- Email: [email protected]
A customer-oriented organisational diagnostic model based on data mining of customer-complaint databases
Expert Systems with Applications 39 (2012) 786–792
Contents lists available at ScienceDirect
Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa
A customer-oriented organisational diagnostic model based on data mining of customer-complaint databases Chi-Kuang Chen a,⇑, An-Jin Shie a, Chang-Hsi Yu b a b
Department of Industrial Engineering & Management, Yuan Ze University, Taiwan, ROC Department of Business Administration, Yu Da University, Taiwan, ROC
a r t i c l e
i n f o
Keywords: Organisational diagnosis Customer complaints Customer-orientated service Data mining Service system
a b s t r a c t The purpose of this paper is to develop a customer-oriented organisational diagnostic model, ‘PARA’ model, based on data mining of customer-complaint databases. The proposed ‘PARA’ model, which is designed to diagnose and correct service failures, takes its name from the initial letters of the four analytical stages of the model: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. In the primary-diagnosis stage, the customer-complaint database is comprehensively analysed to identify themes and categories of complaints. In the advanced-diagnosis stage, a data-mining technique is employed to investigate the relationship between the categories of customer complaints and the deficiencies of the service system. In the review stage, the identified weaknesses of the service system are reviewed and awareness of these weaknesses is enhanced among the organisation’s employees. In the action stage, a strategy of action plans for improvement is developed. An empirical case study is conducted to demonstrate the practical efficacy of the ‘PARA’ model. The paper concludes by summarising the advantages of the proposed model and the implications for future research. Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.
1. Introduction Customer satisfaction is recognised as one of the most important key performance indicators of success. However, it is always difficult to eliminate all causes of customer dissatisfaction and complaints. Customer satisfaction is influenced by such a variety of factors—including the attributes of a product or service, the individual needs of customers, and the service quality provided by front-line personnel—that even a fully ‘customer-focused’ service program cannot eliminate all product or service failures. Most organisations are aware that service failures must be handled appropriately to avoid harm to the organisation’s goodwill and profits (Hart, Heskett, & Sasser, 1990). Service recovery has thus become an increasingly important issue to prevent the loss of customers (Kelley & Davis, 1994; McColl-Kennedy & Sparks, 2003; McColl-Kennedy, Daus, & Sparks, 2003; Sparks & McColl-Kennedy, 1998; Varela-Neira, Vázquez-Casielles, & Iglesias-Arguëlles, 2008). The term ‘service recovery’ refers to remedial actions that are taken to re-establish the satisfaction of customers when product or service failure has occurred (Chaston, 1993; Zemke & Bell, 1990). Many studies have demonstrated that effective service recovery ⇑ Corresponding author. Address: Department of Industrial Engineering & Management, Yuan Ze University, 135 Yuan-Tung Road, Chung-Li 320, Taiwan, ROC. Tel.: +886 3 4638800x2528; fax: +886 3 4638907. E-mail address: [email protected] (C.-K. Chen).
can transform negative evaluations into positive impressions, thus maintaining good relationships with customers (Hoffman, Kelley, & Rotalsky, 1995; Karatepe, 2006; Kelley, Hoffman, & Davis, 1993; Maxham Iii, 2001; Smith, Bolton, & Wagner, 1999; Sparks & McColl-Kennedy, 1998; Spreng, Harrell, & Mackoy, 1995). Appropriate service recovery has also been shown to enhance the trust of customers and increase their willingness to re-purchase (Hung & Wong, 2007; Maxham Iii, 2001; Spreng et al., 1995; Tax & Brown, 1998, 2000; Yu & Dean, 2001). Conversely, ineffective service recovery is one of the main causes of switching behaviour (Keaveney, 1995). Most studies of service recovery (Barlow & Moller, 1996; Boshoff, 1997; Boshoff & Leong, 1998; Johnston & Fern, 1999; Keaveney, 1995; Tax & Brown, 1998, 2000; Wirtz & Mattila, 2004) have focused on the effectiveness of specific remedial actions—such as exchanges of goods, apologies, or offers of compensation. Relatively few have studied the question of how to improve the service system by learning from the experiences of previous service failures and avoiding repetitions. It is the contention of the present study that the prevailing focus on remedial actions and compensation for a service failure is essentially a passive and reactive approach to the problem of service failure, whereas efforts to improve the existing service system represent a creative and proactive strategy. To improve a service system and minimise service failures, it is necessary to collect and analyse customer-complaint data
0957-4174/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2011.07.074
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
periodically and comprehensively. Although some studies have indicated that improvement actions should be based on customer complaints (Bosch & Enríquez, 2005; Gustafsson, Ekdahl, & Edvardsson, 1999; Tax & Brown, 2000; Tax, Brown, & Chandrashekaran, 1998), these authors did not propose a comprehensive model to identify the sources of service failures. Chen, Yu, and Chang (2005), Chen, Yu, and Chang (2006), Chen, Yu, Yang, and Chang (2004), who have investigated the importance of service-system design and management, also emphasised the need to make better use of customer-complaint databases to diagnose failures in service systems. Against this background, the present study seeks to establish a customer-oriented organisational diagnostic model of service failure based on customers’ complaints. The proposed ‘PARA’ model takes its name from the initial letters of the four stages of the model: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. The diagnostic model provides a systematic analysis of service failures based on the customer complaint database. Data-mining techniques are then utilised to establish correlations between the identified categories of customer complaints. The model then develops a strategy of improvement actions for the service system. The model provides constructive customer-focused recommendations for improvements in service delivery through scientific analyses of service failures. The remainder of this paper is organised as follows. The next section reviews the relevant literature on organisational diagnosis. The development of the proposed ‘PARA’ model is then presented. The practical efficacy of the proposed model is then demonstrated in an empirical case study of public-sector services in Taiwan. The paper concludes with a summary of the main advantages of the model and the implications for future research.
2. Literature review of organisational diagnosis The term ‘organisational diagnosis’ is commonly used to refer to a process whereby an external consultant enters an organisation, collects valid data about human experiences within the organisation, and feeds that information back to the organisation to promote increased understanding of the organisation by its members (Alderfer, 1981). The purpose of organisational diagnosis is to establish a widely shared understanding of an organisation, and, based upon that understanding, to determine whether change is desirable. An effective process of organisational diagnosis and intervention promotes congruence between organisational objectives and the structure required to achieve those objectives (Nadler & Tushman, 1980). To achieve such an effective organisational diagnosis, Harrison (1994) suggested that diagnostic practitioners must attend simultaneously to three distinctive facets of the diagnostic process: (i) analysis; (ii) methods; and (iii) interactions. Four approaches to organisational diagnosis are especially worthy of consideration: (i) ‘sharp image’; (ii) ‘open system’; (iii) ‘political’; and (iv) ‘customer complaint’. Harrison and Shirom (1999) proposed a distinctive four-step approach to organisational diagnosis termed ‘sharp-image diagnosis’, which begins with a broad view of the organisation and proceeds to a tightly focused diagnosis of critical problems and challenges. The first step in ‘sharp image diagnosis’ is ‘scouting’, which seeks to clarify the nature of service failures in the organisation and to develop a preliminary view of the organisation’s strengths and weaknesses. In the second diagnostic step, the core problems and challenges are organised as a reference for the examination of other parts of the organisation. In the third diagnostic step, one or more focused models are developed to shape the organisation’s response to its critical challenges. In the fourth
787
diagnostic step, consultants provide clients with an emergent diagnostic model that incorporates the factors of time, resources, reward, and information feedback. According to Harrison and Shirom (1999), ‘sharp-image diagnosis’ bridges theory and practice by responding directly to the distinctive conditions that shape a particular organisation’s operations and its options for change. The outcome of ‘sharp-image diagnosis’ is a model that provides clients with a highly focused, multi-dimensional image of conditions underlying basic problems and critical challenges. The second approach to organisational diagnosis is the ‘open system’ perspective, which views the organisation as a system that obtains inputs from its environment, processes those inputs, and then produces outputs. According to ‘open-system theory’, systems have a tendency to run down if not provided with additional resources from the environment and a tendency to remain in a state of equilibrium if not disturbed (Ashmos & Huber, 1987). The opensystem framework analyses any organisation in terms of the flows of inputs, the processing of those inputs, and the creation of outputs. According to Jackson (1991), organisational diagnosis in accordance with a ‘systems approach’ proceeds by a holistic approach that examines the overall environmental and organisational contexts within which problems arise and within which steps toward organisational improvement are enacted. Systembased diagnosis thus seeks to distinguish between the symptoms of ineffectiveness and the underlying systemic causes of ineffectiveness. According to Senge (1990), this approach also enables organisations to identify the possible side-effects and unintended consequences of remedial actions. However, there are also significant limitations to the open-system approach to organisational diagnosis. In particular, some of the principles of the systems approach are too abstract to be useful, which can lead to the adoption of a superficial approach that overlooks important details of particular organisational operations and ignores the significant differences that exist among organisational contexts. The third approach to organisational diagnosis is to treat an organisation as a political arena in which bargaining and exchange take place among internal and external stakeholders who are seeking their own particular benefits or goals (Bolman & Deal, 1991; Hall, 1999; Morgan, 1986). According to this view of organisational diagnosis, the stakeholders’ involvement in the activities of the organisation can be harmful or beneficial for themselves and/or for the organisation itself (Donaldson & Preston, 1995). Stakeholders include shareholders, staff, customers, suppliers, and so on. They can be categorised as ‘main stakeholders’ (who usually maintain a formal contractual relationship with the organisation and have direct economic influence on it) and ‘secondary stakeholders’ (who include all parties that can influence, or be influenced by, the organisation) (Savage, Nix, Whitehead, & Blair, 1991). In conducting organisational diagnosis in accordance with the political perspective, the following steps are taken (Savage et al., 1991): (i) identify the main stakeholders; (ii) analyse each stakeholder’s position; (iii) examine each stakeholder’s power; and (iv) assess each stakeholder’s capacity for action and impact. The political perspective thus emphasises interpersonal relationships within the organisation, the potential for conflicts among stakeholders, and the influence of stakeholder activity on the budget and other forms of resource distribution (Bartunek, 1993). The role of customers (in their capacity as stakeholders) thus plays some part in this model of organisational diagnosis; however, this approach lacks an operational model with established procedures to take proper account of the opinions of all customers as stakeholders. The fourth approach to organisational diagnosis utilises customer complaints as the driving force for analysis. Bosch and Enríquez (2005) developed a customer-oriented model of organisational diagnosis by incorporating total quality management (TQM), quality function deployment (QFD), and plan-do-check-action
