Do Customer Loyalty Programs Really Work?

Given the popularity of loyalty programs, they are surprisingly ineffective. To stand the best chance of success in tough market conditions, programs must enhance the overall value of the product or service and motivate loyal buyers to make their next purchase.

Reading Time: 31 min 



Customer loyalty schemes have attracted considerable interest as companies practice one of marketing’s most familiar strategies — “If you see a good idea, copy it.” Some banks have offered regular customers credit cards with a range of valuable benefits. In business-to-business markets, loyal customers have traditionally been treated better than those who buy on the spot market. Other well-known customer loyalty schemes are the frequent-flyer programs of the major airlines. These and other well-patronized programs were originally hailed as imaginative ways to instill and maintain loyalty, but, over the years, more and more doubt has been cast on them. Both the academic and trade press have criticized the programs with headlines such as: “A Failure in Competitive Strategy,” “War in the Air: The Scramble for Points Hits Turbulence,” “Frequent Flyer Offers Fail to Boost Loyalty.”1

Loyalty programs that seek to bond customers to a company or its products and services by offering an additional incentive pose an interesting dilemma. Although these schemes often attract widespread customer interest, they are difficult to support, using our current knowledge of competition and buyer behavior. This research suggests that most schemes do not fundamentally alter market structure. They might help to protect incumbents and might be regarded as a legitimate part of the marketer’s armory, but at the cost of increasing marketing expenditures.

Many senior managers now ask their marketing departments to measure the potential contribution of any program developed to implement loyalty marketing. Do these programs really create extra loyalty beyond that which is derived from the relative value of the product or service? Do they encourage customers to spend more? Or do they merely bribe a customer to buy again? In a competitive market, is it really feasible for every organization to increase customer loyalty by implementing a loyalty marketing program?

Underlying the increasing interest in these programs are some marketing managers’ widely held beliefs about customer loyalty:

  • Many customers want an involving relationship with the brands they buy.
  • A proportion of these buyers are loyal to the core and buy only one brand.
  • The hard-core, loyal buyers are a profitable group because there are many of them and they are heavy or frequent buyers.
  • It should be possible to reinforce these buyers’ loyalty and encourage them to be even more loyal.
  • With database technology, marketers can establish personalized dialogues with customers, resulting in more loyalty.

We encountered such beliefs in the rhetoric of relationship marketing, direct marketing, database marketing, and so-called “1:1 marketing.”2 However, in this paper, we compare these beliefs with growing research that challenges their accuracy. First, we describe the origins and aims of loyalty marketing. We then review research that examines empirical patterns of behavioral loyalty. We also consider psychological research on rewards, a major component of most loyalty programs. These two sections provide the context for our comments on frequent-buyer loyalty programs such as those of the major airlines. Finally, we suggest how to design a customer loyalty program that avoids the common traps undermining the effectiveness of many existing schemes.

Why Companies Introduce Customer Loyalty Programs

In the 1970s, European researchers studying business-to-business marketing discovered that suppliers who form close working relationships with their customers tend to have “better” customers.3 That is, the customers are more loyal to their suppliers and often give them a greater share of their business. The customers also reported having “better” suppliers. In short, it is a win-win arrangement.

Subsequent research claimed that loyal customers are more profitable to a firm.4 This profitability was thought to be generated by reduced servicing costs, less price sensitivity, increased spending, and favorable recommendations passed on to other potential customers by loyal buyers. Add to this the claim that it costs much more to entice a new customer to do business with a company than to get a current one to purchase again, and the strategy of gaining and maintaining loyalty seems like the source of sustainable competitive advantage.5

For a company to practice loyalty marketing, however, it must first know who its loyal customers are, which is a lot easier for many business-to-business marketers than for most consumer good marketers. As the number of customers increases, companies have to use database marketing and market research in the absence of personal knowledge. As computerized database technology becomes more sophisticated, so does a firm’s ability to monitor its customers’ behavior. Retailers and packaged-goods manufacturers use these techniques to efficiently target products and services and allocate marketing resources to achieve the maximum return.6 In business markets, companies conduct customer profitability analyses and calculate the lifetime value of their customers. However, the processes are complex and costly.

What began as a strategy for small businesses, business marketing, and catalogue selling has evolved into a new industry. Direct marketing practitioners are creating loyalty programs that tie the buyers of a wide range of consumer goods and services to a particular brand or supplier. But do customers for these products want a relationship with the supplying company? In a small business deal or for a risky purchase, the answer may be yes. For low-involvement products and the types of brands sold by, say, Nestlé, Procter & Gamble, Shell, or Unilever, it is unclear whether customers really want a relationship.

