What to Read Next
SUPERIOR SERVICE IS FASHIONABLE in the business press these days. From the American Banker to Fortune magazine, service cheer-leaders extol the virtues of investing to improve all types of customer service across industries ranging from financial services to manufacturing. One magazine notes that the rewards of service have never been clearer. Similarly, competitors in every sector are declaring their determination to be first in service. As one division manager said recently, “We worship service. I am considered a heretic if I ever question a service initiative.”
Indeed, traditional elements of competitive differentiation are declining. Markets are overcrowded. Competitors imitate product innovations quickly. Proprietary manufacturing techniques are soon diffused throughout the industry. In contrast, service appears to offer an opportunity for positive lasting differentiation, at least theoretically. But many executives are skeptical. They ask: Can companies truly distinguish themselves from competitors based on greater service? Will customers pay for service? In the final analysis, is the payoff from better service likely to be worth the investment?
Like most fads, the current wave of enthusiasm for service contains elements of truth. However, investments in service are two-edged swords—they can create large benefits, or they can be a massive waste of time, effort, and shareholders’ money. Even ones that are successfully implemented (and many are not) may not change customers’ perceptions or behavior. Regardless of whether they are improving the “product” in a service industry or strengthening customer service in a traditional make-and-sell business, managers must adopt the same rigorous attitude they use to develop product strategies and distribution strategies if they are to develop strategically sound service strategies.
And, to do this, they must understand:
- Why service investments fail.
- How customers view service.
- Principles for when to invest (or not) in service initiatives.
Why Service Investments Fail
Clearly, superior service can be a competitive weapon, as demonstrated by the financial success of acknowledged service leaders like American Express, Merck, and McDonalds. However, the fact that superior service can generate a competitive advantage does not mean that every attempt to provide superior service will create a competitive advantage. In this respect, service is no different than product development or distribution. Everyone accepts that investments in superior products or distribution can produce sustainable competitive advantage. But we also accept that not all investments in products or distribution produce advantage or sustainability. So too with service.
Efforts to improve service often fail to produce such results, for a number of reasons. First and foremost, many managers simply think incorrectly about service issues. They approach service issues in “warm, fuzzy” terms, rather than as a set of hard business decisions. They believe superior service is the key to competitive advantage, but consider it unnecessary to prove whether this is so in their particular situation, or to define exactly what is meant by “superior service.” They believe superior service is the result of an all-pervasive “good attitude” among employees, and that any perceived lack of commitment by them will undermine the necessary attitudes. Ironically, these very managers often underinvest in critical aspects of service because they assume that speeches and awards are enough or because they have diffused their organization’s attention across all possible types of service improvements.
However, even when managers focus their efforts and investments, service initiatives often fail, because:
- The investment may fail to produce the desired service levels. Much money has been spent in attempts to raise service levels without the change actually being achieved. For example, a large money center bank spent millions creating a high-service, complex financial account to better serve the service-sensitive affluent market. However, the product proved so complex that the systems supporting it failed. At one point, only 35 percent of that product’s customers were satisfied, and the bank stopped accepting new accounts.
- The effort may improve service in ways the customer does not care about. Probably the most common mistake is incorrectly assuming that improvements in a particular aspect of service will necessarily affect customers’ overall perception of service. For example, a regional bank in the Southeast spent substantial sums to reduce the average time customers waited in line to see a teller from six minutes to three minutes, only to find that satisfaction scores did not move (nor did the bank’s market share). The bank failed to realize that its average wait time was already acceptable to most customers—customers cared only that their wait time never exceeded ten minutes.
- The effort to improve service may be made at the expense of other, more important items. Particularly in businesses with thin margins, managers face tradeoffs between investments in service and investments in upgrading products, extending distribution, adding promotions, and so on. Companies cannot afford everything—they must make tradeoffs. Customers do not always place better service among their greatest needs. Customers of the most successful toy retailer, Toys-R-Us, willingly endure longer checkout lines and less sales assistance in order to obtain a wider array of toys and lower prices.
- The service innovation may be copied quickly. For example, little time passed between the first airline’s introduction of advanced seat selection and imitation of this innovation by others. No one even remembers who offered the service first. Customers cannot reward those they cannot remember.