788
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
(PDCA) in their so-called ‘customer complaint management system’ (CCMS). This diagnostic model involved a seven-step procedure: (i) document the voice of the customer (VOC); (ii) translate VOC into customer demands and problems; (iii) analyse and solve these questions; (iv) explore customer demands; (v) update failure mode and effect analysis (FMEA) to avoid recurrence; (vi) share solutions with affected customers; and (vii) update system performance measurements. According to Bosch and Enríquez (2005), the CCMS model can change an organisation’s perspective on complaint management and transform the process of answering complaints from a trivial activity to an exciting learning experience in which continuous improvement leads to service excellence. However, although this model does take into account the value of customer complaints, it deals with individual customer complaints on a case-by-case basis and does not retrieve useful information from all customer complaints to create a comprehensive basis on which to develop diagnostic strategies for improving the organisation’s service system. This review of the relevant literature has revealed that most of the extant organisational diagnostic models have noted the importance of customer-oriented service. However, the focus and the operational procedures of these models cannot guarantee that the complaints of all customers will be taken into account. The present study addresses the shortcoming in constructing a comprehensive customer-oriented model for organisational diagnosis. The data mining techniques is used to analyze customer complaint database. 3. The ‘PARA’ diagnostic model As noted above, the present study contends that a proactive approach needs to be adopted in dealing with customer complaints and identifying the ‘voice of the customer’ (VOC). In accordance with this view, a customer-oriented organisational diagnostic model based on a comprehensive analysis of customer complaints is developed in this study. The proposed model, which is illustrated in Fig. 1, is designated the ‘PARA’ model from the initial letters of the four stages in the
model’s diagnostic procedure: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. The details of each of these stages of the ‘PARA’ model are described below. The purpose of the first stage, primary diagnosis, is to organise the firm’s original database of customer complaints into categories, and then to analyse these data in terms of the frequencies of complaints in each category. Because the data collected in such databases is usually disorganised and uncategorised, content analysis is used to provide a coherent structure. In contrast to the CCMS model (Bosch & Enríquez, 2005) described above, the ‘PARA’ model utilises a comprehensive analysis of customers’ opinions, rather than analysing individual customer complaints as isolated cases. In the second stage, advanced diagnosis, data mining is used to seek correlations among the categories of customer complaints. Data mining has been described as ‘‘the nontrivial extraction of implicit, previously unknown, and potentially useful information from data’’ (Frawley, Piatetsky-Shapiro, & Matheus, 1992) and ‘‘the science of extracting useful information from large datasets or databases’’ (Hand, Mannila, & Smyth, 2001; Hsia, Shie, & Chen, 2008). The results of the data-mining analyses (in providing correlations among the categories of complaints) are essential to the development of the improvement strategy of action plans in the next stage of the model. In the third stage, review, the correlations among the various categories of customer complaints are examined to seek correlations’ causes. Apart from confirming the causes of the correlations, the purpose of this stage of the model is to raise the awareness of the pertinent issues among managers and their subordinates. To achieve this purpose, a checklist of items derived from the datamining analyses is presented to staff members, who are encouraged to examine the identified service weaknesses and its causes through open discussion. The fourth stage, action, refers to the development of strategies of action plans for improvement of the service system. In this stage, managers work intensively with consultants to devise improving plans to improve the service system according to customer complaint causes proposed in stage 3. The customer, manager
Fig. 1. PARA model of organisational diagnosis.
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
and consultant’ opinions are taken into account to develop the most comprehensive improving plans. The above process enables the organisation to achieve its improvement goals more efficiently and effectively. In summary, the proposed ‘PARA’ model integrates customer complaints, management opinion, and consultant expertise to develop a constructive improvement strategy that enhances service delivery and organisational efficiency. Compared with the extant models of organisational diagnosis noted above, the proposed ‘PARA’ model has several advantages. In particular the new model is able to: (i) convert customer complaints into ready-to-use strategies for the improvement of a service system; (ii) bridge the gap between an organisation’s service and the demands of its customers; and (iii) promote management’s awareness of problems and potential directions for improvement.
4. Empirical case study 4.1. Research setting To demonstrate the efficacy of the proposed ‘PARA’ model, an empirical case study was conducted using the complaint database of public-sector service agencies in Tao-Yuan county, Taiwan. The aim of this exercise in organisational diagnosis was to utilise the complaints of citizens to ascertain the reasons for failures in the service systems of the Tao-Yuan county government agencies. More than 80 government agencies provide a variety of services (including household registration, land management, social welfare, public facility construction, and business management) to the 1.9 million citizens of Tao-Yuan county. The county governor has established a standardised procedure for handling the complaints of citizens. Under the management system, the responsible agency and the designated staff are expected to follow up all complaints on a daily basis. The complaint-management system was computerised in 2006, which enabled citizens to make their complaints online. The standard procedure of the complaints-management system consists of three steps: (i) complaints are sent to the complaintmanagement system; (ii) the complaints are categorised and each complaint is passed to the department responsible for the issue; and (iii) the responsible department should reply to the complainant promptly regarding the proposed action and timeframe for improvement of the service that is the subject of the complaint. The system includes monitoring to ensure that staffs complete the required tasks within three days of the complaint being received. At the time of the present study, the system had encountered significant difficulties. The computerisation of the complaint system in 2006 proved to be so convenient for citizens that the number of complaints increased dramatically. Because complaints were handled on a case-by-case basis, it soon became impossible for government managers to handle all the received complaints within the prescribed time. Moreover, the increased complaint-handling work load meant that agency managers did not have time to search the complaint database for clues to the underlying causes of service-system failures. It was apparent that there was a need for a diagnostic model that was capable of analysing the complaint data systematically and comprehensively with a view to diagnosing systemic service failures within the entire organisation. In response to this situation, the complaint-management system of the public services of Tao-Yuan county was chosen for application of the proposed ‘PARA’ diagnostic model. Customercomplaint data gathered by the system from 1 October 2006 to 30 September 2007 were used as the database for the study. A random selection of 500 records was then analyzed in the study after
789
excluding repeated complaints from the same citizen and complaints with incomplete or unclear descriptions of problems. 4.2. Analysis by ‘PARA’ model 4.2.1. Stage 1: primary diagnosis The primary-diagnosis stage of the ‘PARA’ model employed the technique of ‘content analysis’ to categorise the unstructured data for systematic analysis of customer complaints. A total of 983 themes were initially extracted by content analysis. These themes were then classified into 18 generic categories. Two external coders were invited to examine the reliability of the content analysis. A random sample of 20% of the 983 themes extracted from the data were examined by the two external coders. The degree of mutual agreement was computed as follows:
Degree of mutual agreement ¼
2 number of items completely agreed by two parties number agreed by Party A þ number agreed by Party B
The degrees of mutual agreement between the original content analysis and external coder 1 was 0.864 and that between the original analysis and external coder 2 was 0.77. A reliability index was then computed as follows:
Reliability ¼
2 Degree of average mutual agreement 1 þ ½ð2 1Þ Degree of average mutual agreement
The reliability index in this study was 0.9, which was greater than the threshold value suggested by Kassarjian (1977). The result thus indicated that the content analysis was highly reliable (Kolbe & Burnett, 1991; Weber, 1985). Table 1 shows the customer-complaint categories as derived by the primary diagnosis. Four categories of complaints were most prominent: (i) inadequate public facility design or construction (21.6% of all complaints); (ii) environmental hazards (11.7%); (iii) unreasonable traffic tickets (10.6%); and (iv) ignorance of citizens’ rights (8.0%). These four complaint categories clearly required the closest attention from management. 4.2.2. Stage 2: advanced diagnosis To identify correlations among the 18 customer-complaint categories, the ‘Apriori’ algorithm was utilised as a data-mining technique. The complaint data were first classified with a Boolean-type code (‘1’ or ‘0’). The Apriori algorithm was then applied using Eq. (1) (below) to calculate the confidence from the two supports:
Antecedent [ Consequence Items Support 100 Antecedent Items Support ¼ Confidence ð%Þ
ð1Þ
in which: ‘antecedent items support’ = the percentage of data that tallies with the antecedent rule; and ‘antecedent [ consequence items support’ = the percentage of data that tallies with the union of antecedent items and consequence items support rules. In applying the Apriori algorithm, a minimum level of confidence needs to be defined as a mining threshold. In this case, the minimum level of confidence was set at 80%. Table 2 presents the results of the Apriori algorithm analysis. Three correlation rules were identified beyond 80% (calculated according to the number of antecedent and consequence items). These three correlation rules covered 153 customer-complaint records in total (with rule 1 covering 94 complaint records, rule 2 covering 12, and rule 3 covering 47), which represents 31% of the 500 customer-complaint records included in the study.
