hoek, Janet, Jason Dunnett, Malcolm Wright and Philip Gendall (2000), “Descriptive and Evaluative Attributes: What Relevance to Marketers?”, Journal of Product and Brand Management, 9, 6&7, 415-433.
30
Horne, Suzanne and Steve Worthington (1999), “The Affinity Credit Card Relationship: Can it Really be Mutually Beneficial? ”, Journal of Marketing Management,15, 7, 603-616. Jacoby, Jacob and Robert Chestnut (1978), Brand Loyalty Measurement and Management. New York: John Wiley & Sons. Kalakota, Ravi and Marcia Robinson (1999), e-Business. Reading, MA: Addison Wesley. Kahn, Barbara M., M. U. Kalwani and Donald G. Morrison (1988), “Niching versus Change-of-Pace Brands: Using Purchase Frequencies and Penetration Rates to Infer Brand Positionings”, Journal of Marketing Research, 25, 4, 384-390. Kapferer, Jean-Noël (1999), Strategic Brand Management. 2nd Edition. London: Kogan Page. Keller, Kevin L. (1998), Strategic Brand Management. Upper Saddle River, NJ: Prentice Hall. McAlexander, James H., John W. Schouten and Harold F. Koenig (2002), “Building brand community”, Journal of Marketing, 66 (1), 38-54. Massy, William F., Dave B. Montgomery and Don G. Morrison (1970), Stochastic Models of Buyer Behavior. Boston: MIT Press. Mellens, Michael, Marnik G. Dekimpe and Jan-Benedict E. B. Steenkamp (1996), “A Review of Brand-Loyalty Measures in Marketing”, Tijdschrift voor Economie en Management, 41, 4, 507-533. Morris, Rebecca J. and Charles L. Martin (2000), “Beanie Babies: A Case Study in the Engineering of a High-Involvement / Relationship-Prone Brand”, Journal of Product and Brand Management, 9, 2, 78-96. Niraj, Rakesh, Mahendra Gupta and Chakravarthi Narasimhan (2001), “Customer Profitability in a Supply Chain”, Journal of Marketing, 65, 3, 1-16. OAG Business Travel Lifestyle Survey (1998). Dunstable Bedfordshire, England: Official Airline Guides. Oliver, Richard L. (1997), Satisfaction: A Behavioral Perspective on the Consumer. New York: Irwin/McGraw-Hill. Oliver, Richard L. (1999), “Whence Customer Loyalty”, Journal of Marketing, 63, Special Issue, 33-44. Palmer, Adrian and Rosalind Beggs (1997), “Loyalty Programmes: Congruence of Market Structure and Success”, Proceedings of the Academy of Marketing Conference. Manchester: Manchester Metropolitan University. Pearson, Stewart (1996), Building Brands Directly. London: Macmillan Business. Peppers, Don and Martha Rogers (1997), Enterprise One to One: Tools for Competing in the Interactive Age. New York: Currency Doubleday. Rayner, Sue (1998), Customer Loyalty Schemes: Effective Implementation and Management (2nd edition). London: FT Retail and Consumer Publishing. Reichheld , Frederick F. (1996), The Loyalty Effect. Boston, MA: Harvard Business School Press. Reinartz, Werner J. and V. Kumar (2000), “On the Profitability of Long-Life Customers in a Noncontractual Setting: An Empirical Investigation and Implications for Marketing”, Journal of Marketing, 64, 4, 17-35. Reinartz, Werner J. and V. Kumar (2002), “The Mismanagement of Customer Loyalty”, Harvard Business Review, July, 86-94. Schouten, John W. and James H. McAlexander (1995), “Subcultures of Consumption: An Ethnography of the New Bikers”, Journal of Consumer Research, 22, 1, 43-61. Sharp, Byron and Anne Sharp (1997), “Loyalty Programs and Their Impact on Repeat-Purchase Loyalty Patterns”, International Journal of Research in Marketing, 14, 5, 473-486. Shaw, Robert (1998), Improving Marketing Effectiveness. London, UK: Economist Books. 31
Uncles, Mark D., Kathy Hammond, Andrew S. C. Ehrenberg and R. E. Davis ((1994), “A Replication Study of Two Brand-Loyalty Measures”, European Journal of Operational Research, 76, 2, 375-384. Uncles, Mark D. and Gilles Laurent (1997), “Editorial: Special Issue on Loyalty”, International Journal of Research in Marketing, 14 (5), 399-404. Weilbacher, W. M. (1993), Brand Marketing. Linconwood, IL: NTC Books, Ch. 7. White, Susan and Benjamin Schneider (1998), “Climbing the Commitment Ladder: The Impact on Customer Commitment of Disconfirmation of Service Expectations”, Cambridge, MA: Marketing Science Institute, Report No. 98-108. Wright, C and Leigh Sparks (1999), “Loyalty Saturation in Retailing: Exploring the End of Retail Loyalty Cards?”, International Journal of Retail and Distribution Management, 10, 429-439.