Press releases about launches of frequent-buyer or customer loyalty programs suggest that companies expect these schemes to achieve various objectives or practical measures of success. The most common objective is to retain existing customers and in so doing: (1) maintain sales levels, margins, and profits (a defensive outcome to protect the existing customer base), (2) increase the loyalty and potential value of existing customers (an offensive outcome to provide incremental increases in sales, margins, and profits), and (3) induce cross-product buying by existing customers (defensive or offensive). Usually these desired outcomes refer to specific target segments, for instance, heavy buyers or high-net-worth customers. The underlying belief is that a small percentage of customers generate most of a company’s sales and that these customers can be locked in forever. Companies often invoke the “80/20 law” in support of this viewpoint.

The 80/20 law states that about 80 percent of revenue typically comes from only 20 percent of customers. With such a skewed distribution of customers, it makes sense to concentrate most marketing resources on the 20 percent. The problem for loyalty programs is that the “best” 20 percent are not necessarily loyal buyers, especially in the sense of exclusive loyalty. As we describe in the next section, there is reliable empirical evidence to suggest that many or most heavy users are multibrand loyal for a wide range of products and services. That is, a company’s most profitable customers will probably be the competitors’ most profitable customers as well.

For companies with poor data about their customers, an additional benefit of customer loyalty programs is that members can identify themselves at the point of purchase or service delivery. Membership cards are a quick, efficient way for customers to signal that they deserve special attention. As a by-product, the company gains market research information —another espoused benefit of loyalty schemes. However, such a self-selected group is unlikely to represent all a company’s potential customers. Hence, it is only one source of market research information.

Companies openly discuss all these espoused benefits in public announcements. In practice, however, their decision to launch a program is often motivated as much by fears of competitive parity as anything else, which companies rarely state publicly. Therefore we add the following tactical motives to the previous list: (4) attempting to differentiate a parity brand, (5) preempting the entry of a new (parity) brand, and (6) preempting a competitor from introducing a similar loyalty scheme.

Most airlines and many leading companies, such as American Express, General Motors, Holiday Inn, Toyota/Lexus, John Deere, and Shell, have found enough merit in customer loyalty programs to implement such schemes. While they do not seek all the above benefits from each program, at least some benefits must occur for the scheme to pay off when the full cost is compared to other marketing alternatives. But, does a customer loyalty program offer a better return than an alternative such as a price cut, a move to everyday low pricing, increased advertising, or increasing distribution coverage? For many programs, the answer to this question lies in some interesting academic research that we review in the next two sections.

The Leaky Bucket Theory versus Polygamous Loyalty

Ehrenberg, in what he dubbed the “leaky bucket theory,” observes that many marketing strategies seem to be designed to replace “disloyal” customers who leak away with new ones in order to keep the sales level steady.7 But, while the marketing strategies of customer loyalty programs appear to be designed this way, is the underlying leaky bucket theory true?

To address this question and to establish norms of consumer behavior for different markets, Ehrenberg and his colleagues gathered data from various markets in Britain, continental Europe, Japan, and the United States over more than twenty years. Their research describes people’s purchase habits of products like coffee, ready-to-eat breakfast cereals, newspapers, aviation fuel, toothpaste, laundry detergents, gasoline, television programs, airline tickets, ethical drugs, and even repeat purchases of management development programs at business schools. Their focus was on people’s observed behavior, not their needs, motivations, personalities, or attitudes toward a brand.8 Gordon’s subsequent research provided qualitative support for the emergence of these behavioral norms.9

Typically, the researchers measured five types of behavior over certain time periods: (1) the percentage of consumers buying a brand, (2) the number of purchases per buyer, (3) the percentage who continue to buy the brand (repeat buyers), (4) the percentage who are 100 percent loyal, and (5) the percentage who also buy other specific brands (duplicate buyers). From these data, a company can determine how much of a buyer’s requirement for the category is met by a specific brand. Using statistical analysis, it can predict norms for each type of behavior in a competitive market.10

The empirical facts make it difficult to be enthusiastic about using a loyalty marketing program to create a large group of 100 percent loyal customers. The empirical record and the predictive norms show that only about 10 percent of buyers for many types of frequently purchased consumer goods are 100 percent loyal to a particular brand over a one-year period.11 Even in service situations, exclusive loyalty is confined to a small number of buyers. Moreover, 100 percent loyal buyers tend to be light buyers of the product or service.

The research of Ehrenberg and his colleagues indicates that in stationary markets, customer loyalty is divided among a number of brands, as if there were long-run propensities to buy brands A, B, and C, say, some 70 percent, 20 percent, and 10 percent of the time.12 Hence, invariably, customers do not buy only one brand. “Polygamous loyalty” better describes actual consumer behavior than either brand switching (a conscious once-and-for-all change of allegiance to another brand, as if propensities were 100 percent or zero) or promiscuity (the butterfly tendency to flit from brand to brand without any fixed allegiance, when there are no long-run propensities, only next-purchase probabilities).