790
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
Table 1 Primary diagnosis of customer complaint categories. Customer complaint categories
Frequency
1. Inadequate public facility design or construction 2. Building construction without licence 3. Poor environmental sanitation and hygiene conditions (e.g. garbage disposal, pet droppings on roads and stray animal management) 4. Viatical obstructions (e.g. unauthorised occupation of roads, insufficient parking spaces) 5. Environmental hazards (e.g. dangerous signboards, uneven road surface, damaged or insufficient numbers of security surveillance cameras) 6. Poor quality control of infrastructure constructions (e.g. flooding, water outage) 7. Noise pollution 8. Inefficient maintenance of public facilities (e.g. streetlamp maintenance, leisure park facilities, venues for exercises, bus, traffic signs, sewerage maintenance, roadway maintenance, government website maintenance) 9. Failure to keep promises 10. Unreasonable traffic tickets 11. Slow administrative procedure 12. Reluctant administrative procedures 13. Insufficient professionalism of public servants 14. Poor service attitudes or absence on duty 15. Ignorance of citizens’ rights 16. Reluctance to take responsibilities 17. Failure to provide out-of-SOP (standard operating procedures) service 18. Lack of empathic services
212 4 33
Total ⁄
30 115
% of total complaints 21.6⁄ 0.4 3.4 3.1 11.7⁄
6
0.6
9 36
0.9 3.7
77 104 77 1 9 49 79 42 22
7.8 10.6⁄ 7.8 0.1 0.9 5.0 8.0⁄ 4.3 2.2
78
7.9
983
100.0
indicates the first four highest figures
Rule 1 proposed that ‘inadequate public facility design or construction’ and ‘ignorance of citizens’ rights’ caused citizens to feel that the service provided by the public agency demonstrated a ‘lack of empathy’. The confidence level of this rule was 100%. Rule 2 proposed that a ‘failure to keep promises’ regarding ‘environmental hazards’ was associated with ‘slow administrative procedure’. The confidence level of this rule was also 100%. Rule 3 proposed that ‘unreasonable traffic tickets’ and a ‘lack of empathic services’ was associated with ‘poor service attitudes or absence on duty’. The confidence level of this rule was 80%. 4.2.3. Stage 3: review In stage 3, the results derive from data mining in stage 2 are further reviewed in regard to the causes and effects. It is to identify the service weaknesses and its causes. In this empirical case, rule 1 is selected as a target example for the review because rule 1 involves the most citizen complaint cases (94 cases, see Table 2) with the highest confidence level 100%. The 94 complaint cases are illustrated in the three aspects after one by one review the contents of complaint raised by citizen. They are inadequate city construction design, insufficient operation procedures and apathetic service interface. The three aspects concurrently exist in the service system which results in the 94 citizen complaints cases being raised. The aspects are discussed in details as follows. 4.2.3.1. Inadequate city construction design. Basing on the 94 complaint case-reviews, insufficient city construction designs are mostly occurred. Lacking consideration of citizens’ needs and strict legal regulations are major causes of the insufficient designs or plans. These complaint cases include: (i) inadequate parking space area design; (ii) wrong traffic sign design; and (iii) inadequate road planning. These public agency inadequate city construction
designs cause citizens to feel ignored of their rights and needs, such that generate repetitive complaints from the same problems. Yet the public agency merely focuses on fulfilling law regulations of the city plan without considering citizens’ opinions. The above situations commonly exist in the insufficient bureaucratic service systems of public agencies. 4.2.3.2. Ineffective operation procedures. In addition, most of the citizens frequently complain that operation procedures of the public agency are ineffective. These reasons include: (i) administrative and front-line service standard operation procedures are too verbose to immediately satisfy citizens’ needs; (ii) lack of supplementary measure diminishes the public agency policy efficiency; (iii) operation procedures are restricted by the existent law regulations, such that cannot avoid unnecessary procedures; and (iv) operation procedures lack in integrating services from different departments, which causes citizens feel inconvenient. These reasons usually derive inefficient performance in a bureaucratic service system of the public agency. 4.2.3.3. Apathetic service attitude interface. Apart from the noted complaint reasons above, apathetic service interface of public agency is another major complaint from the citizens. The service interface is that front-line service providers present service attitude and enthusiasm. These complaint cases include: (i) ignoring citizens’ opinions and rights; (ii) presenting unpleasant service attitude; and (iii) lacking empathic service attitude. These illustrate that the public agency services do not conform citizens’ expectation and are ineffective in handling citizens’ opinions and complaints. In fact, most of the public agencies present a defensive attitude in dealing with citizens’ opinions and complaints. Ignoring citizens’ opinions will result in the public agencies’ repeating same faulty service. Besides, the public agency employees often too conveniently rely on previously established standard operation
791
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792 Table 2 Apriori algorithm analysis. Rule ID 1 2 3
Antecedent
Inadequate public facility design or construction Ignorance of citizens’ rights Environmental hazards Failure to keep promises Unreasonable traffic tickets Lack of empathic services
Consequence
Confidence level (%)
Number of complaint cases
Lack of empathic services
100
94
Slow administrative procedure
100
12
Poor service attitudes or absence on duty
80
47
procedures and unwillingly provide flexible service outside of established procedure, to immediately satisfy variety of citizens’ needs. All these reasons generate gaps between a service system and the real needs of customers. 4.2.4. Stage 4: action In this action stage, the public agency managers and scholastic consultants consolidate and analyze the results from stage 3 to figure out improvement strategies of action plan. The study integrated customers, managers and consultants’ opinions and analysis to develop strategies for improving service system of the public agency and minimizing citizens’ complaints. These proposed improving strategies are divided into three categories coincide with stage 3, as follows: 4.2.4.1. Proposed actions to inadequate city construction design. As pointed out in stage 3, inadequate city construction design generates a gap between public agency’s efforts and citizens’ expectation. To bridge this ‘gap’, the four proposed improving strategies of action plan are: holding a public hearing to facilitate the public agency manager’s full understanding of citizens’ opinions or rights in improving the inadequate law limitations; establishing standard operation procedures for speedy response in dealing with citizens’ complaints on city construction design; intensifying different public departments teamwork in constructing effective and multiple needs satisfying designs and plans; and developing evaluation mechanisms in improving outsourcing quality in term of enhancing supplementary measures of the city construction. These are appropriate strategies of action plan to deal with inadequate city construction design and achieve citizens’ expectation. 4.2.4.2. Proposed actions to ineffective operation procedures. Ineffective operation procedures slow down public agency services and cause citizens’ complaints. To improve operation procedures of the public agency, the six proposed strategies of action plan are: developing multiple channels with information communication technology (ICTs) to maintain high accessibility to citizens’ complaints and opinions, which reduces the public agency’s response time; establishing information system to integrate services from different departments, which reduces operation procedures and improves service effectiveness to satisfy citizens; developing an automatic-response system to ensure timely and appropriate response to customers; establishing a mechanism of cross-functional teamwork to provide diverse services; providing employee training on ICTs and other electronic service in reducing the waiting time of citizens; and
changing standard operation procedures according to the survey of citizens’ needs. These strategies can facilitate the public agencies to improve effectiveness of the operation procedures, and improve the relationship between the employee and citizens, which facilitates citizens to agree the public agencies’ efforts. 4.2.4.3. Proposed actions to apathetic service interface. The other complaint cause from citizens is the apathetic service interface of public agency employees. Despite of the public agency provide services which cannot achieve citizens’ expectation, employee’s enthusiastic service attitude will assist in minimizing citizen complaint toward inadequate designs and ineffective operations caused by the public agency. Therefore, improvement of service attitude is a critical remedy for citizens’ dissatisfaction. The five improving employees’ service attitude strategies are: developing a performance-evaluation system to reward frontline service providers who demonstrate good service attitude in customer-oriented service-related activities; offering training courses to reinforce proficient customer-oriented service skills to deal with customers’ complaints and to provide empathic services; establishing standard operation procedures in responding to citizens’ complaints, suggestions, and opinions to achieve consistent service quality; building a customer-oriented service environment to citizens such as: offering comfortable waiting areas, free tea, clean bathroom, clear instruction for each operation procedures and so on; providing ‘one-stop’ service centres (both a traditional service desk and an electronic online service) to provide convenient services-allowing citizens to be serviced in the public agency or online from home; including the public agency dispatches service representatives to service citizens in community centres and in township administrative office. These improving strategies can help the public agency to improve employees’ service attitude as well as redeem the insufficient services, and allow citizens to really consent to the public agency’s efforts. 5. Conclusions The ‘PARA’ (primary diagnosis, advanced diagnosis, review, action) model presented in this paper seeks to have the voice of customer (VOC) taken into account in the diagnosis of failures within a service system. Based on the data mining analysis of a customer complaint database, the model enables a comprehensive diagnosis of service failures and develops appropriate improvement actions. The empirical case study reported in this paper has demonstrated the practical efficacy of the four stages of the ‘PARA’ model in diagnosing service failure and developing an improvement strategy in the context of regional public-sector services in Taiwan. The
