Development of Customer Satisfaction Index (CSI) Model for Pakistan

  • HAMAYUN, Khadija (Department of Management Sciences, COMSATS University Islamabad, Abbottabad Campus) ;
  • HAFEEZ, Shakir (Department of Management Sciences, COMSATS University Islamabad, Abbottabad Campus)
  • Received : 2022.04.10
  • Accepted : 2022.07.05
  • Published : 2022.07.30


To measure economic performance, customer satisfaction indices are constructed. This study proposes an index for banking and telecom, a significant evaluative system for comparing and enhancing customer satisfaction across the industries. The study suggests and examines amendments and improvements to the prior indices and incorporates ignored indicators to propose a punier index for Pakistan. The study is a pioneer in integrating online and offline indices into a single comprehensive model. The study is enriched by the Theory of Reasoned Action and Technological Acceptance Model. A sample of 320 respondents was used. The sample was divided based on gender and marital status. To authenticate the theoretical model, PLS-SEM was applied. We discovered nine latent variables that define customer satisfaction and conclude that a single model can be utilized for e-commerce enterprises as well. The index scores are comparable to the American index for banking and the Turkish index for telecom. Multi-group analysis (MGA) was used to comprehend the differences among the groups. This reveals that customization, design, reliability, and responsiveness induce satisfaction in telecom male and married customers. For the banking industry, the difference exists in complaint handling, customization, corporate image, perceived price, reliability, responsiveness, sentiments, convenience, and security to satisfaction links, image and complaint handling to loyalty links.


1. Introduction

The significance of economic activity is defined by customer satisfaction because in the final analysis customer satisfaction matters most rather than the amount of production or consumption in an economy. The product of customer satisfaction is customer loyalty, which fallouts in profitability and high company performance. According to the marketing concept, satisfying customers is a step towards the achievement of marketing target and profitability, so organizations make the efforts and strategies to satisfy their customers (Kotler et al., 2017). Therefore, CSI is an evaluative system that compares and enhances customer satisfaction across industries (Fornell et al., 1996). The index scores explicitly work as an intangible economic indicator, which is utilized for evaluating the economic feasibility of companies, industries, and international trade (Fornell, 1992).

The CSI is mostly based on non-financial measures. Cakir et al. (2019) claimed that significant research growth in the area of retail measures is observed from 2008 to 2018. Foremost growth is contributed by indicators of non-financial from 2015 to 2018. Retail research in the financial measure is shown by a relation of 55% to 45%. Though in 2017–18 the rise in research and awareness for non-financial measures is most significant as compared to financial measures. Therefore, a continuous shift toward non-financial measures is noticed (Cakir et al., 2019; Hassan et al., 2020). Correspondingly, the prior indices ignored some noteworthy nonfinancial measures, providing a gray area for investigation. Thus, the inspiration to construct an index is determining the vital antecedents influencing customer satisfaction and loyalty.

The customer satisfaction measurement models and methods could be utilized to evolve and modify the customer satisfaction indices. The reliability and validity of CSI models made them applied in a wide variety of fields but focused on the firms, and firm customers are relatively outdated. Most of the index’s construct is based on the degree to which how services and products provide important consumer customization and how consistently promised customization is carried out (Park et al., 2008; Keskar & Panday, 2020). The indices are developing and studied in the marketing field and indices for electronic commerce are developing in Management Information systems continuously in the last decade. However, a single index for both concerns has not yet been developed and tested. Prior indices are purely for offline businesses; on the other extreme, for online businesses, there is a separate index.

The banking and telecom industries are major global economic sectors (Hassan et al., 2020). The banking sector dominates Pakistan’s financial sector -with increasing assets composition year by year (Hassan et al., 2020). These assets increased by 74% during the last eight years and contribute 7.7% (60% service sector) to GDP by the year 2021 (Economic survey of Pakistan 2020–21). Similarly, telecom is the fastest-growing industry and a crucial factor in the growth of Pakistan’s economy. Its tele density is at 82%, with 172 million mobile, 2.2 million land subscribers and international bandwidth connectivity is 3.1 terabytes. This sector will contribute 278 billion in 2020 (Pakistan telecommunication authority 2021). Thus, for these indus- tries, customer satisfaction is the most important foundation for preserving customer loyalty. Therefore, evaluating the different aspects of the service quality could pinpoint the weak areas to enhance the quality as a result of capturing loyal customers. All this can do through CSI. In Pakistan, there are no such widely accepted government criteria for gauging customer satisfaction. Besides, it is very crucial to address which index type provides the most important information and competitive advantage for the different sectors? The prior models unexploited the notions of changing environment consumerism, particularly fast growing technologies and smart services. Thus, to fill these deficiencies, there is a chronic need for the development of a new model. The proposed Pakistan customer satisfaction index (PCSI) is constructed from a consumer point of view with a concentration on technologies.