Polygamous loyalty is readily apparent, for instance, in the soft drink and breakfast cereal markets but extends well beyond to car rentals, fast-food outlets, and business airline travel. It is also evident in customers’ multiple memberships in loyalty schemes. For example, surveys of European business airline travelers show that more than 80 percent are members of more than one airline loyalty scheme. In 1993, the average membership of airline loyalty clubs was 3.1 per traveler, and the figure seems to be rising.13

Many reasons for the generalized patterns of divided loyalty or polygamy are straightforward. For example, people buy different brands for different occasions or for variety. Or the brand may have been the only one in stock or may have offered better value because of a special deal. Most buyers who change brands are not lost forever or disloyal; usually, they prefer to buy a number of brands rather than one particular brand.

Given the amount of research that supports these patterns of buyer behavior, it seems unlikely that a loyalty program could fundamentally alter this behavior, especially in established, competitive markets (where copycat responses are most likely). Even a path-breaking scheme may alter only short-run probabilities. Once the market has settled down again, or a competitor has launched a similar scheme, patterns of divided loyalty reemerge. Then the issue is whether the longer-run propensities settle at old or new levels.

In a market in which competitive response occurs quickly, the market is expected to settle at its old levels. For example, at the beginning of 1995, there were no national loyalty schemes in the British grocery market, but once one retailer broke rank, all others followed within months. Despite the amount of loyalty building, market shares have been reasonably steady. Much the same had happened ten years earlier when American Airlines launched the first frequent-flyer scheme; within weeks, other carriers began to follow suit.14

When aggregated across many product and service markets, these research results suggest that the marketing mix of a brand (its product or service formulation, price, promotion, and distribution) determines its market share, and, once this settles down, the level of brand loyalty is strongly correlated with market share. Consequently, although marketers might give brands some fancy names, it is better to think about them as either big or small, rather than strong or weak in terms of loyalty (see Figure 1).15

A secondary effect is that big brands tend to have more buyers, and somewhat more of these buy slightly more frequently. Equally, small brands suffer “double jeopardy” in that there tend to be fewer buyers who buy the brand less frequently.16 If a loyal buyer is one who again purchases the brand frequently, the bigger the brand, the larger the number of more, loyal buyers. Add to this the empirical fact that some large-share brands may have a slightly higher proportion of heavy users and loyal buyers than predicted by the double jeopardy phenomenon, and there is more bad news for loyalty marketing programs designed to rescue small brands.

Most brands lie along the double jeopardy line (see Figure 1).17 However, there may also be big brands that exhibit some signs of super loyalty; i.e., they have more frequent buyers than double jeopardy would predict. At the other end of the scale, a niche brand is small, with a relatively higher proportion of buyers who are more loyal than predicted by double jeopardy. In its early days, the ecological Body Shop was a good example of a niche brand, but as it increased its distribution and appeal, it became a “normal” brand. As with most so-called niche brands, those that survive and prosper end up along the double jeopardy line, and they cease to be a niche.

The change-of-pace brand is one with a higher than expected market share but a less than expected proportion of loyal buyers. Many low-alcohol beers fit this description; people may buy them only when they are in a special situation such as before driving a car. Or they may have a premium beer as a predinner drink in a restaurant. The significant point is that super-loyalty, niche, and change-of-pace brands are much less common than big or small double jeopardy brands. Any of these brands can be profitable, because profit is determined by margin, not by the type or size of the brand.

A more effective way to grow a brand’s market share or become big is to get more people to buy the brand rather than try to get current customers to buy it more often. Fader and Schmittlein’s research suggests that one of the most effective ways to get more buyers is to gain more distribution outlets. It’s as simple as that.18 Leading global players like British Airways, Coca-Cola, and Ford are only too aware of this, which is why they strive so hard to protect their air routes, distribution channels, and dealerships. Loyalty programs and other marketing tactics such as price cuts and promotions are effective only in the long term, to the extent that they entice more distributors to stock the brand or build presence in the marketplace.

In summary, research implies that, in many established, competitive markets, purchasing of products and services is characterized by a number of empirical regularities. Given that these regularities are so widespread, it is difficult to increase brand loyalty above the market norms with an easy-to-replicate “add on” like a customer loyalty program. We do not deny that companies can have a short-term lucky break, or that they may be forced to act because of competition. But, for any customer loyalty program to be as effective as possible, given the prevailing competitive conditions, it must leverage the brand’s value proposition in the eyes of customers (that is, the balance of benefits relative to price).

Linking Customer Rewards to Loyalty Programs

Products and services provide functional, economic, and psychological benefits or solutions to buyers. These are the prime sources of customer value. As research suggests, the relative amount of customer value and distribution coverage drives market share and the number of loyal customers that a brand will acquire. The rewards that many customer loyalty programs offer are designed to alter this relationship.