792
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
study has demonstrated three advantages of the ‘PARA’ model compared with extant organisational diagnostic models: The core concept of the ‘PARA’ model is customer focus, which differs from the managerial and consultant perspective adopted by most other diagnostic models. The ‘PARA’ model thus addresses the gaps that exist between a service system and the real needs of customers. The proposed model offers a comprehensive data mining of customer complaint databases, rather than dealing with customer complaints on a case-by-case basis. The ‘PARA’ model integrates the views of customers, managers, and consultants in developing a consolidated improvement strategy of actions to eliminate the causes of service failure. Several suggestions for further research arise from the present study. First, in the ‘primary diagnosis’ stage of the model, future research might choose to use different approaches to ascertain the VOC, rather than the database of customer complaints used in the present study. Secondly, future research might choose to use other techniques of artificial intelligence—such as artificial nerve network (ANN), TRIZ theory, or text-mining techniques—to analyse the VOC in the advanced diagnosis stage of the ‘PARA’ model. Finally, the ‘PARA’ model might be applied in other service settings, such as service systems in the private sector or non-profit organisations. Acknowledgements This research is supported by National Science Council, Taiwan (NSC-97-2221-E-155-051-MY3). We also thank Miss. Hsiu-Chu Liang (Section leader in Tao-Yuan County Government, Taiwan) and colleagues for their assistance. References Alderfer, C. P. (1981). The marketing orientation of OD. The Journal of Applied Behavioral Science, 17, 324–325. Ashmos, D. P., & Huber, G. P. (1987). The systems paradigm in organizational theory: Correcting the record and suggesting the future. Academy of Management Review, 12(4), 607–621. Barlow, J., & Moller, C. (1996). A Complaint Is A Gift. Using Customer Feedback As A Strategic Tool. San Francisco: Berrett-Koehler Publisher. Bartunek, J. M. (1993). The multiple cognitions and conflicts associated with second order organizational change. In J. K. Murnighan (Ed.), Social Psychology in Organizations: Advances in Theory and Research (pp. 322–349). NJ: Prentice Hall. Bolman, L. G., & Deal, T. E. (1991). Reframing Organizations: Artistry. Choice and Leadership. San Francisco: Jossey-Bass. Bosch, V. G., & Enríquez, F. T. (2005). TQM and QFD: Exploiting a customer complaint management system. International Journal of Quality and Reliability Management, 22(1), 30–37. Boshoff, C. (1997). An experimental study of service recovery options. International Journal of Service Industry Management, 8(2), 110–130. Boshoff, C., & Leong, J. (1998). Empowerment, attribution and apologising as dimensions of service recovery. An experimental study. International Journal of Service Industry Management, 9(1), 24–47. Chaston, I. (1993). Delivering customer satisfaction within SME client-banker relationship. The Service Industries Journal, 13(1), 98–111. Chen, C. K., Yu, C. H., & Chang, H. C. (2005). An empirical analysis of customeroriented service activities in the Taiwanese public sector. Total Quality Management and Business Excellence, 16(7), 887–901. Chen, C. K., Yu, C. H., & Chang, H. C. (2006). ERA model: A customer-orientated organizational change model for the public service. Total Quality Management and Business Excellence, 17(10), 1301–1322. Chen, C. K., Yu, C. H., Yang, S. J., & Chang, H. C. (2004). A customer-oriented serviceenhancement system for the public sector. Managing Service Quality, 14(5), 414–425. Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20(1), 65–91.
Frawley, W., Piatetsky-Shapiro, G., & Matheus, C. (1992). Knowledge discovery in databases: An overview. Ai Magazine, 13(3), 57–70. Gustafsson, A., Ekdahl, F., & Edvardsson, B. (1999). Customer focused service development in practice: A case study at Scandinavian Airlines System (SAS). International Journal of Service Industry Management, 10(4), 344–358. Hall, R. H. (1999). Organizations: Structures, Processes, and Outcomes (7 ed.). NJ: Prentice Hall. Hand, D., Mannila, H., & Smyth, P. (2001). Principles of Data Mining. Cambridge: MIT Press. Harrison, M. I. (1994). Diagnosing Organizations: Methods, Models, and Processes (2 ed.). London: Sage Publications. Harrison, M. I., & Shirom, A. (1999). Organizational Diagnosis and Assessment: Bridging Theory and Practice. Thousand Oaks: Sage Publications. Hart, C. W., Heskett, J. L., & Sasser, W. E. Jr., (1990). The profitable art of service recovery. Harvard business review, 68(4), 148–156. Hoffman, K. D., Kelley, S. W., & Rotalsky, H. M. (1995). Tracking service failures and employee recovery efforts. Journal of Services Marketing, 9(2), 49–62. Hsia, T. C., Shie, A. J., & Chen, L. C. (2008). Course planning of extension education to meet market demand by using data mining techniques – An example of Chinkuo technology university in Taiwan. Expert Systems with Applications, 34(1), 596–602. Hung, H., & Wong, Y. H. (2007). Organisational perception of customer satisfaction: Theories and evidence. Service Industries Journal, 27(4), 495–507. Jackson, M. C. (1991). Systems Methodology for the Management Sciences. New York: Plenum Press. Johnston, R., & Fern, A. (1999). Service recovery strategies for single and double deviation scenarios. Service Industries Journal, 19(2), 69–82. Karatepe, O. M. (2006). The effects of selected antecedents on the service recovery performance of frontline employees. Service Industries Journal, 26(1), 39–57. Kassarjian, H. H. (1977). Content analysis in consumer research. Journal of Consumer Research, 4(JUNE), 8–18. Keaveney, S. M. (1995). Customer switching behavior in service industries: An exploratory study. Journal of Marketing, 59(2), 71–82. Kelley, S. W., & Davis, M. A. (1994). Antecedents to customer expectations for service recovery. Journal of the Academy of Marketing Science, 22(1), 52–61. Kelley, S. W., Hoffman, K. D., & Davis, M. A. (1993). A typology of retail failures and recoveries. Journal of Retailing, 69(4), 429–452. Kolbe, R. H., & Burnett, M. S. (1991). Content-analysis research: An examination of applications with directives for improving research reliability and objectivity. Journal of Consumer Research, 18, 243–250. Maxham Iii, J. G. (2001). Service recovery’s influence on consumer satisfaction, positive word-of-mouth, and purchase intentions. Journal of Business Research, 54(1), 11–24. McColl-Kennedy, J. R., Daus, C. S., & Sparks, B. A. (2003). The role of gender in reactions to service failure and recovery. Journal of Service Research, 6(1), 66–82. McColl-Kennedy, J. R., & Sparks, B. A. (2003). Application of fairness theory to service failures and service recovery. Journal of Service Research, 5(3), 251–266. Morgan, G. (1986). Images of organizations. CA: Sage. Nadler, D., & Tushman, M. (1980). A model for diagnosing organizational behavior. Organizational Dynamics, 9(2), 35–51. Savage, G. T., Nix, T. W., Whitehead, C. J., & Blair, J. D. (1991). Strategies for assessing and managing organizational stakeholders. Academy of Management Executive, 5(2), 61–75. Senge, P. M. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. New York: Doubleday. Smith, A. K., Bolton, R. N., & Wagner, J. (1999). A model of customer satisfaction with service encounters involving failure and recovery. Journal of Marketing Research, 36(3), 356–372. Sparks, B. A., & McColl-Kennedy, J. R. (1998). The application of procedural justice principles to service recovery attempts: Outcomes for customer satisfaction. Advances in Customer Research, 25, 156–161. Spreng, R. A., Harrell, G. D., & Mackoy, R. D. (1995). Service recovery: Impact on satisfaction and intentions. Journal of Services Marketing, 9(1), 15–23. Tax, S. S., & Brown, S. W. (1998). Recovering and learning from service failure. Sloan Management Review, 40(1), 75–88. Tax, S., & Brown, S. (2000). Service recovery: Research insights and practices. Handbook of Services Marketing and Management, 271–286. Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences: Implications for relationship marketing. Journal of Marketing, 62(2), 60–76. Varela-Neira, C., Vázquez-Casielles, R., & Iglesias-Arguëlles, V. (2008). The influence of emotions on customer’s cognitive evaluations and satisfaction in a service failure and recovery context. Service Industries Journal, 28(4), 497–512. Weber, R. P. (1985). Basic Content Analysis. London: Sage. Wirtz, J., & Mattila, A. S. (2004). Consumer responses to compensation, speed of recovery and apology after a service failure. International Journal of Service Industry Management, 15(2), 150–166. Yu, Y. T., & Dean, A. (2001). The contribution of emotional satisfaction to consumer loyalty. International Journal of Service Industry Management, 12(3), 234–250. Zemke, R., & Bell, C. (1990). Service recovery: Doing it right the second time. Training, 27(6), 42–48.