2. Literature Review

The literature review reveals that previous index models focused on a few relationship indicators like quality relationships and value relationships for demonstrating customer satisfaction and loyalty (Grigoroudis & Siskos, 2004; Deng et al., 2013; Dam & Dam, 2021). Other researchers (Gohain et al., 2018; Asyraf et al., 2019) explore the association among the numerous variables such as value, switching cost, commitment, trust, and service quality. Some other researchers (Bruhn & Grund, 2000; Virvilaite & Daubaraite, 2011) recommends expectations, complaint behavior, and corporate image. Hence, differentiating factors and their influence on customer satisfaction exist differently across countries (Grigoroudis & Siskos, 2004; Kalia et al., 2021). All the important revealed determinants are not included in a single comprehensive model to construct an index. This implies a need for more comprehensive and holistic customer satisfaction and loyalty review that covers all the ignored and significant determinants and finds additional performance satisfaction indicators at all levels of society.

2.1. Indices Evolution

The primordial national CSI was Swedish Customer Satisfaction Barometer (SCSB) established in 1989 (Fornell, 1992). This model has two satisfaction antecedents (Customer Expectation and Perceived Value) predicted to have a positive effect. The consequences are derived from Hirschman’s (1970) theory of voice and exit. Germany introduced its barometer in 1992 with a single-item approach and not based on a structural model (Meyer & Dornach, 1996). The American Customer Satisfaction (ACSI) was a cause-and-effect model (Fornell, 1992) based on the original SCSB and two well-entrenched theories (Quality, Satisfaction, and Performance (QSP) paradigm and Hirschman’s Exit Voice theory 1970). The discrepancy between SCSB and ACSI includes perceived quality, discrete from perceived value. In 1999 European Customer Satisfaction Index (ECSI) was constructed for 11 European countries based on ACSI. The model excludes complaint behavior and includes corporate Image as a latent variable. It divides perceived quality into the perceived quality of hardware and software (O’Loughlin & Coenders, 2004). Then, in 2001 Norwegian index (NCSB)was introduced with the addition of commitment and exclusion of expectation as it proved insignificant in ACSI. Johnson et al. (2001) discussed the detailed evolution of indices. Afterward, every country has worked to create its own index.

As for as the online customer satisfaction index is concerned the first step taken by (Cho & Park, 2001) with a few factors, e.g., internet shopping, consumer attitude, and creating an e-customer user satisfaction index (ECUSI). Kim (2005) has constructed a pure online index with ten customer satisfaction factors extracted from marketing, e-commerce, and management information systems literature. Hsu (2008) proposed a model of e-CSI based on ACSI. The researcher replaces trust with expectation as economic activities are based on trust. The e-SQ replaces service quality, and a new relationship between trust and loyalty was added. Tabaei and Fathian (2012) have developed an index based on customers’ needs and wants, along with a customer relationship management strategic map. Security concerns are treated as latent variables rather than items. Shin (2014) has developed an index for mobile phone users. This index is modified, advanced, and better suited to emerging smart technologies. The smart services model was evaluated using perceived utilitarian and hedonic performance aspects. Furthermore, service quality was measured by using the SERVQUAL scale. A comprehensive literature review was conducted based on e-commerce customer satisfaction from 2000–2017 and developed a conceptual model (Deyalage & Kulathunga, 2019). This review revealed fifty-four determinants of customer satisfaction; among them, five are the most cited and investigated factors affecting e-commerce customer satisfaction.

2.2. Prior Models Evaluation

National customer satisfaction indices should be modified as time, conditions, knowledge, and the environment changes. The satisfaction and loyalty constructs are enduring, but there is also no reason to trust that current models precisely describe these constructs in different situations and times. In all prior models, the quality-to-value link is particularly problematic, and interpretation of this path is difficult, especially the quality-to-value direct path. Because, in the value equation, quality is the most important part. The path’s causal part is questionable as it is impossible to know how much cause-and-effect impact by quality to value has and how much is accurate by definition. In a causal model, there must be a rationale regarding the mechanism by which one constructs influences another (Bagozzi, 1994). Expectation measures review discloses that they are specifically related to quality instead of value. Therefore, the rationale for the expectation value link is indistinct and insignificant (Johnson et al., 2001). Hence, it is proposed that perceived price replaces the value for removing tautology. Because market research suggests that price is a cue to quality rather than the opposite. Consumers consider price information and always try to get the product at a reasonable price and make purchasing decisions mostly based on reference prices. Evaluation of price based on marketing instrument is important in terms of customer’s revisit, satisfaction, loyalty, and vice versa (Bassey, 2014; Celil et al., 2019). For that purpose, businesses make efforts to improve price-quality. Whereas Priori models lack this connection. Furthermore, the direct effect of price on loyalty is incorporated too. Because the Pakistani market is particularly price-conscious, prices catch the attention of customer repurchase evaluation.

