CHAPTER 6 CUSTOMER EXPECTATION MEASUREMENT MODEL
6.3 The Expectation Measurement Model
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 116 ~
occurrence of loss events, Loss Given Event (LGE) represents the ratio of transaction or exposure which would be disbursed as loss, given that event and Risk Profile Index (RPI) represents the bank specific risk profile which can be considered a capability of a bank for solving the risk problem. Besides, i is the business line and j represents the risk type.
In brief, operation risk is built by the concept of required capital (i.e. probability) which includes several indicators to calculate and represent the risk value. This study attempts to use the analogy of operational risk to compute the representative value of the determinants of influencing customer expectations. Hence, this approach can enhance the integrity and rationality with using Fechner‟s Law of customer expectation management.
6.3 The Expectation Measurement Model
As addressed in Chapter 4, the customer expectation management engine is composed of four theoretical methods that tightly cooperate with the expectation measurement module. The expectation measurement module is to measure the likely performance of selected determinants delivered by the aforementioned methods via calculating the computable indicators (e.g. the numbers of determinants, the average variation of customer expectation, provider capability, and so on). The expectation measurement module generates two outputs, which include the scores of customer expectation and feasible service tactics, to the aforementioned methods. The aforementioned methods, then, deliver the outputs from the expectation measurement module and information of the context in service encounters to the service component execution module in order for high-quality service experiences. Thus, the expectation measurement module is a critical function for realizing customer mental status, and ensures the integrity and effectiveness of service experience delivery of the customer
‧
6.3.1 Phase of Customer Expectation Measurement Model
The measure used for expectations is based on a mathematical model based on Fechner‟s Law (Thurstone, 1929) and operation risk (Basel Committee, 2001). Figure 6-2 represents the reasoning process of the expectation measurement model, which involves three separate stages, namely: expectation determinants, expectation measurement model and customer expectations. Furthermore, the measurement model also contains feedback which can continuously refresh a real-time database to measure customer expectations.
Expectation determinants stage
The input of the expectation measurement model comprises the combinations of determinants obtained from the methods of customer expectation management engine.
According to Zeithaml et al. (1993), these determinants include enduring service intensifiers, personal needs, transitory service intensifiers, perceived service alternatives, customer self-perceived service role, situational factors, expected service, explicit service promises, implicit service promise and word-of-mouth communications (Zeithaml et al., 1993).
Expectation measurement model stage
This step calculates the scores of the desired and adequate expectations, while
Figure 6 - 2 The process of expectation measurement model
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 118 ~
managing customer expectations are adequate expectation raise, adequate expectation abatement, desired expectation raise and desired expectation abatement. According to the management objectives and combinations of determinants, the stimuli can be computed using a stimulus intensity formula based on an analogical mapping between the factors considered by the operation risk and the stimulus intensity factors regarded in the dynamic service context. After obtaining the stimuli value, the expectation measurement model calculates the adequate or desired expectation scores based on Fechner‟s Law (i.e. the magnitude of the sensations can be calculated based on the magnitude of the external stimulus).
Customer expectations stage
Accordingly, the outputs of the expectations measurement model include adequate expectation score, desired expectation score and list of recommended expectation tactics. Once service providers understand actual customer expectations based on these outputs, they can propose suitable services to assist their customers in achieving their business goals (e.g. customer satisfaction, repeated customers, and business profit). Additionally, the expectation tactics list provides a reference. This list of appropriate expectation tactics can be mapped to specific service components to influence customer expectations via the aforementioned methods. After implementing the expectation tactics (namely, service components), the aforementioned methods should store the values of expectation variation and their capabilities indicators in the real time database. The expectation measurement model can then use the feedback control to reflect actual customer expectations.
6.3.2 Applying Fechner’s Law to Expectation Measurement Model
According to Zeithaml et al. (1993), there are two customer expectation levels (that is, desired expectation and adequate expectation) which can be influenced by determinants. This study describes the details of two expectation measurement models
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 119 ~ based on the theory of Fechner‟s Law.