Contents lists available at ScienceDirect
Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa
A customer-oriented organisational diagnostic model based on data mining of customer-complaint databases Chi-Kuang Chen a,⇑, An-Jin Shie a, Chang-Hsi Yu b a b
Department of Industrial Engineering & Management, Yuan Ze University, Taiwan, ROC Department of Business Administration, Yu Da University, Taiwan, ROC
a r t i c l e
i n f o
Keywords: Organisational diagnosis Customer complaints Customer-orientated service Data mining Service system
a b s t r a c t The purpose of this paper is to develop a customer-oriented organisational diagnostic model, ‘PARA’ model, based on data mining of customer-complaint databases. The proposed ‘PARA’ model, which is designed to diagnose and correct service failures, takes its name from the initial letters of the four analytical stages of the model: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. In the primary-diagnosis stage, the customer-complaint database is comprehensively analysed to identify themes and categories of complaints. In the advanced-diagnosis stage, a data-mining technique is employed to investigate the relationship between the categories of customer complaints and the deficiencies of the service system. In the review stage, the identified weaknesses of the service system are reviewed and awareness of these weaknesses is enhanced among the organisation’s employees. In the action stage, a strategy of action plans for improvement is developed. An empirical case study is conducted to demonstrate the practical efficacy of the ‘PARA’ model. The paper concludes by summarising the advantages of the proposed model and the implications for future research. Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.
1. Introduction Customer satisfaction is recognised as one of the most important key performance indicators of success. However, it is always difficult to eliminate all causes of customer dissatisfaction and complaints. Customer satisfaction is influenced by such a variety of factors—including the attributes of a product or service, the individual needs of customers, and the service quality provided by front-line personnel—that even a fully ‘customer-focused’ service program cannot eliminate all product or service failures. Most organisations are aware that service failures must be handled appropriately to avoid harm to the organisation’s goodwill and profits (Hart, Heskett, & Sasser, 1990). Service recovery has thus become an increasingly important issue to prevent the loss of customers (Kelley & Davis, 1994; McColl-Kennedy & Sparks, 2003; McColl-Kennedy, Daus, & Sparks, 2003; Sparks & McColl-Kennedy, 1998; Varela-Neira, Vázquez-Casielles, & Iglesias-Arguëlles, 2008). The term ‘service recovery’ refers to remedial actions that are taken to re-establish the satisfaction of customers when product or service failure has occurred (Chaston, 1993; Zemke & Bell, 1990). Many studies have demonstrated that effective service recovery ⇑ Corresponding author. Address: Department of Industrial Engineering & Management, Yuan Ze University, 135 Yuan-Tung Road, Chung-Li 320, Taiwan, ROC. Tel.: +886 3 4638800x2528; fax: +886 3 4638907. E-mail address: [email protected] (C.-K. Chen).
can transform negative evaluations into positive impressions, thus maintaining good relationships with customers (Hoffman, Kelley, & Rotalsky, 1995; Karatepe, 2006; Kelley, Hoffman, & Davis, 1993; Maxham Iii, 2001; Smith, Bolton, & Wagner, 1999; Sparks & McColl-Kennedy, 1998; Spreng, Harrell, & Mackoy, 1995). Appropriate service recovery has also been shown to enhance the trust of customers and increase their willingness to re-purchase (Hung & Wong, 2007; Maxham Iii, 2001; Spreng et al., 1995; Tax & Brown, 1998, 2000; Yu & Dean, 2001). Conversely, ineffective service recovery is one of the main causes of switching behaviour (Keaveney, 1995). Most studies of service recovery (Barlow & Moller, 1996; Boshoff, 1997; Boshoff & Leong, 1998; Johnston & Fern, 1999; Keaveney, 1995; Tax & Brown, 1998, 2000; Wirtz & Mattila, 2004) have focused on the effectiveness of specific remedial actions—such as exchanges of goods, apologies, or offers of compensation. Relatively few have studied the question of how to improve the service system by learning from the experiences of previous service failures and avoiding repetitions. It is the contention of the present study that the prevailing focus on remedial actions and compensation for a service failure is essentially a passive and reactive approach to the problem of service failure, whereas efforts to improve the existing service system represent a creative and proactive strategy. To improve a service system and minimise service failures, it is necessary to collect and analyse customer-complaint data
0957-4174/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2011.07.074
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
periodically and comprehensively. Although some studies have indicated that improvement actions should be based on customer complaints (Bosch & Enríquez, 2005; Gustafsson, Ekdahl, & Edvardsson, 1999; Tax & Brown, 2000; Tax, Brown, & Chandrashekaran, 1998), these authors did not propose a comprehensive model to identify the sources of service failures. Chen, Yu, and Chang (2005), Chen, Yu, and Chang (2006), Chen, Yu, Yang, and Chang (2004), who have investigated the importance of service-system design and management, also emphasised the need to make better use of customer-complaint databases to diagnose failures in service systems. Against this background, the present study seeks to establish a customer-oriented organisational diagnostic model of service failure based on customers’ complaints. The proposed ‘PARA’ model takes its name from the initial letters of the four stages of the model: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. The diagnostic model provides a systematic analysis of service failures based on the customer complaint database. Data-mining techniques are then utilised to establish correlations between the identified categories of customer complaints. The model then develops a strategy of improvement actions for the service system. The model provides constructive customer-focused recommendations for improvements in service delivery through scientific analyses of service failures. The remainder of this paper is organised as follows. The next section reviews the relevant literature on organisational diagnosis. The development of the proposed ‘PARA’ model is then presented. The practical efficacy of the proposed model is then demonstrated in an empirical case study of public-sector services in Taiwan. The paper concludes with a summary of the main advantages of the model and the implications for future research.
2. Literature review of organisational diagnosis The term ‘organisational diagnosis’ is commonly used to refer to a process whereby an external consultant enters an organisation, collects valid data about human experiences within the organisation, and feeds that information back to the organisation to promote increased understanding of the organisation by its members (Alderfer, 1981). The purpose of organisational diagnosis is to establish a widely shared understanding of an organisation, and, based upon that understanding, to determine whether change is desirable. An effective process of organisational diagnosis and intervention promotes congruence between organisational objectives and the structure required to achieve those objectives (Nadler & Tushman, 1980). To achieve such an effective organisational diagnosis, Harrison (1994) suggested that diagnostic practitioners must attend simultaneously to three distinctive facets of the diagnostic process: (i) analysis; (ii) methods; and (iii) interactions. Four approaches to organisational diagnosis are especially worthy of consideration: (i) ‘sharp image’; (ii) ‘open system’; (iii) ‘political’; and (iv) ‘customer complaint’. Harrison and Shirom (1999) proposed a distinctive four-step approach to organisational diagnosis termed ‘sharp-image diagnosis’, which begins with a broad view of the organisation and proceeds to a tightly focused diagnosis of critical problems and challenges. The first step in ‘sharp image diagnosis’ is ‘scouting’, which seeks to clarify the nature of service failures in the organisation and to develop a preliminary view of the organisation’s strengths and weaknesses. In the second diagnostic step, the core problems and challenges are organised as a reference for the examination of other parts of the organisation. In the third diagnostic step, one or more focused models are developed to shape the organisation’s response to its critical challenges. In the fourth
787
diagnostic step, consultants provide clients with an emergent diagnostic model that incorporates the factors of time, resources, reward, and information feedback. According to Harrison and Shirom (1999), ‘sharp-image diagnosis’ bridges theory and practice by responding directly to the distinctive conditions that shape a particular organisation’s operations and its options for change. The outcome of ‘sharp-image diagnosis’ is a model that provides clients with a highly focused, multi-dimensional image of conditions underlying basic problems and critical challenges. The second approach to organisational diagnosis is the ‘open system’ perspective, which views the organisation as a system that obtains inputs from its environment, processes those inputs, and then produces outputs. According to ‘open-system theory’, systems have a tendency to run down if not provided with additional resources from the environment and a tendency to remain in a state of equilibrium if not disturbed (Ashmos & Huber, 1987). The opensystem framework analyses any organisation in terms of the flows of inputs, the processing of those inputs, and the creation of outputs. According to Jackson (1991), organisational diagnosis in accordance with a ‘systems approach’ proceeds by a holistic approach that examines the overall environmental and organisational contexts within which problems arise and within which steps toward organisational improvement are enacted. Systembased diagnosis thus seeks to distinguish between the symptoms of ineffectiveness and the underlying systemic causes of ineffectiveness. According to Senge (1990), this approach also enables organisations to identify the possible side-effects and unintended consequences of remedial actions. However, there are also significant limitations to the open-system approach to organisational diagnosis. In particular, some of the principles of the systems approach are too abstract to be useful, which can lead to the adoption of a superficial approach that overlooks important details of particular organisational operations and ignores the significant differences that exist among organisational contexts. The third approach to organisational diagnosis is to treat an organisation as a political arena in which bargaining and exchange take place among internal and external stakeholders who are seeking their own particular benefits or goals (Bolman & Deal, 1991; Hall, 1999; Morgan, 1986). According to this view of organisational diagnosis, the stakeholders’ involvement in the activities of the organisation can be harmful or beneficial for themselves and/or for the organisation itself (Donaldson & Preston, 1995). Stakeholders include shareholders, staff, customers, suppliers, and so on. They can be categorised as ‘main stakeholders’ (who usually maintain a formal contractual relationship with the organisation and have direct economic influence on it) and ‘secondary stakeholders’ (who include all parties that can influence, or be influenced by, the organisation) (Savage, Nix, Whitehead, & Blair, 1991). In conducting organisational diagnosis in accordance with the political perspective, the following steps are taken (Savage et al., 1991): (i) identify the main stakeholders; (ii) analyse each stakeholder’s position; (iii) examine each stakeholder’s power; and (iv) assess each stakeholder’s capacity for action and impact. The political perspective thus emphasises interpersonal relationships within the organisation, the potential for conflicts among stakeholders, and the influence of stakeholder activity on the budget and other forms of resource distribution (Bartunek, 1993). The role of customers (in their capacity as stakeholders) thus plays some part in this model of organisational diagnosis; however, this approach lacks an operational model with established procedures to take proper account of the opinions of all customers as stakeholders. The fourth approach to organisational diagnosis utilises customer complaints as the driving force for analysis. Bosch and Enríquez (2005) developed a customer-oriented model of organisational diagnosis by incorporating total quality management (TQM), quality function deployment (QFD), and plan-do-check-action