Theoretically, complaining is considered a natural consequence of low satisfaction, not an opportunity to increase satisfaction. Consumer raises their voice to complain due to the failure to deliver the promised services and cause dissatisfaction. Due to technological advancement and e-commerce, consumers feel free to complain (Stevens et al., 2018). Methodologically, the complaint handling and recovery systems certainly happen former to the customer being surveyed. In the last two-decade researchers realized the complaint management potential and recovery systems to increase satisfaction. They emphasize complaint resolution rather than complaints per se (Johnson et al., 2001). Thus, complaint handling should be the driver of satisfaction rather than a consequence. Several researchers have found theoretically and empirically significant relationships between complaint handling and customer satisfaction (Stevens et al., 2018; Shams et al., 2020).

2.3. Theoretical Background and Conceptual Model

The prior index models were mostly based on the Quality, Satisfaction, and Performance (QSP) paradigm, Hirschman’s Exit, Voice, and loyalty theory 1970 (EVL), and Oliver’s (1980) expectations confirmation/disconfirmation theory (ECT). The study is enriched by two more diverse theoretical paradigms, the theory of reasoned action (TRA) (Fishbein & Ajzen, 1975) demonstrating the association among beliefs, attitudes, norms, intentions, and behavior and posits that humans make rational decisions depending on the information at hand. Technological Acceptance Model (TAM) (Davis, 1989) to anticipate and explain consumer behavior in regards to the acceptance and application of new technology. The theory explains the reasons for which the participants found the system useful and felt satisfied.

Customer satisfaction is mostly defined by a company’s service quality (Prasilowati et al., 2021). According to the ACSI, satisfaction is mostly quality pull rather than price and value pull. According to all prior indices and related studies, service quality has a significant positive relationship with customer satisfaction and corporate image (Gohain et al., 2018; Asyraf et al., 2019). For measuring consumer perception of service quality, the study employs a five-dimensional (Figure 1) service performance (SERVPERF) scale (Cronin & Taylor, 1992) based on the SERVQUAL scale. SERVQUAL is a disconfirmationbased scale, whereas SERVPERF is a performance-only scale. SERVPERF scale was used first time in constructing an index. In 2001 Norway and Taiwan 2009 used the SERVQUAL scale for the same purpose.

Figure 1: Conceptual Model

Organizations offer self-service technologies (SST) to improve the consumer experience, customer retention and decrease expenses. Hassan et al. (2020) defined SST as a technical interface through which consumers get services without being involved with firms’ employees and complete transactions more quickly and conveniently anytime, anywhere. Lin and Hsieh (2011) argued that consumers’ technology engagement and significant moves for information contribute to technology advancement and reduce the risk and uncertainty. Therefore, the focus has switched to determining how SST improves customer satisfaction and loyalty through successful business management and technology integration. Few studies were based on a hierarchical measurement model with multiple dimensions and manifest items to seize the entire domain of SST (Hassan et al., 2020). PCSI uses the SSTQUAL scale with seven dimensions (Figure 1) developed by (Lin & Hsieh, 2011) as an antecedent of customer satisfaction. For the scale validation, a variety of different validity and reliability tests in various situations were used (Fatmawati & Permatasari, 2019).

The Internet provides opportunities for building strong interactions and retaining relationships with customers. Organizations invest a lot of finance for creating efficient websites with traditional functionality and integrated marketing to attract customers and fulfill their needs (Ting et al., 2016). Sellers should focus on the issues and associated variables for competitive advantages and an acceptable share of the e-market. According to the literature, more than 40% of researchers examined fifty-four factors in various contexts and aspects, with five of them being the most frequently cited (customer services, purchasing process convenience, product information quality, website design, security perception) (Deyalage & Kulathunga, 2019). Three determinants of customer services, purchasing process convenience, and product information quality was selected for the study. The reason behind choosing three out of five indicators is to avoid redundancy, which would compromise the study’s validity and reliability. Because SSTQUAL discusses the two indicators website design and security perception. So, choosing the three indicators is logical.