To explore loyalty programs’ potential to alter normal patterns of behavior, we need to examine three psychological effects: (1) the extent to which loyalty is to the brand (a direct effect) or to the program (an indirect effect), (2) how buyers value the rewards offered, and (3) the effect of timing. We consider each in turn.

· Direct or Indirect Effect.

Rothschild and Gaidis have used behavioral learning theory to suggest that the type of incentives that many customer loyalty schemes offer may induce loyalty to the program (deal loyalty) rather than to the core product or service (brand loyalty).19 The extent to which this is desirable depends on the buyer’s level of involvement with the product. For many low-involvement products, the incentive and not the product can become the primary reward, especially if the incentive is exotic and out of proportion to the money spent. This might create a point of product differentiation, but once the incentive is taken away, the prime reason for purchase disappears. Many gasoline company loyalty schemes are caught in this trap, locking them onto a treadmill of continuous promotions.

However, for high-involvement products and services, which are typically accompanied by a small incentive, the product and not the incentive is the primary reward. For example, in the General Motors rebate scheme (the GM card), which allows participants to build up savings toward the cost of a new GM car, the car and not the accumulation of a discount is paramount. This seems to be more desirable because the creation of a point of difference reinforces the longer-term value proposition of the product itself.

In this way, we can classify customer loyalty programs by whether their explicit rewards directly support the value proposition of the product or service offered to customers (e.g., the GM card), or whether the rewards are designed to motivate loyalty by a more indirect route (e.g., free air travel from gasoline retailers). We suggest that loyalty programs that directly support the value proposition and positioning of the target product better fit the goals of loyalty marketing.

· How Buyers Value Rewards.

Psychologists have long been interested in the role of rewards in behavior modification and learning. They have developed numerous cognitive-learning theories that give insight into how the rewards of customer loyalty programs might help to achieve loyalty to the product rather than the program. O’Brien and Jones suggest five elements that combine to determine a program’s value: (1) the cash value of the redemption rewards (e.g., the ratio of the cost of an airline ticket to the dollar purchases necessary to accumulate frequent-flyer points), (2) the range of choice of these rewards (e.g., choice of flight destinations), (3) the aspirational value of the rewards (e.g., exotic free travel is more desirable than a cash-back offer), (4) the perceived likelihood of achieving the rewards (e.g., how many points are required to qualify for a flight), and (5) the scheme’s ease of use.20 To this, we add the psychological benefits of belonging to the program and accumulating points.

· Timing.

The potential of a loyalty program to attract members depends not only on the value of its rewards but also on when the rewards are available. Research in psychology suggests that when a loyalty program’s redemption rewards are delayed, they are less powerful.21 Many accumulating benefit programs, such as frequent-flyer schemes, try to alleviate this problem by regularly sending their members a statement of accumulated points. Typically, the statements are accompanied by promotional material about the aspirational value and ease of achieving various rewards.

We suggest that more immediate rewards are preferable to delayed rewards, and that direct support of the target product’s value proposition increases the chance that the program will build loyalty for the product and not just the program. We classify different types of loyalty schemes according to the reward’s support of the product or service value proposition and the reward’s timing (see Figure 2). From the customers’ perspective, the instant gratification programs (in sections one and three of the figure) should be preferable to those with delayed gratification (in sections two and four). However, from the sponsor’s perspective, programs that make an explicit link between product and program and have a bearing on long-term behavior (in sections one and two) should be preferable to those with indirect rewards (in sections three and four). From either perspective, those in section four appear to be least preferable, yet many current loyalty schemes fall into this category, for example, the Australian “Fly Buys” frequent-buyer program.

Sponsors offer membership in the Fly Buys program at no cost. Members accumulate points toward free air travel and hotel accommodation by using a bank credit card and/or a magnetic-strip membership card with the scheme’s sponsors — a major retailer, a car rental company, and a gasoline company. Within a year of launch, 1.7 million Australians had taken this free option, which represented 10 percent of the Australian population, with one member in every four households.22 The customers’ immediate rewards are psychological, namely, a feeling of participation, the anticipation of future rewards, and a sense of belonging. The delayed rewards are a bimonthly summary of accumulated points and sometimes the qualification for a reward.

A customer cost-benefit analysis helps explain why the least desirable loyalty program (according to Figure 2) has attracted so many members. On the cost side of the equation, joining the program is free, and each transaction is handled with an easy-to-use magnetic strip card. Many people perceive rewards, however, as very difficult to achieve, so the sponsors have run a TV ad campaign to counter the perception.23 On the reward-benefit side, many people value the aspirational aspect of air travel, and others just like being part of a program.