788
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
(PDCA) in their so-called ‘customer complaint management system’ (CCMS). This diagnostic model involved a seven-step procedure: (i) document the voice of the customer (VOC); (ii) translate VOC into customer demands and problems; (iii) analyse and solve these questions; (iv) explore customer demands; (v) update failure mode and effect analysis (FMEA) to avoid recurrence; (vi) share solutions with affected customers; and (vii) update system performance measurements. According to Bosch and Enríquez (2005), the CCMS model can change an organisation’s perspective on complaint management and transform the process of answering complaints from a trivial activity to an exciting learning experience in which continuous improvement leads to service excellence. However, although this model does take into account the value of customer complaints, it deals with individual customer complaints on a case-by-case basis and does not retrieve useful information from all customer complaints to create a comprehensive basis on which to develop diagnostic strategies for improving the organisation’s service system. This review of the relevant literature has revealed that most of the extant organisational diagnostic models have noted the importance of customer-oriented service. However, the focus and the operational procedures of these models cannot guarantee that the complaints of all customers will be taken into account. The present study addresses the shortcoming in constructing a comprehensive customer-oriented model for organisational diagnosis. The data mining techniques is used to analyze customer complaint database. 3. The ‘PARA’ diagnostic model As noted above, the present study contends that a proactive approach needs to be adopted in dealing with customer complaints and identifying the ‘voice of the customer’ (VOC). In accordance with this view, a customer-oriented organisational diagnostic model based on a comprehensive analysis of customer complaints is developed in this study. The proposed model, which is illustrated in Fig. 1, is designated the ‘PARA’ model from the initial letters of the four stages in the
model’s diagnostic procedure: (i) ‘primary diagnosis’; (ii) ‘advanced diagnosis’; (iii) ‘review’; and (iv) ‘action’. The details of each of these stages of the ‘PARA’ model are described below. The purpose of the first stage, primary diagnosis, is to organise the firm’s original database of customer complaints into categories, and then to analyse these data in terms of the frequencies of complaints in each category. Because the data collected in such databases is usually disorganised and uncategorised, content analysis is used to provide a coherent structure. In contrast to the CCMS model (Bosch & Enríquez, 2005) described above, the ‘PARA’ model utilises a comprehensive analysis of customers’ opinions, rather than analysing individual customer complaints as isolated cases. In the second stage, advanced diagnosis, data mining is used to seek correlations among the categories of customer complaints. Data mining has been described as ‘‘the nontrivial extraction of implicit, previously unknown, and potentially useful information from data’’ (Frawley, Piatetsky-Shapiro, & Matheus, 1992) and ‘‘the science of extracting useful information from large datasets or databases’’ (Hand, Mannila, & Smyth, 2001; Hsia, Shie, & Chen, 2008). The results of the data-mining analyses (in providing correlations among the categories of complaints) are essential to the development of the improvement strategy of action plans in the next stage of the model. In the third stage, review, the correlations among the various categories of customer complaints are examined to seek correlations’ causes. Apart from confirming the causes of the correlations, the purpose of this stage of the model is to raise the awareness of the pertinent issues among managers and their subordinates. To achieve this purpose, a checklist of items derived from the datamining analyses is presented to staff members, who are encouraged to examine the identified service weaknesses and its causes through open discussion. The fourth stage, action, refers to the development of strategies of action plans for improvement of the service system. In this stage, managers work intensively with consultants to devise improving plans to improve the service system according to customer complaint causes proposed in stage 3. The customer, manager
Fig. 1. PARA model of organisational diagnosis.
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
and consultant’ opinions are taken into account to develop the most comprehensive improving plans. The above process enables the organisation to achieve its improvement goals more efficiently and effectively. In summary, the proposed ‘PARA’ model integrates customer complaints, management opinion, and consultant expertise to develop a constructive improvement strategy that enhances service delivery and organisational efficiency. Compared with the extant models of organisational diagnosis noted above, the proposed ‘PARA’ model has several advantages. In particular the new model is able to: (i) convert customer complaints into ready-to-use strategies for the improvement of a service system; (ii) bridge the gap between an organisation’s service and the demands of its customers; and (iii) promote management’s awareness of problems and potential directions for improvement.
4. Empirical case study 4.1. Research setting To demonstrate the efficacy of the proposed ‘PARA’ model, an empirical case study was conducted using the complaint database of public-sector service agencies in Tao-Yuan county, Taiwan. The aim of this exercise in organisational diagnosis was to utilise the complaints of citizens to ascertain the reasons for failures in the service systems of the Tao-Yuan county government agencies. More than 80 government agencies provide a variety of services (including household registration, land management, social welfare, public facility construction, and business management) to the 1.9 million citizens of Tao-Yuan county. The county governor has established a standardised procedure for handling the complaints of citizens. Under the management system, the responsible agency and the designated staff are expected to follow up all complaints on a daily basis. The complaint-management system was computerised in 2006, which enabled citizens to make their complaints online. The standard procedure of the complaints-management system consists of three steps: (i) complaints are sent to the complaintmanagement system; (ii) the complaints are categorised and each complaint is passed to the department responsible for the issue; and (iii) the responsible department should reply to the complainant promptly regarding the proposed action and timeframe for improvement of the service that is the subject of the complaint. The system includes monitoring to ensure that staffs complete the required tasks within three days of the complaint being received. At the time of the present study, the system had encountered significant difficulties. The computerisation of the complaint system in 2006 proved to be so convenient for citizens that the number of complaints increased dramatically. Because complaints were handled on a case-by-case basis, it soon became impossible for government managers to handle all the received complaints within the prescribed time. Moreover, the increased complaint-handling work load meant that agency managers did not have time to search the complaint database for clues to the underlying causes of service-system failures. It was apparent that there was a need for a diagnostic model that was capable of analysing the complaint data systematically and comprehensively with a view to diagnosing systemic service failures within the entire organisation. In response to this situation, the complaint-management system of the public services of Tao-Yuan county was chosen for application of the proposed ‘PARA’ diagnostic model. Customercomplaint data gathered by the system from 1 October 2006 to 30 September 2007 were used as the database for the study. A random selection of 500 records was then analyzed in the study after
789
excluding repeated complaints from the same citizen and complaints with incomplete or unclear descriptions of problems. 4.2. Analysis by ‘PARA’ model 4.2.1. Stage 1: primary diagnosis The primary-diagnosis stage of the ‘PARA’ model employed the technique of ‘content analysis’ to categorise the unstructured data for systematic analysis of customer complaints. A total of 983 themes were initially extracted by content analysis. These themes were then classified into 18 generic categories. Two external coders were invited to examine the reliability of the content analysis. A random sample of 20% of the 983 themes extracted from the data were examined by the two external coders. The degree of mutual agreement was computed as follows:
Degree of mutual agreement ¼
2 number of items completely agreed by two parties number agreed by Party A þ number agreed by Party B
The degrees of mutual agreement between the original content analysis and external coder 1 was 0.864 and that between the original analysis and external coder 2 was 0.77. A reliability index was then computed as follows:
Reliability ¼
2 Degree of average mutual agreement 1 þ ½ð2 1Þ Degree of average mutual agreement
The reliability index in this study was 0.9, which was greater than the threshold value suggested by Kassarjian (1977). The result thus indicated that the content analysis was highly reliable (Kolbe & Burnett, 1991; Weber, 1985). Table 1 shows the customer-complaint categories as derived by the primary diagnosis. Four categories of complaints were most prominent: (i) inadequate public facility design or construction (21.6% of all complaints); (ii) environmental hazards (11.7%); (iii) unreasonable traffic tickets (10.6%); and (iv) ignorance of citizens’ rights (8.0%). These four complaint categories clearly required the closest attention from management. 4.2.2. Stage 2: advanced diagnosis To identify correlations among the 18 customer-complaint categories, the ‘Apriori’ algorithm was utilised as a data-mining technique. The complaint data were first classified with a Boolean-type code (‘1’ or ‘0’). The Apriori algorithm was then applied using Eq. (1) (below) to calculate the confidence from the two supports:
Antecedent [ Consequence Items Support 100 Antecedent Items Support ¼ Confidence ð%Þ
ð1Þ
in which: ‘antecedent items support’ = the percentage of data that tallies with the antecedent rule; and ‘antecedent [ consequence items support’ = the percentage of data that tallies with the union of antecedent items and consequence items support rules. In applying the Apriori algorithm, a minimum level of confidence needs to be defined as a mining threshold. In this case, the minimum level of confidence was set at 80%. Table 2 presents the results of the Apriori algorithm analysis. Three correlation rules were identified beyond 80% (calculated according to the number of antecedent and consequence items). These three correlation rules covered 153 customer-complaint records in total (with rule 1 covering 94 complaint records, rule 2 covering 12, and rule 3 covering 47), which represents 31% of the 500 customer-complaint records included in the study.