Along with firms’ internal factors, some situational factors also influence customer satisfaction. The broader concept of satisfaction allows it to include other important factors besides just assessing product and service quality and values (Palacios et al., 2020). If they didn’t follow the expectations disconfirmation will cause. Situational factors are contextual factors that describe the environment for consumers related to specific issues of time and space of service consumption (Zarei et al., 2019). Crucial service situational factors are time factors and task interruption due to any reason. These factors cause consumers to complete their tasks rapidly and conveniently by adapting available options. Karma et al. (2015) and Prasadh and Suresh (2017) suggested to incorporates the situational factors in customer satisfaction studies as it is not addressed in the literature. During time pressure, the seller’s tactics perceived by the consumer as aggressive causes dissatisfaction and less trust in salespersons. Therefore, firms take a critical part in facilitating customers within the situation of time scarcity and other interruptions.

Consumer sentiments are referred to the specific emotions and values of consumers. According to the theory of emotions, people’s attitudes determine satisfaction and lead to emotions (Prasadh & Suresh, 2017). The emotional responses in different contexts influence customer satisfaction and make the company sensitive to customer values and emotions. Sentiments are significant antecedents of satisfaction examined by various researchers. (Ali et al., 2016; Karma et al., 2015). When a customer looks for a service provider, they show some trust and have a dialogue with them to minimize the perceived risk associated with the product and service. The substantial growth in online review pushes the firms to read and make holistic interpretations of opinions about themselves and competitors. These opinions and reviews are based on the consumer’s values and sentiments, an important source of understanding and managing the consumer’s sentiments and expectations. Kristensen and Eskildsen (2012) argued that the expectation dimension controversy of CSI can be removed by replacing it with consumer sentiments if the consumer sentiments are proved as a better measure of expectation.

Corporate image is reputation, customer perceptions, beliefs, impressions, and behavior towards an organization (Turkyılmaz & Ozkan, 2007). A satisfied customer has a favorable corporate image, impressions, positive behavior, and spread positive words of mouth. In turn, attract the new customers. Corporate image, due to its functional components directly associated with loyalty and its significant effect, is empirically proved by the researcher (Chien & Chi, 2019; Munoz et al., 2020). Hence, the study models satisfaction as input to corporate image and treated as a consequence rather than a driver.

The market’s expectations are an important determinant of overall satisfaction. For any product and service, customers have some positive or negative expectations regarding quality, price, and other values added to services. The prior experiences with the product and services, words of mouth, and consumer characteristics provide the basis for expectations. Satisfaction and service quality are directly affected by expectations. To answer the question, what derives expectations? It is very important for the firms. Influences of the individual’s expectations vary from time to time, and service quality also varies according to the customer expectations level (Munoz et al., 2020).

Satisfaction is referred to as a person’s contentment, and pleasure from his expectations compared with product, outcome, and perceived performance (Kotler et al., 2016). Customer interaction with the product and service results from positive or negative experiences irrespective of nature and form. The positive customer experience has the outcomes like satisfaction, revisit intentions, trust, loyalty, and repurchase intentions.

Loyalty is experienced, built, and described as a strong desire to continually repurchase the same product and service brand in the future (Oliver, 1999), regardless of marketing efforts, attractive alternatives, and situational influences. Satisfied and loyal customers are the source of improved financial performance and increased revenue. Besides the benefits, loyalty is the greatest challenge for businesses. A complex relationship exists between satisfaction and loyalty. Abundant studies proved the positive relationship between these two constructs and are still continuously under study (Dam & Dam, 2021; Alnsour et al., 2021). The aforementioned discussion provides the ground for constructing the following proposed research model. The hypothesized relationships among antecedents (consumer expectation, service quality, situational factors, SST, consumer sentiments, perceived price, corporate image, compliant handling) and consequences (customer loyalty and corporate image) are portrayed in Figure 1.

3. Methodology

3.1. Target Population and Sample Size

The target population consists of customers of the banking industry and mobile phone services. The reason for selecting these industries is the fierce competition and rising use of new and smart technologies for a fast and easy customer service interface. Consumers can access these services through smart technologies 24/7 at very reasonable charges. Pakistan has 169 million mobile phone subscribers Pakistan Telecommunication Authority (PTA, 2020) and 50.565 million bank accounts for a population of 207.77 million State Bank of Pakistan (SBP 2018). The target sample was chosen using a non-probability convenient sampling method. Raosoft’s (2004) sample size calculator was used for calculating the required sample size. Consequently, the sample of 385 for each industry was considered sufficient for the study.