- The margins in the business may not support the cost of additional service. Service costs are insidious. The full cost of an increase in service due to additional effort per transaction, for example, is often difficult to detect in advance. Many companies have initiated new services only to find margins falling to unacceptable levels. Others have had to drop traditional services as margins fell. For example, “full service” lines at gas stations bear only a faint resemblance to the “ordinary” service levels at those same stations twenty years ago.
How can companies avoid these pitfalls? Certainly, they should not simply give up on service as a competitive weapon. They should accept the premise that superior service may provide an important advantage. But they should not blindly support every proposed service initiative. They must see service decisions as resource allocation decisions, just as product decisions and distribution decisions are. Many problems can be avoided by approaching service investments with the same rigor used for investments in products. Service investments should be expected to produce a value to customers disproportionate to the cost of investment and a demonstrable change in customer behavior. If behavior does not change, the investment cannot pay off. To achieve this result, companies must first understand how customers perceive service—in fine detail.
How Customers View Service
Most companies already measure customers’ perceptions of their service—badly. They survey customers to calculate “service satisfaction” scores: the percentage of customers rating service as excellent, good, fair, or poor. However, this approach provides little usable data about customers’ actual perceptions of service and little guidance about the leverage of improving service. To sufficiently understand customers’ views of service, companies must understand their service at a much finer level of detail. The four propositions that follow facilitate this process.
- Service, in the customer’s view, consists of many separate interactions with the company (“service encounters”), and each encounter has a number of “attributes” (for example, time required, level of courtesy, and competence). Each encounter and attribute must be analyzed separately.
- Customers’ expectations about service are a function of whether the encounter is environmental, transactory, or assistance-based in nature.
- Individual service encounters produce one of five states of satisfaction: O.K., dissatisfied, irritated, angry, or excited. Each state has direct implications for customer behavior.
- Links exist between service and customer behavior, but they are not always straightforward.
We look at each of these more closely.
Service Consists of Many Elements
Even in the most simple make-and-sell business (let alone service businesses), customers have many types of encounters with the company, all of which affect their perception of service: sales visits, calls to locate delinquent orders, delivery, after-sales service, etc. Each encounter has a number of measurable attributes. For example, when withdrawing cash at a bank, a customer is affected by line wait time, cleanliness of the branch, friendliness and competence of the teller, limits on the size of the transaction, requirements for presenting identification, and so on.
Service analysis must be conducted individually on each significant attribute. In most cases improvements to attributes must be pursued individually; for example, investments to reduce counter wait times at a parts distributor will probably not affect complaints about stock-outs. Therefore, to improve service, companies should focus on individual attributes, rather than on overall satisfaction.
Fortunately, attributes vary in importance. Most go unnoticed by customers, unless they are performed so badly as to be intrusive. A few attributes are strongly differentiating, such that small changes can significantly affect overall impressions. These attributes usually relate closely to the fundamental rationale for the purchase itself—for example, good food at a restaurant or wide acceptance of a credit card. These few attributes constitute the “service promise” and, as such, demand special attention. For example, to a large segment of the business community, precise airline schedule differences matter overwhelmingly. Business flyers regularly switch airlines for a thirty-minute difference in flight times. It would take very rude flight attendants over many flights to cause a business flyer to switch and endure the thirty-minute schedule disadvantage. The schedule is the service promise; the attendants are peripheral.
Service Encounters and Customer Expectations
Many companies also find analyzing service difficult because they use the same concepts to describe all encounters with the customer. In our experience, service encounters (whether part of the service promise or not) can be grouped into three categories; each category has a distinct set of requirements for success.
- Environmental service encounters are interactions that customers rarely notice unless they are performed so badly as to be intrusive. Examples include cleanliness of motel rooms and getting baggage after a flight. The key customer need here is that the service be performed satisfactorily without the customer needing to be consciously aware of it. Even though environmental service encounters can be part of the service promise, they are usually relatively unimportant unless performed badly. For example, no one notices when the accountant’s tax calculations are correct, but people certainly notice if the calculations prove incorrect.
- Transactory encounters are ones in which customers are engaged in entirely routine transactions with the provider (for example, checking out at a grocery store, inquiring as to availability of a stock item). Customers expect to “get what they came for” after exerting a certain effort. The expected level of effort depends on their experiences with similar transactions in the past. Their needs are for the transaction to be completed perfectly and for the required effort to be acceptably low.