790
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
Table 1 Primary diagnosis of customer complaint categories. Customer complaint categories
Frequency
1. Inadequate public facility design or construction 2. Building construction without licence 3. Poor environmental sanitation and hygiene conditions (e.g. garbage disposal, pet droppings on roads and stray animal management) 4. Viatical obstructions (e.g. unauthorised occupation of roads, insufficient parking spaces) 5. Environmental hazards (e.g. dangerous signboards, uneven road surface, damaged or insufficient numbers of security surveillance cameras) 6. Poor quality control of infrastructure constructions (e.g. flooding, water outage) 7. Noise pollution 8. Inefficient maintenance of public facilities (e.g. streetlamp maintenance, leisure park facilities, venues for exercises, bus, traffic signs, sewerage maintenance, roadway maintenance, government website maintenance) 9. Failure to keep promises 10. Unreasonable traffic tickets 11. Slow administrative procedure 12. Reluctant administrative procedures 13. Insufficient professionalism of public servants 14. Poor service attitudes or absence on duty 15. Ignorance of citizens’ rights 16. Reluctance to take responsibilities 17. Failure to provide out-of-SOP (standard operating procedures) service 18. Lack of empathic services
212 4 33
Total ⁄
30 115
% of total complaints 21.6⁄ 0.4 3.4 3.1 11.7⁄
6
0.6
9 36
0.9 3.7
77 104 77 1 9 49 79 42 22
7.8 10.6⁄ 7.8 0.1 0.9 5.0 8.0⁄ 4.3 2.2
78
7.9
983
100.0
indicates the first four highest figures
Rule 1 proposed that ‘inadequate public facility design or construction’ and ‘ignorance of citizens’ rights’ caused citizens to feel that the service provided by the public agency demonstrated a ‘lack of empathy’. The confidence level of this rule was 100%. Rule 2 proposed that a ‘failure to keep promises’ regarding ‘environmental hazards’ was associated with ‘slow administrative procedure’. The confidence level of this rule was also 100%. Rule 3 proposed that ‘unreasonable traffic tickets’ and a ‘lack of empathic services’ was associated with ‘poor service attitudes or absence on duty’. The confidence level of this rule was 80%. 4.2.3. Stage 3: review In stage 3, the results derive from data mining in stage 2 are further reviewed in regard to the causes and effects. It is to identify the service weaknesses and its causes. In this empirical case, rule 1 is selected as a target example for the review because rule 1 involves the most citizen complaint cases (94 cases, see Table 2) with the highest confidence level 100%. The 94 complaint cases are illustrated in the three aspects after one by one review the contents of complaint raised by citizen. They are inadequate city construction design, insufficient operation procedures and apathetic service interface. The three aspects concurrently exist in the service system which results in the 94 citizen complaints cases being raised. The aspects are discussed in details as follows. 4.2.3.1. Inadequate city construction design. Basing on the 94 complaint case-reviews, insufficient city construction designs are mostly occurred. Lacking consideration of citizens’ needs and strict legal regulations are major causes of the insufficient designs or plans. These complaint cases include: (i) inadequate parking space area design; (ii) wrong traffic sign design; and (iii) inadequate road planning. These public agency inadequate city construction
designs cause citizens to feel ignored of their rights and needs, such that generate repetitive complaints from the same problems. Yet the public agency merely focuses on fulfilling law regulations of the city plan without considering citizens’ opinions. The above situations commonly exist in the insufficient bureaucratic service systems of public agencies. 4.2.3.2. Ineffective operation procedures. In addition, most of the citizens frequently complain that operation procedures of the public agency are ineffective. These reasons include: (i) administrative and front-line service standard operation procedures are too verbose to immediately satisfy citizens’ needs; (ii) lack of supplementary measure diminishes the public agency policy efficiency; (iii) operation procedures are restricted by the existent law regulations, such that cannot avoid unnecessary procedures; and (iv) operation procedures lack in integrating services from different departments, which causes citizens feel inconvenient. These reasons usually derive inefficient performance in a bureaucratic service system of the public agency. 4.2.3.3. Apathetic service attitude interface. Apart from the noted complaint reasons above, apathetic service interface of public agency is another major complaint from the citizens. The service interface is that front-line service providers present service attitude and enthusiasm. These complaint cases include: (i) ignoring citizens’ opinions and rights; (ii) presenting unpleasant service attitude; and (iii) lacking empathic service attitude. These illustrate that the public agency services do not conform citizens’ expectation and are ineffective in handling citizens’ opinions and complaints. In fact, most of the public agencies present a defensive attitude in dealing with citizens’ opinions and complaints. Ignoring citizens’ opinions will result in the public agencies’ repeating same faulty service. Besides, the public agency employees often too conveniently rely on previously established standard operation
791
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792 Table 2 Apriori algorithm analysis. Rule ID 1 2 3
Antecedent
Inadequate public facility design or construction Ignorance of citizens’ rights Environmental hazards Failure to keep promises Unreasonable traffic tickets Lack of empathic services
Consequence
Confidence level (%)
Number of complaint cases
Lack of empathic services
100
94
Slow administrative procedure
100
12
Poor service attitudes or absence on duty
80
47
procedures and unwillingly provide flexible service outside of established procedure, to immediately satisfy variety of citizens’ needs. All these reasons generate gaps between a service system and the real needs of customers. 4.2.4. Stage 4: action In this action stage, the public agency managers and scholastic consultants consolidate and analyze the results from stage 3 to figure out improvement strategies of action plan. The study integrated customers, managers and consultants’ opinions and analysis to develop strategies for improving service system of the public agency and minimizing citizens’ complaints. These proposed improving strategies are divided into three categories coincide with stage 3, as follows: 4.2.4.1. Proposed actions to inadequate city construction design. As pointed out in stage 3, inadequate city construction design generates a gap between public agency’s efforts and citizens’ expectation. To bridge this ‘gap’, the four proposed improving strategies of action plan are: holding a public hearing to facilitate the public agency manager’s full understanding of citizens’ opinions or rights in improving the inadequate law limitations; establishing standard operation procedures for speedy response in dealing with citizens’ complaints on city construction design; intensifying different public departments teamwork in constructing effective and multiple needs satisfying designs and plans; and developing evaluation mechanisms in improving outsourcing quality in term of enhancing supplementary measures of the city construction. These are appropriate strategies of action plan to deal with inadequate city construction design and achieve citizens’ expectation. 4.2.4.2. Proposed actions to ineffective operation procedures. Ineffective operation procedures slow down public agency services and cause citizens’ complaints. To improve operation procedures of the public agency, the six proposed strategies of action plan are: developing multiple channels with information communication technology (ICTs) to maintain high accessibility to citizens’ complaints and opinions, which reduces the public agency’s response time; establishing information system to integrate services from different departments, which reduces operation procedures and improves service effectiveness to satisfy citizens; developing an automatic-response system to ensure timely and appropriate response to customers; establishing a mechanism of cross-functional teamwork to provide diverse services; providing employee training on ICTs and other electronic service in reducing the waiting time of citizens; and
changing standard operation procedures according to the survey of citizens’ needs. These strategies can facilitate the public agencies to improve effectiveness of the operation procedures, and improve the relationship between the employee and citizens, which facilitates citizens to agree the public agencies’ efforts. 4.2.4.3. Proposed actions to apathetic service interface. The other complaint cause from citizens is the apathetic service interface of public agency employees. Despite of the public agency provide services which cannot achieve citizens’ expectation, employee’s enthusiastic service attitude will assist in minimizing citizen complaint toward inadequate designs and ineffective operations caused by the public agency. Therefore, improvement of service attitude is a critical remedy for citizens’ dissatisfaction. The five improving employees’ service attitude strategies are: developing a performance-evaluation system to reward frontline service providers who demonstrate good service attitude in customer-oriented service-related activities; offering training courses to reinforce proficient customer-oriented service skills to deal with customers’ complaints and to provide empathic services; establishing standard operation procedures in responding to citizens’ complaints, suggestions, and opinions to achieve consistent service quality; building a customer-oriented service environment to citizens such as: offering comfortable waiting areas, free tea, clean bathroom, clear instruction for each operation procedures and so on; providing ‘one-stop’ service centres (both a traditional service desk and an electronic online service) to provide convenient services-allowing citizens to be serviced in the public agency or online from home; including the public agency dispatches service representatives to service citizens in community centres and in township administrative office. These improving strategies can help the public agency to improve employees’ service attitude as well as redeem the insufficient services, and allow citizens to really consent to the public agency’s efforts. 5. Conclusions The ‘PARA’ (primary diagnosis, advanced diagnosis, review, action) model presented in this paper seeks to have the voice of customer (VOC) taken into account in the diagnosis of failures within a service system. Based on the data mining analysis of a customer complaint database, the model enables a comprehensive diagnosis of service failures and develops appropriate improvement actions. The empirical case study reported in this paper has demonstrated the practical efficacy of the four stages of the ‘PARA’ model in diagnosing service failure and developing an improvement strategy in the context of regional public-sector services in Taiwan. The