3.2. Data Collection

The questionnaire was created using Google Forms and distributed on selected social media platforms. The data was collected using a 7-point Likert scale based on the measurement items of the model’s latent variables. The survey was conducted from April 2021 to December 2021. Items are altered and changed from previous research as per the requirement to improve the reliability and validity of construct measures. Of 385 responses, only 319 bank responses and 321 telecom responses (80%) were valid and selected for the final study. The response rate is comparable to prior studies.

3.3. Respondent’s Demographics

The respondent’s characteristics data reveal that 59% and 62% were married (telecom and bank), whereas unmarried constitutes 41% and 38%, respectively. 73% and 77% were male, while 27% and 23% were female in the telecom and bank industry. 21–40-year age was the largest age group of respondents representing the highly educated young people have masters and higher-level degrees. The largest portion of the respondents were job holders, comprised of 44% of telecom and 55% of bank. This group contains a mixed distribution of income levels, of which 26% and 18% prefer not to show their income, which shows purchasing power and sensitivity towards the price and security. Most of the customers belong to their respective telecom company or bank for 3–5 years.

4. Results

4.1. Pilot Study, Reliability and Validity

The smart PLS was employed to estimate the model. The PCSI model is a reflective measurement model. The SEM is defined by the Inner model (structural model) specifies the latent variable’s relationships and the Outer model (measurement model) specifies the association between the latent and their manifest variables (Hair et al., 2018).

The pilot survey was conducted to refine the question- naire. The loadings between the construct and their manifest variables meet the threshold value of 0.5. The reliability and validity of a CSI model’s results determine its applicability (Munoz et al., 2020). Therefore, the diagnostic measures e.g., Cronbach’s alpha and composite reliability (CR) are used to measure the internal consistency. Cronbach’s alpha and CR exceed the 0.7 threshold value, which confirms the unidimensionality of the manifest variables. Convergent validity is determined via Average Variance Extracted (AVE). AVE is also according to the threshold of 0.50 for all the variables, as shown in Table 1.

Table 1: Quality of Measurement Model

The measures’ weights optimize the items’ ability to explain the dependent variable. The PLS estimated weights of the items are used to construct the index score (ranges from 0–100-point scale). These scores serve as uniform and analogous measurement systems for organized benchmarks across firms over time.

4.2. Discriminant Validity

The discriminant validity can be checked through two main methods Fornell-Larcker criteria and cross-loadings. According to the Fornell-Larcker criteria, if the square root of AVE of the latent variable is more significant than its correlation with other latent variables, then there is discriminant validity. The study meets the criteria and establishes the discriminant validity, and explain that constructs share more variance with their indicators than other constructs, as shown in Table 2 and 3.

Table 2: Bank Discriminant Validity

Correlation is significant at 0.05.

Table 3: Telecom Discriminant Validity

The correlation is significant at 0.05.

All the constructs are conceptually and empirically different from each other for both industries. In 2001 Norwegian index evaluated banking and found the SERVQUAL violated the discriminant validity of 1% for the banks. The diagonal italic and bold figures represent the square root of AVE, and the downward figures represent correlations.

4.3. Path Coefficients

The path coefficients show how strong the links between the dependent and independent variables are. Table 4 reports the results, where most of the paths are significant, and insignificant paths are marked red. 32 (72.7%) out of 46 paths are significant in the predicted direction. SERVPERF has a ratio of 50% significant and 50% insignificant paths, from which tangible construct of telecom is at the boundary of p-value 0.051. Empathy, design, and tangibles constructs are insignificant. Banks should take the serious initiative to improve their image and empathy. Most important indicators are the assurance (β = 0.170, 0.33, t = 2.081, 2.087), responsiveness (β = 0.175, 0.24, t = 2.93, 3.03), for both industries. Knowledgeable, polite, helping staff, and timely service provision psychologically affect the customer perceptions. Cronin and Taylor (1992) proved SERVPERF empirically across four industries, bank, fast food, dry cleaning, and pest control, to corroborate the superiority of scale. These results are consistent with the ACSI, NCSB, Taiwan, and Danish index results.

Table 4: Path Coefficients

Table 5: PLS-MGA (Gender, Marital Status)

For both industries, promising results are received in terms of pure price construct and anticipated in correspondence with satisfaction and loyalty (β = 0.146, t = 1.790), (β = 0.321, t = 5.290). Rather than an indirect effect on loyalty via satisfaction, the price has a direct and considerable effect on satisfaction and loyalty. The results are consistent as the price is important for the customer. Similar results are provided by Johnson et al. (2001).