With such a huge market penetration, this program would seem to be a runaway success. But, while the sponsors had hoped that rewards would be sufficient to increase sales volumes, in practice, they have been fairly elusive.24 Also, with so many members, the scheme is expensive to maintain.

Both the Fly Buys case and the research on rewards suggest that designing a loyalty program to disrupt established patterns of behavioral loyalty is difficult. Despite these difficulties, many schemes have recently been launched. In the next section, we review some claimed benefits.

The Claimed Benefits of Loyalty Programs

Advocates of loyalty programs contend that they are profitable because:

  • The costs of serving loyal customers are less.
  • Loyal customers are less price sensitive.
  • Loyal customers spend more with the company.
  • Loyal customers pass on positive recommendations about their favorite brands or suppliers.

These alluring benefits go far to explain the interest in customer loyalty programs. But, with one notable exception,25 there is little well-documented empirical research to substantiate these claims. As we have outlined, evidence from behavioral loyalty and reward research suggests the opposite. Hence, before willingly accepting that customers in loyalty programs are always more profitable, let’s examine each source of increased profitability.

When there are specific start-up costs involved in serving a new customer, such as prospecting, credit checks, and entering the customer’s account details in a database, the costs exceed those of serving a repeat customer. However, it is not at all clear why the costs of serving a very regular, loyal, repeat customer should in principle be different from those of serving any other type of repeat customer. Why some transactions will differ in cost has more to do with the type of transaction, not the loyalty of the customer or his or her membership in a loyalty program. The key variables driving cost are first purchase versus repeat purchase, size and type of order, special versus standard order, and so on, not loyal versus divided-loyal customers.

Why would loyal customers be less price sensitive? They may be, but then again, they may not. It depends on how important they think price is and on the value proposition that the brand offers. Although one frequent claim of brand-equity researchers is that brand loyalty and higher prices are positively correlated, this does not automatically mean that more loyal buyers are less price sensitive.26 It may simply mean that these people buy a brand at a higher price because they perceive it to be better. For example, usually a brand is clearly in only one price category. Less price-conscious people then have the opportunity to buy at either the cheaper or higher price depending on whether the brand can offer a good reason (functional or psychological) to justify its higher price. It is perceived brand value, not brand loyalty, that drives price insensitivity.

Alternatively, loyal customers may come to expect a price discount or better service. In other words, what are the rewards to the customer for his or her loyalty? If we consider the double jeopardy relationship, these loyal customers are likely to be slightly more frequent buyers and hence may expect a volume discount.

The next assumption is that loyal customers spend more with the company. This may be simply because they buy more of the product category than less loyal customers (e.g., business air travelers versus holiday travelers). As such, it is the weight of purchase that matters most, not necessarily customer loyalty. And, as we argue in the next section, this is more likely to be a function of the better value offered than of any add-on loyalty program.

The last assumption contends that loyal customers pass on favorable comments. While this seems a sensible assertion, there is little research to indicate what percentage of loyal customers help a company to market its products. The interesting question here is whether only loyal customers, or only those in a loyalty program, are likely to do this, or is it simply that satisfied customers are the ones that say nice things about good products and services. If satisfaction is the key driver of positive recommendations, any satisfied customer should provide this benefit. The only way that a loyalty program can give extra leverage to a company’s word-of-mouth marketing is if the loyal customers offer substantially more, or more effective, positive comments. We are not aware of research that demonstrates this.

In short, the contention that loyal customers are always more profitable is a gross oversimplification. Each company needs to use its customer data to determine the truth of these assertions. Here is where the behavioral loyalty research has relevance. If a particular brand fits the market conditions under which these research findings hold, there are some interesting implications for the introduction of a customer loyalty program. The market conditions are:

  • There is open competition.
  • Products and services are functionally equivalent in broad terms (and are therefore substitutable).
  • There is little tendency for any brand to uniquely appeal to a particular subgroup of consumers (despite brands within a product category being superficially different).
  • There is little dynamic variation over time in competing brands’ market shares (despite considerable marketing activity under the surface to maintain these shares).27

We can hear your sighs of relief. “This doesn’t apply to our brand!,” you say. But are you sure? Certainly there are some products and services — those for state monopolies, highly innovative new products, and products whose success depends on fad and fashion — that do not appear to match these market conditions. Moreover, some trivial differences certainly seem to affect choice.28 However, most airlines, banks, beers, executive education programs, frozen foods, hardware stores, instant coffees, mineral waters, gas stations, plastic pipes, cars, stationery suppliers, TV soap operas, and so on compete directly with each other within their product category. This usually means that when specific products and services are not conveniently available, most people will not hesitate to buy a similar brand or may deliberately seek variety.