- Assistance-based encounters are ones where customers cannot precisely define desired outcomes in advance, and thus need the input of the provider to make the best decision. These encounters usually involve some level of risk to customers (e.g., it will cost more than they hoped, they will be embarrassed by their lack of knowledge, or they will learn later that they were not told of a better alternative). The customers’ needs go beyond those associated with transactory encounters; they need assurance that they are dealing with someone competent and trustworthy, and that they matter to this provider. To understand the distinction between assistance-based and transactory services, contrast customers’ feelings when they are purchasing gasoline and when their cars are being repaired. In the former case, customers expect that it will take little effort, that the car will be filled with the correct grade of gasoline, and that they will be charged the correct amount. (They expect the transaction to be completed perfectly and with little effort on their part.) When the car needs repair, other elements apply. Customers expect to exert a different level of effort (but not too much). They also want (but do not necessarily anticipate) the car to be fixed correctly and completely—even if they cannot fully define the problem. In addition, they need psychological reassurance that all repairs were truly necessary. They want the garage to be responsive to their timing needs, take extra care to explain the problem, and give them the chance to authorize the work.
If service providers misunderstand the category of encounter, they inevitably create problems for themselves. When providers underrate the encounter, customers see them as unresponsive. When a provider overrates the encounter, customers see them as intrusive. For example, we have all been annoyed when an airline reservation clerk treated us “like a number” when we needed help figuring out schedules and routes. To her, this was a routine transaction; to us it was a need for assistance. Conversely, we have all been annoyed when an over-zealous waiter in a restaurant attempted to explain every detail of the menu before taking our order. (He mistook a transaction for a need for assistance.) Many attempts by retail banks to sell additional products by teller-based “active” selling efforts have produced negative reactions in customers. By contrast, McDonalds’s simple question, “Would you like a drink with that?” sells additional products but keeps the encounter transactory, as customers expect.
States of Satisfaction
As noted above, customers expect that individual encounters will require a certain level of effort on their part (ranging from zero to quite high) and that the encounter will be completed perfectly a certain percentage of the time. This percentage varies according to the type of encounter and the current level of service in the industry. For example, the same customers expect airline reservations to be correct every time but consider themselves lucky when home repair and maintenance companies perform adequately the first time. I believe customers experience one of five states of satisfaction for an individual encounter.
- “O.K.” occurs when expectations are met but not exceeded. In most cases, this results when a customer exerts the expected level of effort and the transaction is completed perfectly (e.g., when the dry cleaner correctly starches a shirt collar). The customer is technically satisfied, but holds no strong positive feelings toward the provider—feeling instead that another provider could have performed just as well.
- “Irritated” occurs when the transaction succeeds, but the provider is overly intrusive, as described earlier.
- “Dissatisfied” occurs when the transaction does not succeed or requires more than the expected level of effort.
- “Angry” occurs when the transaction does not succeed, despite substantially greater-than-expected effort, or because of a “mistake” by the provider that clearly (in the customer’s mind) should have been avoided.
- “Excited” occurs when the customer is pleasantly surprised by the encounter—when the level of effort required is substantially below expectation, when the customer had low expectations of success, or when the success of the transaction is redefined upward (e.g., by being “upgraded” by an airline or hotel, or by becoming friends with one’s banker). Customers will not be excited by exerting the expected level of effort and successfully completing a transaction that they already expected to go perfectly. In this case, they can only be neutral toward the provider. Only a pleasant surprise can produce an excited response. Customers must perceive that they should not have expected this level of service and probably could not receive it from other providers.
Linking Service Satisfaction and Customer Behavior
When service is analyzed correctly, one can establish a link between customer satisfaction on the most important elements of service (usually the service promise) and customer behavior. For example, a manufacturer of a type of consumer durable with inherent reliability problems (because of its complexity) recendy discovered a tight correlation between the level of customer satisfaction with service performed under warranty and repurchase loyalty over a several-year period. However, the links between satisfaction and behavior are not always easily discerned, for the following reasons.
- The relationship is nonlinear, involving two critical thresholds. There appear to be thresholds of service for affecting customer behavior. In the consumer durable example, when satisfaction rose above a certain threshold, repurchase loyalty climbed rapidly. In contrast, when satisfaction fell below a different threshold, customer loyalty declined equally rapidly. However, between these thresholds, loyalty was relatively flat. I believe this twin threshold framework applies to a wide variety of service situations.