The Pakistan telecom industry does not have an effective complaint handling system for creating loyalty. The results show no direct effect of complaint handling on loyalty (β = 0.159, t = 0.633) but have a significant positive path coefficient towards satisfaction (β = 0.198, t = 1.890). Whereas banking complaint handling has a significant and positive path coefficient towards both satisfaction (β = 0.206, t = 2.642) and loyalty (β = 0.102, t = 2.311). Similarly, the Norwegian model reveals complaint handling has an insignificant path coefficient towards satisfaction and a significant path coefficient towards loyalty. Shams et al. (2020) demonstrated the significance of complaint handling in banking and telecom.

Customer satisfaction is dependent upon the situational factors with positive relationships (β = 0.135, t = 2.150) for banks and (β = –0.079, t = 2.097) negatively related to telecom. Karma et al. (2015) and Palacios et al. (2020) in their adaption study proved situational factors as significant in the context of banking.

Satisfaction has a direct effect on both the image and loyalty with significant path coefficients, whereas the significant path coefficient is also shown from image to loyalty. Customer loyalty is almost 100% for ACSI telecommunication. In ECSI the European customers are more satisfied with banking, and the lowest rating was received for the telecommunication sector.

The role of the sentiments in affecting customer satisfaction and loyalty shows trust and perceived risk as to the critical dimensions, but customer dialogues generate problems with overall results for banking. However, the path coefficient for the telecom has a significant positive value for satisfaction (β = 0.167, t = 1.561). Kristensen and Eskildsen (2012) proved no relationship with these factors, though individually, they are significant. Sweden index reveals a dominant influence identified in customer dialogue on customer loyalty for banking.

The expectation construct proved to be insignificant for both industries (β = 0.035, t = 0.123) and (β = 0.079, t = 1.199), same as ACSI and ECSI.

The study shows that online availability drivers have a direct effect on satisfaction (β = 0.118, t = 2.789), (β = 0.151, t = 3.011) for both industries. The SST dimensions individually and collectively explain the variances in satisfaction significantly. Out of a total of 14 SSTQUAL paths, 4 paths are insignificant towards satisfaction. Design (β = 0.060, t = 0.777), (β = 0.088, t = 1.353), enjoyment (β = 0.006, t = 0.122) of the banking industry, and functionality (β = 0.068, t = 1.346) of telecom has no direct effect on satisfaction. Hollebeek and Rather (2019) explored a significant association between technology and customer satisfaction. Table 4 also shows the variance explained in the dependent variables. The R2 63.5% (telecom) and 70.7% (bank) predictors explained the variance in satisfaction as compared to ACSI of 57%. Whereas R2 for loyalty is 64% and 66% for telecom and banking. The variance in the image is relatively low 40% and 50%, respectively. Furthermore, the banking industry shows more variation in satisfaction than loyalty as compared to telecom.

4.4. Index Scores

In the ACSI equation (Fornell et al., 1996), Wi is estimated unstandardized weights, xi is the measurement item of overall customer satisfaction. n is the number of measurement variables. In ACSI the three overall satisfaction indicators were used which range from 1 to 10. ACSI has applied in 7 sectors including banking and telecom.

\(\text { Score }=\frac{\sum_{i=1}^{h} w_{i} \overline{x_{i}}-\sum_{i=1}^{h} w_{i}}{9 \sum_{i=1}^{h} w_{i}} \times 100\)\(\text { Score }=\frac{\sum_{i=1}^{h} w_{i} \bar{x}_{i}-\sum_{i=1}^{h} w_{i}}{9 \sum_{i=1}^{h} w_{i}} \times 100\)

Based on this equation, the study proposes the following equation for calculating satisfaction. Where Wi and xi are unstandardized weights and manifest variables of customer satisfaction. For PCSI, n = 3, there are three indicators of satisfaction, and the observed range is from 1 to 7. The PCSI score for banking is 72%, and telecom is 68%, respectively. ACSI gives a satisfaction score of 74% for banking and 77% for telecom. The Norwegian banking sector has a 56.4% satisfaction score. Turkish telecom index score was 64.09%.