Our argument is that these market conditions hold more often than you may think. Given these effects, in the next section, we make some suggestions for designing a loyalty program that offers the best chance of providing a positive return on its investment.

How to Design an Effective Loyalty Program

Here we focus on loyalty programs that attempt to lock in a customer by offering an accumulating benefit, which increases the switching costs to the buyer over time (the programs in section two and four of Figure 2). In economic terminology, they try to change the customer’s choice process from operating in a spot market to operating in a multiperiod, contractual relationship market. The programs are potentially dangerous because they require a company’s long-term commitment and funding. The programs in section one and three of Figure 2 are short-term promotions; they might well stimulate sales for the duration of the promotion but do not have any long-term behavioral after-effects.29

For frequent-buyer loyalty programs, the level of customer involvement is an important consideration. For a high-involvement purchase, the consumer is likely to be involved with both the category purchase decision (Will I fly or go by train from London to Paris?) and the brand choice (Will I fly British Airways or Air France?). For low-involvement decisions, the level of involvement is likely to be low for both decisions, although somewhat higher for the category purchase decision. We suggest that loyalty programs will be more effective for high- than low-involvement products and services, primarily because low-involvement products are often bought by consumers out of habit, while, for high-involvement products, consumers might form a relationship with the supplier (the difference between the habitual purchase of Nescafé, say, and joining Club-Med).

In either case, to maximize the program’s chances of success, the following guidelines are helpful:

1. Design the loyalty program to enhance the value proposition of a product or service. To the extent that frequent-flyer clubs and the GM card enhance the value of the airline or car, these schemes are more worthwhile than a frequent-buyer scheme, such as Fly Buys, which does not. In this respect, probably the least useful rewards for customer loyalty are free gifts such as lottery tickets; these are nice to receive but tend to be only short-term tactical froth that can devalue the brand.

2. Fully cost the loyalty program. There are a number of highly visible costs, such as those associated with program launch, database creation and maintenance, value of rewards claimed, and issue of regular activity statements. The costs of frequent-flyer programs are reportedly between 3 percent and 6 percent of an airline’s revenue.30

Many airlines peg their advertising spending at approximately 3 percent of revenue. Many other loyalty costs are less visible, namely, the opportunity cost of managers’ time spent on the loyalty program rather than on other marketing activities, and the effectiveness of the loyalty program compared with an alternative use of the funds. It is wise, therefore, to be cautious in allocating the marketing budget to a loyalty program. If competitors make a countermove, resources will be needed for a robust response.

3. Design a reward scheme that maximizes the buyer’s motivation to make the next purchase. Most existing reward schemes achieve this only indirectly because they don’t account for customers’ current situations (Are they light or heavy buyers? How many reward points have they accumulated?).

In many reward schemes, for each dollar spent, a participant gains the same number of points (see Figure 3, section A). In the Fly Buys program, the more a member buys, the more he or she flies. Airlines often use a variation of this scheme; an economy class air-fare results in 1 point per mile or dollar spent, business class 1.5 points, and first class 2 points. The incentive is to spend enough to gain access to different reward levels. Some other reward schemes are more transparent to the buyer (see section B). They offer more reward points for each additional dollar spent so that the next purchase is increasingly more valuable to the buyer. The three zones (1, 2, and 3) in the response function are designed to balance the costs to the company of having too many unprofitable participants with the customers’ motivation to participate in the scheme.

In zone 1, light buyers may join but will not gain many rewards. Unless a company can be extremely effective in cross-selling other products and services to these buyers, it doesn’t want them in a loyalty program. Maintenance of a database and regular communication with too many of these buyers can be costly; however, it is important not to alienate them. They may not think of themselves as light or insignificant buyers. The response function is steepest in zone 2. Here the aim is to motivate the midsize buyer to continue buying and to allocate more purchases to the product. These customers are likely to be the most profitable type, big enough to be profitable to serve, but not so big that they request a volume discount. The response function is flat in zone 3, which implies a company should limit the rewards available to super-heavy buyers. The value proposition is already good for these customers, and their current level of buying suggests that it will be difficult to entice them to buy even more, unless the company enhances the value proposition itself or gains wider distribution.

The point here is that whatever reward scheme a company adopts, it should design it with the profitability of different types of customers in mind. Ideally, this will motivate the most profitable customers to give a company a higher share of their business. But realistically, in a competitive market where copycat schemes are inevitable, the aim may be no more than to ensure that the company maintains market share (with the attendant level of loyalty and divided loyalty).

4. Consider specific market situations in planning. (For some typical situations, together with comments on whether a customer loyalty program is appropriate, see Table 1. Because the situations are usually interrelated, our comments are guidelines, not inviolate prescriptions.)


There are three primary lessons from the research and examples we cite here.