- Each service element is one part of a larger package. Customers purchase all product and service attributes as a package. Only careful analysis can sort out the impact of individual service encounters. And, of course, individual service encounters and attributes differ in importance. For example, most people are willing to suffer through irritating wait times in order to visit the doctor they prefer, presumably because the quality of advice is far more important than the difference in time required.
- Behavior lags behind satisfaction. Dissatisfaction with a single transaction is unlikely to cause the customer to switch loyalties, particularly where switching barriers are high — as they are, for example, in bank checking accounts. Conversely, a single transaction producing a state of “excited” is unlikely to lead to new loyalty. Customers have many encounters, and states of mind build overtime. For example, a string of encounters in a short period producing dissatisfaction can lead to anger, then to a behavior change. Or a string of encounters producing an excited state can lead to the belief that the provider is unique. (I do not believe, however, that a series of “O.K.” transactions can ever lead to the customer being excited. Such a series is unlikely to generate loyalty because the customer lacks the feeling that the provider is anything more than what he can expect of others.) In any case, actual behavior is likely to lag, due to the natural inertia of customers and the difficulty of finding alternative providers.
Investing in Service
The frameworks and concepts outlined above can help managers think more rigorously about service investment decisions. Data on actual satisfaction with individual encounters and attributes, and an understanding of the relationship between customer satisfaction and behavior, can be part of a thorough analysis of likely payoff on service investments. Managers can then use these analyses to guide decisions, whether they are attempting to become the premium service provider in the industry or to invest selectively to achieve maximum payoff on a more limited investment.
Becoming the Service Leader
A few companies have the opportunity to use service to truly differentiate themselves from competition. There are a number of intangible benefits associated with being perceived as the premium service provider, as Merck, American Express, McDonalds, and Walt Disney can attest. However, the number of companies for which this is the appropriate strategy is limited. Each industry can support one or two service leaders at most. To determine whether they ought to pursue such a strategy, managers should ask themselves four sets of questions:
- Is there a large, unsatisfied need for service in this market (i.e., specific encounters with which many customers feel dissatisfied, irritated, or angry)? Or are most customers satisfied with current providers’ service? Is there a large potential for creating an “excited” reaction?
- Can this market be segmented on the basis of service? Is the group of customers that values particular service encounters and attributes more highly than other product/delivery attributes sufficiently large and attractive to support the cost of a superior service strategy? If so, which encounters should be improved? Which attributes?
- Can we definitively fill the existing service gap through a targeted, contained set of programs and actions?
- If we begin to fill the service gap and others imitate our actions, will we be better off than if we never initiated the effort, or will we simply return to the current competitive position but with higher costs?
Depending on the answers to these questions, managers can decide whether to seek the service leader reputation. However, if the company does not currently have the reputation in the market as the superior service provider, it may be difficult or impossible ever to achieve that reputation. Examples of companies with middling reputations for service successfully transforming themselves into the recognized service leaders are few (Scandinavian Air System being the best known). Perhaps the only companies that can aspire to a service leader position are those that have pursued such a strategy from the outset. Since customer perceptions lag behind reality, it may never be possible to overcome the old perception.
For most companies, the service leader position is an unrealistic aspiration. Usually service investments must be traded off against other investments on direct, tangible benefits and therefore must justify themselves by directly reducing costs, increasing revenues, or both.
In some cases, the tradeoff decision is straightforward, because certain service investments in “production” settings (such as high-volume order processing or telephone operations) pay for themselves through reducing operating expenses by eliminating handlings, generating fewer customer inquiries, and so forth. A conceptually straightforward cost benefit analysis can help determine whether such investments are justified.
However, most service investments are intended to affect customers’ experiences directly. Companies seeking to maximize return on such “strategic” service investments should invest to improve service only when they clearly understand which of three impacts on customers they are trying to achieve: minimum effective service parity, loyalty of current customers, or acquisition of new customers. Although related, these objectives are not synonymous. All too often, companies undertake service investments for a vague combination of the three and fail on all of them because they do not focus on each individually. A single initiative can have more than one objective, but it will succeed or fail in achieving each objective individually, based on criteria specific to that objective.