\(\text { Score }=\frac{\sum_{i=1}^{3} w_{i} x_{i}-\sum_{i=1}^{3} w_{i}}{6 \sum_{i=1}^{3} w_{i}} \times 100\)\(\text { Score }=\frac{\sum_{i=1}^{3} w_{i} x_{i}-\sum_{i=1}^{3} w_{i}}{6 \sum_{i=1}^{3} w_{i}} \times 100\)

4.5. Multi-Group Analysis (PLS-MGA)

A surfeit of recent studies has advised investigating the role of demographic variables. So, in the second part, we demeanor (PLS-GMA). The sample was divided based on gender and marital status. In the case of gender, a significant difference is noticed in customization (0.788), design (0.146), reliability (0.658), and responsiveness (0.520) to satisfaction link (telecom). On the other hand, the difference exists in complaint handling (0.464), customization (0.956), corporate image (0.063), perceived price (0.028), reliability (0.037), responsiveness (0.017), sentiments (0.040), and security (0.003) to satisfaction links, image and complaint handling to loyalty link (bank). Based on marital status, significant differences are found in customization (0.333), design (0.064), and reliability (0.940) to satisfaction link; and corporate image (0.884) to loyalty link for telecom. Significant differences are found in the case of bank convenience (0.150), customization (0.556), responsiveness (0.901), sentiments (0.030), and security (0.002) to satisfaction and corporate image (0.302) to loyalty. The study established that “the same size does not fit all” since the differences observed are based on gender and marital status.

5. Conclusion and Limitations

The CSI is used as a firm’s former, existing, and forthcoming performance indicator, to forecast the company’s profitability and market value. The five requirements must be fulfilled by the index. Customer satisfaction should be measured by differentiated indicators. Consequences of customer satisfaction should be identified. A causal relationship among the indicators should be identified through the structural model. The results should be comparable at different levels. This study proposes the PCSI model, which is an extended, modified, and advanced form of prior indices. This model integrated the offline and online customer satisfaction indices to provide a comprehensive scale for measuring cumulative customer satisfaction. The emergence of the COVID-19 pandemic induced isolation and lockdown, and business activities come to a halt. In such situations, industries’ online activities facilitate their customers effectively and efficiently to fulfill their needs and continue to grow. The model confirms theories of ECT, TRA, TAM, and EVL, by enriching with new determinants of intentions, attitude, service quality, and easing conditions. The study also shows that consumer perceptions of service quality, as well as other stimuli, have an impact on their tenacious feelings, attitudes, and behaviors. This helps the marketing managers examine the success factors and design planned interventions.

The results offer perceptions to advance the strategies in the service delivery process to address how to serve customers best to retain them. Furthermore, to address the question - What makes customers satisfied or dissatisfied? How is the company handling complaints, and whether are the procedures effective? How can customer satisfaction be improved? As compared to the competitors, where is the firm standing regarding customer satisfaction? Modernization of equipment makes the services more professional and attractive while making the customer more satisfied.

The limitation of the study includes that the model should be tested in different sectors and tested periodically. So, the results can compare with other national indices. The economic and financial latent variables should be incorporated into the model for broad conception. Consumer sentiment dimensions should be tested separately as trust and risk show significant results, but customer dialogues make the actor insignificant. The relationship between the situational factors, sentiments, online availability, and SST with loyalty should be tested for further insights, which the study lacks. PLS-SEM provided satisfactory results. We recommend supplementary development in perspective based on Bayesian approach extension and generalized structured component analysis via Choi and Hwang (2020).