First, a major reason for the launch of many customer loyalty schemes is competition. Companies may want to preempt a competitor (and possibly secure first-mover advantages) or respond to a competitor’s scheme (as in most of the frequent-flyer clubs).

Second, apart from purely defensive reasons, if a loyalty program does not support the product or service value proposition, it might be justified in enticing more distributors to handle the product, a demand-pull effect. As noted earlier, for many products and services, there is a positive relationship between distribution coverage and market share.

Third, the behavioral loyalty research we reviewed here suggests that brand loyalty is more likely to come from the market in which a company operates and the brand it has already than from an add-on customer loyalty program. In other words, in most cases, all that a customer loyalty program will do is cost money to provide more benefits to customers — not all of which will be seen as relevant to the brand’s value proposition and/or positioning. The program is unlikely to significantly increase the relative proportion of loyal customers or profitability.

These lessons suggest that customer loyalty programs that (1) neutralize a competitor’s program, (2) broaden the availability of the product or service, or (3) directly enhance the product or service value proposition may be worthwhile. But they also suggest that (4) it is probably a mistake for a company to introduce a frequent-buyer program if it is selling a parity brand in a competitive market and merely add a me-too scheme. Competitors are sure to counter the move with something of equal perceived value. If they offer a price cut, the value of such an immediate reward may be more motivating than the promise of a potential but delayed reward. If they counter with a loyalty program, it is likely to be a better program in the hope of winning back defecting customers.

We have reviewed a body of research that indicates that many programs widely discussed in the business press may be seriously flawed. Like some recent management strategies or fads, many customer loyalty programs seem to have been adopted too quickly, without much thought.31 Our aim here is to generate a more critical analysis of the schemes, especially those added on to prevent the gradual loss of customers.



1. T.J. Kearney, “Frequent Flyer Programs: A Failure in Competitive Strategy, with Lessons for Management,” Journal of Consumer Marketing, volume 7, Winter 1990, pp. 31–40;

D. Churchill, “War in the Air: The Scramble for Points Hits Turbulence,” The Sunday Times (London), 21 November 1993;

P. Meller, “Frequent Flyer Offers Fail to Boost Loyalty,” Marketing, 4 November 1993.

2. D. Peppers and M. Rogers, The One to One Future: Building Relationships One Customer at a Time (New York: Currency/Doubleday, 1993);

S. Pearson, Building Brands Directly (Basingstoke, England: Macmillan Press, 1996);

R.C. Blattberg and J. Deighton, “Manage Marketing by the Customer Equity Test,” Harvard Business Review, volume 74, July–August 1996, pp. 136–144.

3. H. Hakansson, ed., International Marketing and Purchasing of Industrial Goods (Chichester, England: John Wiley, 1982).

4. F.F. Reichheld and W.E. Sasser, “Zero Defections: Quality Comes to Services,” Harvard Business Review, volume 68, September–October 1990, pp. 105–111;

F.F. Reichheld with T. Teal, The Loyalty Effect (Boston: Harvard Business School Press, 1996);

F.F. Reichheld, “Learning from Customer Defections,” Harvard Business Review, volume 74, March–April 1996, pp. 56–69;

R.T. Rust and A.J. Zahorik, “Customer Satisfaction, Customer Retention, and Market Share,” Journal of Retailing, volume 69, Summer 1993, pp. 193–215.

5. Quote by V. Jenkins, chairman of the direct marketing agency, Clemenger Direct Response, in:

A. Biziorek, “Desperately Seeking Loyalty,” Australian Professional Marketing, September 1994, pp. 15–18; and

Reichheld and Sasser (1990).

6. Blattberg and Deighton (1996).

7. A.S.C. Ehrenberg and G.J. Goodhardt, Understanding Buyer Behavior (New York: J. Walter Thompson and the Market Research Corporation of America, 1977).

8. For a discussion of behavioral loyalty, see:

A.S. Dick and K. Basu, “Customer Loyalty: Toward an Integrated Conceptual Framework,” Journal of the Academy of Marketing Science, volume 22, 1994, pp. 99–113.

9. W. Gordon, “Retailer Brands — The Value Equation for Success in the 90s,” Journal of the Market Research Society, volume 36, number 3, 1994, pp. 165–181; and

W. Gordon, “Taking Brand Repertoires Seriously,” Journal of Brand Management, volume 2, number 1, 1994, pp. 25–30.

10. A.S.C. Ehrenberg, Repeat-Buying: Facts, Theory, and Applications, 2nd. ed. (London: Charles Griffin and Co.; New York: Oxford University Press, 1988).

11. Ibid.;

M. Uncles, A.S.C. Ehrenberg, and K. Hammond, “Patterns of Buyer Behavior: Regularities, Models, and Extensions,” Marketing Science, volume 14, number 3, 1995, pp. G71–G78; and

A.S.C. Ehrenberg and M. Uncles, “Dirichlet-Type Markets: A Review” (Sydney, Australia: University of New South Wales; London: South Bank Business School, working paper, 1996).