Achieving effective parity.
At a minimum, companies must provide a level of service where they do not develop an adverse reputation and customers do not leave at an unacceptable rate. Too high a rate of dissatisfaction, irritation, or anger is likely to result in overall satisfaction below the lower threshold identified earlier; loyalty and reputation drop off rapidly. Investments to reach minimum standards cannot be “traded off against other investments; they are a cost of doing business and should be considered required investments.
However, achieving effective parity often requires less investment than managers might expect, for three reasons. First, as noted above, most service encounters and attributes do not matter to customers except in extreme situations. Second, as implied by the two-threshold framework described earlier, most customers are indifferent to fairly wide variations in the level of service provided for most encounters, once the lower threshold of service is reached. Finally, customers have imprecise impressions as to the actual level of service being provided, and it is often difficult for customers to compare one provider’s service offerings to those of competitors. Thus, two providers may offer significantly different levels of service in a particular encounter (when measured by objective standards), yet be effectively at parity. In this case, the competitor with the higher service levels has wasted its investment. For example, the regional bank mentioned earlier did not improve its competitive position despite its investment in reducing teller wait times.
A strategy to maintain effective parity with competitors begins with ranking all encounters by their importance to customers and their current impact on customers’ impressions of the company, then:
- measuring the minimum acceptable level of service by seeking a tolerably low level of dissatisfied, irritated, or angry customers as a result of a particular encounter, not a high percentage of customers highly satisfied with that encounter;
- working to eliminate customer anger on an encounter-by-encounter basis, beginning with the encounters that dominate customer views and that currently cause the highest incidence of anger;
- (once anger has been reduced to an acceptable level) remaining in the lower to middle portion of the parity band, avoiding the temptation to “gold plate” each service attribute; and
- constandy monitoring competitors to remain “in the ball park” with respect to important encounters. As competitors’ service levels shift upward over time, an attribute that was formerly at effective parity can slip to being a problem.
In summary, the strategic approach to achieving effective parity in most encounters is to invest only to the degree necessary to be acceptably close to competitors in customers’ eyes.
Retaining the loyalty of current customers.
Often the dominant strategic role of proposed service investments is to improve the ability to retain the loyalty of existing customers. However, in many cases, these investments have little impact on customers. In my view, investments to retain loyalty work best near either the “excited” or the “dissatisfied” end of the service spectrum (above and below the two thresholds described earlier), not in the middle.
At the high end, investments to increase the number of customers excited about service can lead directly to a higher frequency of repeat purchases, etc. Once companies understand this positive service elasticity, they will be well equipped to model the return on new service initiatives aimed at existing customers.
At the low end, companies invest to minimize service-related attrition. However, if a company is already above the minimum acceptable threshold, but near the lower end of the service satisfaction band, investments to incrementally change position may not be warranted. The goal of such defensive investments should not be to eliminate service-related attrition entirely, but to minimize the economic cost of service-related attrition. The benefit of reducing service-related attrition (by eliminating the service problems that cause it) must be balanced against the investment required. The problem is analogous to a quality control issue in which the marginal benefit of a reduction in errors is balanced against the marginal additional cost of quality control needed to produce the lower number of errors. In industries such as consumer banking, customers face barriers to switching (e.g., the effort to close and reopen checking accounts), so their inertia is extremely high. In these cases, if the firm’s overall service offering is acceptable, investments to achieve small changes in satisfaction may not be warranted.
To analyze the cost and benefit of investing to reduce attrition, managers should determine the cost of the contemplated investment and translate that cost into the reduction in attrition (i.e., the number of customers “saved”) required to justify the investment. They can then compare the required reduction in attrition to the portion of customers who cite that particular service problem as a reason for leaving (keeping in mind the fact that customer action lags behind dissatisfaction), as well as the percentage reduction in complaints expected as a result of the investment itself. They will then be in a better position to judge whether the investment is likely to produce the necessary payoff.
Acquiring new customers.
The most difficult goal for service investments to achieve is to acquire new customers. This is not simply a matter of obtaining good references from existing customers. While it always helps to have satisfied customers, the strategic leverage of raising average satisfaction scores in the hope of acquiring new customers is low, because it is difficult for current customers to communicate to potential customers that a company’s service is truly superior.