6.3.2.1 Desired Expectation Measurement Model
Desired services are the high level expectation of customers. It means that the service customers hope to receive (Zeithaml et al., 1993). Besides, the desired expectation is incredibly stable and changeless. For example, some customers always concern about the high-quality of services or the lower prices, so, in other words, their basic needs can‟t change. Consequently, the desired expectation measurement model can be approximately applied by Fechner‟s Law, yet the difference between them is that the mental perceptions of customer would increase slowly in the desired expectation measurement model in terms of the stability of desired expectations.
Hence, the equation of the desired expectation measurement model can be modified in the form,
E
D= K
＊㏒
αI
,in which
E
D is the desired expectation value of the customer affected by the external stimuli, andI
is the stimulus magnitude of the expectation determinants that would be computed through the approach of operation risk. In addition,K
is the constant which can represent the type of customers. According to the dissimilar type of customers, their mental perceptions must be quite different when they touch on the expectation determinants. α is to represent the desired expectation of customers.‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 120 ~
Figure 6-3 clearly shows the curve of the desired expectation measurement model. When the intensity of stimuli enlarges with time, the desired expectation value of a customer also increases. Zeithaml et al., (1993) argued that the adequate expectation is easily influenced by expectation determinants but the desired expectation is considerably changeless. Hence, in Figure 6-3 shows that the desired expectation value would increase slowly, given the intensity of stimuli extremely heightens.
6.3.2.2 Adequate Expectation Measurement Model
Adequate services are the low level expectation of customers, which means customers can accept this level of services (Zeithaml et al., 1993). As mentioned above, the adequate expectation of a customer is often changeable and unstable in contrast to his/her desired expectation (Zeithaml et al., 1993). In other words, the adequate expectation of a customer can be easily influenced by the determinants. The more determinants the providers use, the more mental effects customers have.
Consequently, this study defines that the shape of the adequate expectation nearly belongs to the S curve (as depicted in Figure 6-4).
E
DFigure 6 - 3 Desired expectation measurement model I
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 121 ~
The curve means that the adequate expectation value would increase continually, if the external stimuli enlarge with time. When the intensity of stimuli reaches a certain magnitude (I’), even though the intensity exceeds this magnitude, the adequate expectation value would nearly keep stable for a fixed value. The reason why the adequate expectation becomes steady ultimately is that the adequate expectation has attained the desired expectation based on the concept of the zone of tolerance (Zeithaml et al., 1993). Namely, the zone of tolerance becomes extremely narrow during this period in terms of the overlap between the adequate expectation and desired expectation.
According to Fechner‟s Law, the adequate expectation measurement model can be written in the form
E
A= K
＊㏒
βI
,in which
E
A is the adequate expectation value of the customer influenced by the external stimuli, I is the stimulus magnitude of the expectation determinants that would be computed through the approach of operation risk, and K is the constant which can represent the type of customers. β is to represent the adequate expectation of customers. Figure 6-5 shows the expectation measurement model which combines the desired expectation measurement model and adequate expectation measurementE
AI’
Figure 6 - 4 Adequate expectation measurement model
I
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 122 ~ model.
6.3.3 Applying Operation Risk to Expectation Measurement Model
As mentioned above, the stimuli within the service delivery are the combinations of expectation determinants. It is different from the only single stimulus of Fechner‟s Law. Hence, this study tries to apply the analogy of the operation risk to support what the stimuli are meticulously formed. The concept of required capital is the core though of operation risk. In order to confront the huge capital loss that operational risks could bring about, banks have to prepare appropriate money (i.e. required capital) ahead according to their capabilities, probabilities of loss events or exposure of risks.
This study refers to the formula of required capital and modifies the elements to form the new formula for calculating the magnitude of the stimuli. Accordingly, the formula of stimulus intensity can be written in the form
SI = UDI * PSE * AEV * CPI.
Table 6-1 shows the analogical mapping relationship between the required capital and stimulus intensity formulas. There are four key elements in this equation, and the details are as follows.
Figure 6 - 5 Expectation measurement model
E
DE
AE
I
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 123 ~
Table 6 - 1 The mapping table of the required capital and stimulus intensity formulas Required Capital
Mapping
Stimulus Intensity Exposure Indicator Usage of Determinant Indicator Probability of Loss Event Probability of Success Event
Loss Given Event Average Expectation Variation
Risk Profile Indicator Capability Profile Indicator
RC = EI * PE * LGE * RPI SI = UDI * PSE * AEV * CPI
Usage of determinant indicator (UDI)
This indicator represents the effect of using expectation determinants for managing customer expectations. According to Fechner‟s Law, the stimulus intensity of a human would become large, if there are many irritants to influence him/her.