12. Ibid.

13. OAG Business Traveller Lifestyle Survey 1993 (Dunstable, Bedfordshire, England: Official Airline Guides, 1993).

14. M. Uncles, “Do You or Your Customers Need a Loyalty Scheme?,” Journal of Targeting, Measurement, and Analysis for Marketing, volume 2, number 4, 1994, pp. 335–350;

M. Uncles, “The Seven Perils of Loyalty Programmes,” The Marketing Society Review, Autumn 1994, pp. 18–20;

P. Hawkes, “The Customer Loyalty Challenge,” Admap, January 1996, pp. 47–48; and

S. Rayner, Customer Loyalty Schemes: Effective Implementation and Management (London: FT Retail and Consumer Publishing, FT Management Report, 1996).

15. A.S.C. Ehrenberg, “If You’re So Strong, Why Aren’t You Bigger?,” Admap, October 1993, pp. 13–14, 20;

A.S.C. Ehrenberg, “New Brands and the Existing Market,” Journal of the Market Research Society, volume 33, number 4, 1991, pp. 285–299;

B. Kahn, M.U. Kalwani, and D. Morrison, “Niching versus Change-of-Pace Brands: Using Purchase Frequencies and Penetration Rates to Infer Brand Positionings,” Journal of Marketing Research, volume 25, November 1988, pp. 384–390.

16. A.S.C. Ehrenberg, G.J. Goodhardt, and P. Barwise, “Double Jeopardy Revisited,” Journal of Marketing, volume 54, July 1990, pp. 82–91.

17. As the length of the observation period gets longer (e.g., from a month to a year), more people will buy a brand, their average purchase frequency will rise, the percent who are repeat buyers will increase, and rates of 100 percent loyalty will fall. Loyalty, therefore, has to be defined with reference to the observation period.

18. P.S. Fader and D.C. Schmittlein, “Excess Behavioral Loyalty for High-Share Brands: Deviations from the Dirichlet Model for Repeat Purchasing,” Journal of Marketing Research, volume 30, November 1993, pp. 478–493.

See also:

N.R. Barnard, A.S.C. Ehrenberg, K. Hammond, and M. Uncles, “Dirichlet Discrepancies: Some Empirical Results” (London: South Bank Business School, working paper, 1995).

19. M.L. Rothschild and W.C. Gaidis, “Behavioral Learning Theory: Its Relevance to Marketing and Promotions,” Journal of Marketing, volume 45, Spring 1981, pp. 70–78.

20. L. O’Brien and C. Jones, “Do Rewards Really Create Loyalty?,” Harvard Business Review, volume 73, May–June 1995, pp. 75–82.

21. R.R. Bootzin, G.H. Bower, J. Crocker, and E. Hall, Psychology Today (New York: McGraw-Hill, 1991).

22. Fly Buys Escape: The Magazine for Members, August 1995, p. 3.

23. “Fly Buys,” Ad News, 16 December 1994, p. 24.

24. B. Sharp and A. Sharp, “Loyalty Programs and Their Impact on Behavioural Loyalty Patterns” (Adelaide, Australia: University of South Australia, MSC Technical Report Number 4025, 1996).

25. Reichheld with Teal (1996).

26. Brand loyalty is one of the components in many definitions of brand equity. See, for example:

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

27. Ehrenberg (1988);

Uncles et al. (1995); and

Ehrenberg and Uncles (1996).

28. G.S. Carpenter, R. Glazer, and K. Nakamoto, “Meaningful Brands from Meaningless Differentiation: The Dependence on Irrelevant Attributes,” Journal of Marketing Research, volume 31, August 1994, pp. 339–350.

29. A.S.C. Ehrenberg, K. Hammond, and G. Goodhardt, “The After-Effects of Price-Related Consumer Promotions,” Journal of Advertising Research, volume 34, July–August 1994, pp. 11–21.

30. “Extra Lift for Airlines,” Asian Business, August 1993, pp. 44–46.

31. F. Hilmer and L. Donaldson, Management Redeemed (New York: Simon & Schuster, 1996).

Reprint #:


More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.

Comments (3)
Emiliano Porciani
Does anyone have a deeper resesrch in loyalty programs to share? I work at a universal bank in argentina and no sure in the effectiveness of loyalty programs.
Partha Saha
Nice article! Customer loyalty is a process to build an emotional bonding between customers and organization. According to Gartner, 80% of your company’s revenue comes from just 20% of your existing customers. Check these 15 mind blowing customer loyalty statistics that will prove this article's every word .
badry jaafar
provide very valuable facts and arguments for anyone who wish to improve their customer service.