32
Table 1: Management Checklist Context • Established repeat-purchase markets • Where there is direct competition between branded products and services • Demand-side success is assumed to be of crucial importance Assessing Customer Loyalty • What underlying model of customer loyalty is being assumed – model 1, 2 or 3? (Section 2, Figure 1) • How is this assumption influencing thinking about loyalty-building initiatives – explicitly and implicitly? (Section 3) • Know your customers – what are the relative sizes and distinguishing features of the CBCs, CBAs and CBBs? (Section 4) • What steps have been taken to address the sampling problem? (Section 4, Figure 2) Assessing Loyalty Programs • What demand-side goals are there for the loyalty program – maintaining customer loyalty or enhancing it? How will these goals be set and assessed? (Section 1) • In general, will the program focus on the most profitable customers? What time frame is to be used to assess their profitability? (Section 4) • What is the appeal of the program for these customers? (Section 5 (a)) • How will the program be used in combination with other marketing activities? (Section 5 (a)) • Will these initiatives grow share and sales revenues? (Section 5 (b)) • Can the customer data be analyzed in useful ways? (Section 5 (c)) • Are the sales and cost data reliable? Is the evidence contradictory? Are you relying on studies with weak experimental designs? (Section 5 (c)) • What benchmarks have been chosen to assess the loyalty program and are these appropriate? (Section 5 (c)) • How will the overall profitability of the program be calculated? (Section 5 (c)) Assessing Major Traps • Are there, in fact, too few customers who will actually use it? Does the scheme have little appeal for customers? (Section 5 (a)) • Or, has the scheme been too indiscriminant – perhaps all three types of customers have joined because they see it as a (relatively) free option and/or a reward for their current purchase behavior? (Section 5 (a)) • Are customers more loyal to the scheme than the brand? Is this a problem and, if so, what can be done to rectify the problem? (Section 5 (a)) • What are the chances of a competitor retaliating to nullify the impact of the program? Have competitors already launched a counter-initiative (section 5 (b))? • Do you simply want to maintain the status quo (at a higher cost to all competitors)? (Section 5 (c)) • Is the need to service large numbers of members driving up running costs? (Section 5 (c))
33
Figure 1: Conceptualizations of Customer Loyalty Model 1 Strong attitudes & positive beliefs toward the brand
Attitudinal-loyalty to the brand (mainly seen as single-brand loyalty monogamy)
The influence of significant others, community membership & identity Model 2 Habitual revealed behavior
Behavioral-loyalty to the brands (mainly seen as divided-loyalty to a few brands - polygamy)
Satisfactory experience & weak commitment to brands
Model 3 Purchase situation, usage occasions & variety seeking
Co-determinants of buying brand(s) (mainly seen as weak loyalty or no loyalty - promiscuity)
An individual’s circumstances & characteristics
34
Figure 2: Summary of the Different Approaches to Customer Loyalty
Model of loyalty
1
2
3
(1 + 3)
Concept of loyalty
CBC (extreme)
CBA (norm)
CBB (extreme)
Distribution of customertypes across buyers of a product
CBB
Typical producttype
CBA
CBC
‘High-identity’ products
CBB
CBA
‘Branded’ brands
35
CBC
CBB
CBA
‘Commodified’ products
CBC
Figure 3: Customer Brand Acceptance (CBA)
Brand Attributes •little meaningful differentiation
Satisfaction &
•functional equivalence
Habit Formation
•weak strength
Need Arousal
Individual Customers
Set of
Purchase of
Acceptable Brands
Acceptable & Accessible
in a Category
Brands
Unexpected
•characteristics
Purchase Situation
•current circumstances
Factors
•significant others
36