For example, imagine a person moving to a city and asking various people about the service quality of different banks. Unless one of the people happens to be dissatisfied or angry, all the references will probably consist of adjectives like “good,” “fine,” or “great service” The new arrival cannot hope to distinguish among the banks; the different adjectives simply reflect different speech patterns.
Investments intended to acquire new customers will pay off only when they demonstrably affect customers’ perceptions of value. To do this they should meet four conditions:
- They must improve an encounter or attribute in which the customer values higher levels of performance. As described earlier, a few critical encounters and attributes constitute the “service promise.” A superior promise that is delivered upon is usually central to attracting customers, but not always. Research by one consulting firm showed that the most highly rated service firms were not necessarily the ones that excelled in encounters rated as “most important” by customers. Rather, the most highly rated firms were ones that excelled in those encounters that most providers in the industry performed poorly. Managers should target those encounters that customers indicate are central to the perception of the company’s overall level of performance and in which most providers currently do not adequately meet customers’ expectations.
- The investment should result in a new standard of service that causes an “excited” reaction from the customer. As noted earlier, “excited” occurs only when a new service system succeeds where most others fail, reduces the amount of effort required to complete a transaction or redefines upwards the success of a transaction. Specifically, the new standard of service must convince the customer that the provider’s service is “something special” compared with the service of other providers. For example, Federal Express created a new industry by redefining the standard for time required to ship time-sensitive packages. It focused on cutting the time by over half, rather than on achieving perfect reliability against the previous standard. Similarly, McDonalds’s execution against its famous Quality-Service-Cleanliness-Value standard is certainly not perfect, but imperfect execution against that standard is significantly better than perfect execution against the lower standards of many McDonalds’s competitors. Finally, as noted earlier, perfect execution of the existing standard will not produce an “excited” reaction, because the customer often already expects that standard to be met perfectly.
- The difference in service must be easily describable. Unless the company or current customers can convey to newcomers a tangible difference in the new level of service vis-á-vis that of competitors, the ability to acquire new customers will be lost. For example, while other California department stores strove to improve customer service in a variety of subtle ways, Nordstrom’s created a new service system whereby individual salespeople became (in effect) personal buyers for the customers. Word of mouth soon spread the message about the new, superior service, and Nordstrom’s experienced rapid growth. However, describability is not just a “message” issue. Some service improvements are inherently more communicable than others. Incremental reduction in wait times is difficult to describe; elimination of all waiting is not. Improvements in airline cleanliness or safety are probably not communicable, because of the implication of a previous problem.
- The new level of service must be difficult for significant competitors to copy. Obviously, easily copied service investments are not worth initiating; the initiator bears the risk of experimentation, but imitators can erase any successful difference quickly. Managers should look to the inherent differences between themselves and competitors for clues as to inimitable innovations. For example, American Express was able to offer Global Assist to its cardholders (a service that puts travelers in touch with local doctors and attorneys in emergencies) because its travel offices had already established the network of doctors and lawyers in order to serve its travel agency customers. VISA was forced to create the network from scratch, an expensive and time-consuming proposition. Thus, American Express had an exclusive service advantage for some time, and a permanent cost advantage.
An example of a service investment meeting the criteria above is a portable device recently advertised by two national rental car companies; it allows an attendant to approach the car and process the return there, so the customer no longer needs to come into a central office. This service improves a key service attribute (time and effort required to return the car at airports, when time is always short); it establishes a new standard of service (as opposed to simply reducing the line wait times at the central facility); it is easily describable to newcomers (as shown by the TV commercials); and it is difficult for smaller competitors to copy (because of the fixed cost of installing the system).
In summary, winning service strategies are ones based on the same level of rigorous thought as winning product and distribution strategies. Service, like products and distribution, can be a competitive weapon only if managers clearly understand the customers’ probable perception of its value and its likely impact on their behavior. Managers should ask difficult questions. (Whom is this initiative intended to affect? Will it make a big difference in their experience with us? How will potential customers learn about the improvement in service? Can competitors copy it?) And they should target their improvements, realizing that no company will perform all services perfectly. (Even the generally excellent Disney theme parks have slow food service operations.) In general, if customers care deeply about service, companies should be the best. If customers don’t care, why bother?