792
C.-K. Chen et al. / Expert Systems with Applications 39 (2012) 786–792
study has demonstrated three advantages of the ‘PARA’ model compared with extant organisational diagnostic models: The core concept of the ‘PARA’ model is customer focus, which differs from the managerial and consultant perspective adopted by most other diagnostic models. The ‘PARA’ model thus addresses the gaps that exist between a service system and the real needs of customers. The proposed model offers a comprehensive data mining of customer complaint databases, rather than dealing with customer complaints on a case-by-case basis. The ‘PARA’ model integrates the views of customers, managers, and consultants in developing a consolidated improvement strategy of actions to eliminate the causes of service failure. Several suggestions for further research arise from the present study. First, in the ‘primary diagnosis’ stage of the model, future research might choose to use different approaches to ascertain the VOC, rather than the database of customer complaints used in the present study. Secondly, future research might choose to use other techniques of artificial intelligence—such as artificial nerve network (ANN), TRIZ theory, or text-mining techniques—to analyse the VOC in the advanced diagnosis stage of the ‘PARA’ model. Finally, the ‘PARA’ model might be applied in other service settings, such as service systems in the private sector or non-profit organisations. Acknowledgements This research is supported by National Science Council, Taiwan (NSC-97-2221-E-155-051-MY3). We also thank Miss. Hsiu-Chu Liang (Section leader in Tao-Yuan County Government, Taiwan) and colleagues for their assistance. References Alderfer, C. P. (1981). The marketing orientation of OD. The Journal of Applied Behavioral Science, 17, 324–325. Ashmos, D. P., & Huber, G. P. (1987). The systems paradigm in organizational theory: Correcting the record and suggesting the future. Academy of Management Review, 12(4), 607–621. Barlow, J., & Moller, C. (1996). A Complaint Is A Gift. Using Customer Feedback As A Strategic Tool. San Francisco: Berrett-Koehler Publisher. Bartunek, J. M. (1993). The multiple cognitions and conflicts associated with second order organizational change. In J. K. Murnighan (Ed.), Social Psychology in Organizations: Advances in Theory and Research (pp. 322–349). NJ: Prentice Hall. Bolman, L. G., & Deal, T. E. (1991). Reframing Organizations: Artistry. Choice and Leadership. San Francisco: Jossey-Bass. Bosch, V. G., & Enríquez, F. T. (2005). TQM and QFD: Exploiting a customer complaint management system. International Journal of Quality and Reliability Management, 22(1), 30–37. Boshoff, C. (1997). An experimental study of service recovery options. International Journal of Service Industry Management, 8(2), 110–130. Boshoff, C., & Leong, J. (1998). Empowerment, attribution and apologising as dimensions of service recovery. An experimental study. International Journal of Service Industry Management, 9(1), 24–47. Chaston, I. (1993). Delivering customer satisfaction within SME client-banker relationship. The Service Industries Journal, 13(1), 98–111. Chen, C. K., Yu, C. H., & Chang, H. C. (2005). An empirical analysis of customeroriented service activities in the Taiwanese public sector. Total Quality Management and Business Excellence, 16(7), 887–901. Chen, C. K., Yu, C. H., & Chang, H. C. (2006). ERA model: A customer-orientated organizational change model for the public service. Total Quality Management and Business Excellence, 17(10), 1301–1322. Chen, C. K., Yu, C. H., Yang, S. J., & Chang, H. C. (2004). A customer-oriented serviceenhancement system for the public sector. Managing Service Quality, 14(5), 414–425. Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20(1), 65–91.
Frawley, W., Piatetsky-Shapiro, G., & Matheus, C. (1992). Knowledge discovery in databases: An overview. Ai Magazine, 13(3), 57–70. Gustafsson, A., Ekdahl, F., & Edvardsson, B. (1999). Customer focused service development in practice: A case study at Scandinavian Airlines System (SAS). International Journal of Service Industry Management, 10(4), 344–358. Hall, R. H. (1999). Organizations: Structures, Processes, and Outcomes (7 ed.). NJ: Prentice Hall. Hand, D., Mannila, H., & Smyth, P. (2001). Principles of Data Mining. Cambridge: MIT Press. Harrison, M. I. (1994). Diagnosing Organizations: Methods, Models, and Processes (2 ed.). London: Sage Publications. Harrison, M. I., & Shirom, A. (1999). Organizational Diagnosis and Assessment: Bridging Theory and Practice. Thousand Oaks: Sage Publications. Hart, C. W., Heskett, J. L., & Sasser, W. E. Jr., (1990). The profitable art of service recovery. Harvard business review, 68(4), 148–156. Hoffman, K. D., Kelley, S. W., & Rotalsky, H. M. (1995). Tracking service failures and employee recovery efforts. Journal of Services Marketing, 9(2), 49–62. Hsia, T. C., Shie, A. J., & Chen, L. C. (2008). Course planning of extension education to meet market demand by using data mining techniques – An example of Chinkuo technology university in Taiwan. Expert Systems with Applications, 34(1), 596–602. Hung, H., & Wong, Y. H. (2007). Organisational perception of customer satisfaction: Theories and evidence. Service Industries Journal, 27(4), 495–507. Jackson, M. C. (1991). Systems Methodology for the Management Sciences. New York: Plenum Press. Johnston, R., & Fern, A. (1999). Service recovery strategies for single and double deviation scenarios. Service Industries Journal, 19(2), 69–82. Karatepe, O. M. (2006). The effects of selected antecedents on the service recovery performance of frontline employees. Service Industries Journal, 26(1), 39–57. Kassarjian, H. H. (1977). Content analysis in consumer research. Journal of Consumer Research, 4(JUNE), 8–18. Keaveney, S. M. (1995). Customer switching behavior in service industries: An exploratory study. Journal of Marketing, 59(2), 71–82. Kelley, S. W., & Davis, M. A. (1994). Antecedents to customer expectations for service recovery. Journal of the Academy of Marketing Science, 22(1), 52–61. Kelley, S. W., Hoffman, K. D., & Davis, M. A. (1993). A typology of retail failures and recoveries. Journal of Retailing, 69(4), 429–452. Kolbe, R. H., & Burnett, M. S. (1991). Content-analysis research: An examination of applications with directives for improving research reliability and objectivity. Journal of Consumer Research, 18, 243–250. Maxham Iii, J. G. (2001). Service recovery’s influence on consumer satisfaction, positive word-of-mouth, and purchase intentions. Journal of Business Research, 54(1), 11–24. McColl-Kennedy, J. R., Daus, C. S., & Sparks, B. A. (2003). The role of gender in reactions to service failure and recovery. Journal of Service Research, 6(1), 66–82. McColl-Kennedy, J. R., & Sparks, B. A. (2003). Application of fairness theory to service failures and service recovery. Journal of Service Research, 5(3), 251–266. Morgan, G. (1986). Images of organizations. CA: Sage. Nadler, D., & Tushman, M. (1980). A model for diagnosing organizational behavior. Organizational Dynamics, 9(2), 35–51. Savage, G. T., Nix, T. W., Whitehead, C. J., & Blair, J. D. (1991). Strategies for assessing and managing organizational stakeholders. Academy of Management Executive, 5(2), 61–75. Senge, P. M. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. New York: Doubleday. Smith, A. K., Bolton, R. N., & Wagner, J. (1999). A model of customer satisfaction with service encounters involving failure and recovery. Journal of Marketing Research, 36(3), 356–372. Sparks, B. A., & McColl-Kennedy, J. R. (1998). The application of procedural justice principles to service recovery attempts: Outcomes for customer satisfaction. Advances in Customer Research, 25, 156–161. Spreng, R. A., Harrell, G. D., & Mackoy, R. D. (1995). Service recovery: Impact on satisfaction and intentions. Journal of Services Marketing, 9(1), 15–23. Tax, S. S., & Brown, S. W. (1998). Recovering and learning from service failure. Sloan Management Review, 40(1), 75–88. Tax, S., & Brown, S. (2000). Service recovery: Research insights and practices. Handbook of Services Marketing and Management, 271–286. Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences: Implications for relationship marketing. Journal of Marketing, 62(2), 60–76. Varela-Neira, C., Vázquez-Casielles, R., & Iglesias-Arguëlles, V. (2008). The influence of emotions on customer’s cognitive evaluations and satisfaction in a service failure and recovery context. Service Industries Journal, 28(4), 497–512. Weber, R. P. (1985). Basic Content Analysis. London: Sage. Wirtz, J., & Mattila, A. S. (2004). Consumer responses to compensation, speed of recovery and apology after a service failure. International Journal of Service Industry Management, 15(2), 150–166. Yu, Y. T., & Dean, A. (2001). The contribution of emotional satisfaction to consumer loyalty. International Journal of Service Industry Management, 12(3), 234–250. Zemke, R., & Bell, C. (1990). Service recovery: Doing it right the second time. Training, 27(6), 42–48.