Namely, the more determinants providers propose, the larger magnitude the stimulus procures. Let D = {D1, D2,…, Dj} be the set of all determinants for managing customer expectations during service delivery and UDI = {UDI1, UDI2,…, UDIn} be the set of all combinations of determinants. Each combination UDIcontains a subset of determinants chosen from D. Let W= {W1, W2,…, Wl} be the set of the weight, which can influence service providers‟ business goals, of each combination UDI.
Each D has the exclusive weight. Hence, UDI = .
Average Expectation Variation (AEV)
Each combination UDId has its average expectation variation. The expectation variation means the difference between the initial expectation and the terminal expectation while providers implement the determinants in each round. Then, accumulating and averaging the total expectation variations are to realize what significant effects the determinants can provoke. Consequently, if the average expectation variation of an UDId is large, the magnitude of the stimulus would be
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 124 ~
great. Let EV = {EV1, EV2,…, EVj} be the set of values of expectation variation of an UDId. Let AEV = {AEV1, AEV2,…, AEVj} be the set of all values of average expectation variation. The equation of average expectation variation can be written as follows,
AEVi = .
Probability of Success Event (PSE)
The definition of success event is that service providers attempt to utilize certain UDId to reach the average expectation variation. As mentioned above, the average expectation variation is accumulated by each calculation of customer expectation.
Hence, this indicator is to calculate the probability of success event by capturing and updating the real time data. Different UDId would have its probability for achieving the average expectation average. Let PSE = {PSE1, PSE2,…, PSEn} be the set of all values of probability of success event and EA = {EA1, EA 2,…, EA m} (or ED = {ED1, ED 2,…, ED m}) be the set of all values of adequate expectation (desired expectation).
Accordingly, each probability of success event PSEi = .
Capability Profile Indicator (CPI)
According to the operation risk, the risk profile indicator is to evaluate the capabilities of a bank for dealing with the operation risks. In this expectation measurement model, we would like to use the capability profile indicator to assess the competence of service providers. The service provider, therefore, with high capabilities (such as many resources, high capitals or collaborative partnership) can lead to the high magnitude of the stimulus. The equation of capability profile indicator can be written as follows
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 125 ~
CPI = ,
where Ci represents the competence indicators of the provider and Wi represents the weights of each competence indicator. In addition, the capability profile indicator should be based on the domain-specific applications.
6.3.4 Scenario Demonstration
This study employs a B2C scenario of the exhibition service system to demonstrate the utility of our expectation measurement model. In general, exhibitors with high capabilities can increase their visitors‟ expectation level for their customer franchise, and they will propose useful services with different weighting (such as recommendation service, advertisements, or warranty service) to visitors. An exhibitor (i.e. service provider) can deliver services to a visitor (i.e. customer) within one encounter through this exhibition service system. Hence, these services that service providers would like to deliver to customers can be mapped to expectation determinants. According to the past results (that is, the customer expectation variations) of using these services, the levels of specific expectation determinants can be represented as the UDIt. Besides, the exhibition service system can analyze the historical data of the successful probability of using specific services from the exhibition database and compute the PSEt. According to these services, the exhibition service system can also compute the average expectation value AEVt of using specific services in the past. The exhibition service system can also derive the value of the exhibitor‟s capability (CPIt) based on the exhibitor‟s numbers of frontline employees, numbers of booths, numbers of service category, or capital assets in the exhibition.
The exhibition service system immediately transforms the four values into a value of stimulus intensity (SIt) used to calculate the value of the visitor‟s expectation.
Meanwhile, visitors can generally be divided into diverse classifications (based on age,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
~ 126 ~
gender, occupation, or consumption aims) representing K. Consequently, the service exhibition system can acquire the values of two expectation levels (namely, ED and EA). When the exhibitor realizes the expectation values of visitors, it can flexibly modulate proper services in order to increase visitors‟ expectations. In other words, the expectation measurement model can enable exhibitors to deliver innovative service experience by closely grasping what visitors want in real time during the service delivery.