  1. Ali, F., Amin, M., & Cobanoglu, C. (2016). An integrated model of service experience, emotions, satisfaction, and price acceptance: An empirical analysis in the Chinese hospitality industry. Journal of Hospitality Marketing and Management, 25(4), 449-475.
  2. Alnsour, I. A., Alnsour, I. R., & Alotoum, F. J. (2021). Customers' satisfaction using loyalty rewards programs: Evidence from Jordanian banks. Journal of Asian Finance, Economics, and Business, 8(11), 0297-0305.
  3. Asyraf, A., Zainudin, A., Norfadzilah, R., Hazimi, F., & Puspa, L. G. (2019). Assessing the effects of service quality on customer satisfaction. Management Science Letters, 9, 13-24.
  4. Babin, B. J., James, K. W., Camp, K., Jones, R. P., & Parker, J. M. (2019). Pursuing personal constructs through quality, value, and satisfaction. Journal of Retailing and Consumer Services, 51, 33-41.
  5. Bagozzi, R. P. (1994). Structural equation models in marketing research: Basic principels. In: Bagozzi, R. P. (Ed), Principles of marketing research, (pp. 317-385). Cambrige, MA. Blackwell.
  6. Bassey, F. O. (2014). The effect of perceived price fairness on customer satisfaction and loyalty [Master thesis, Eastern Mediterranean University, North Cyprus].
  7. Bruhn, M., & Grund, M. A. (2000). Theory, development, and implementation of national customer satisfaction indices: The Swiss Index of Customer Satisfaction (SWICS). Total Quality Management, 11(7), 1017-1028.
  8. Celil, C.A., Yilmaz, A., & Oya, Y. (2019). The impact of perceived price justice and satisfaction on loyalty: The mediating effect of revisit intention. Tourism Review, 74(3), 443-462.
  9. Chien, L., & Chi, S.Y. (2019). Corporate image as a mediator between service quality and customer satisfaction: Difference across categorized exhibitors. Heliyon, 5(3), e01307.
  10. Cho, N., & Park, S. (2001). Development of electronic commerce user-consumer satisfaction index (ECUSI) for Internet shopping. Industrial Management and Data Systems, 101(8), 400-406.
  11. Cronin, J. J., & Taylor, S.A. (1992), Measuring Service Quality: A Reexamination and Extension, Joumal of Marketing, 56 (3), 55-68.
  12. Dam, S. M., & Dam, T. C. (2021). Relationships between service quality, brand image, customer satisfaction, and customer loyalty. Journal of Asian Finance, Economics, and Business, 8(3), 0585-0593.
  13. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340.
  14. Deng, W. J., Yeh, M. L., & Sung, M. L. (2013). A customer satisfaction index model for international tourist hotels: Integrating consumption emotions into the American Customer Satisfaction Index. International Journal of Hospitality Management, 35, 133-140.
  15. Fatmawati, I., & Permatasari, F. B. (2019). The role of technology acceptance and E-service quality in predicting customer satisfaction and loyalty. Advances in Economics, Business and Management Research, 86, 54.
  16. Fishbein, M., & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.
  17. Fornell, C. (1992). A national customer satisfaction barometer, the Swedish experience. Journal of Marketing, 56(1), 6-21.
  18. Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E. (1996). The American customer satisfaction index: Nature purpose and findings. Journal of Marketing, 60(4), 7-18.
  19. Cakir, G., Bezbradica, M., & Helfert, M. (2019) The Shift from Financial to Non-financial Measures During Transition into Digital Retail - A Systematic Literature Review. In: Business Information Systems. Springer, 189-200.
  20. Gohain, K., Thambiah, S., & Hong, T. C. (2018). A comprehensive model with six service quality dimensions, satisfaction, and loyalty: An empirical investigation into the public hospitals in Malaysia. International Journal of Economics and Management Sciences, 07(1), 508.
  21. Grigoroudis, E., & Siskos, Y. (2004). A survey of customer satisfaction barometers: Some results from the transportation-communications sector. European Journal of Operational Research, 152(2), 334-353.
  22. Hair, J. F., Risher, J. J., Sarstedt, M., & Ringle, C. M. (2019). The results of PLS-SEM article information. European Business Review, 31(1), 2-24.
  23. Hassan, M., Iqbal, M. S., & Habibah, U. (2020). Self-service technology service quality: Building loyalty and intention through technology trust in the Pakistani service sector. SAGE Open, 10(2), 1-19.
  24. Hirschman, A. O. (1970). Exit, voice, and loyalty- responses to decline in firms, organizations, and states. Cambridge, MA: Harvard University Press.
  25. Hollebeek, L. D., & Rather, R. A. (2019). Service innovativeness and tourism customer outcomes. International Journal of Contemporary Hospitality Management, 31(11), 4227-4246.
  26. Hsu, S. H. (2008). Developing an index for online customer satisfaction: Adaptation of American customer satisfaction index. Expert Systems with Applications, 34(4), 3033-3042.
  27. Johnson, M. D., Gustafsson, A., Andreassen, T. W., Lervik, L., & Cha, J. (2001). The evolution and future of national customer satisfaction index models. Journal of Economic Psychology, 22(2), 217-245.
  28. Kalia, P., Kaushal, R., Singla, M., & Parkash, J. (2021). Determining the role of service quality, trust and commitment to customer loyalty for telecom service users: a PLS-SEM approach. The TQM Journal, 33(7), 377-396.
  29. Karma, N. G., Ibrahim, S. B., & Ali, A. H. (2015). The moderating effect of situational factors and the key factors influencing mobile banking adoption. International Journal of Management and Economics Invention, 10, 452-464.
  30. Keskar, M. Y., Pandey, N., & Patwardhan, A. A. (2020). Development of a conceptual framework for Internet banking customer satisfaction index. International Journal of Electronic Banking, 2(1), 417.
  31. Kim, H. R. (2005). Developing an index of online customer satisfaction. Journal of Financial Services Marketing, 10(1), 49-64.
  32. Kotler, P., Kartajaya, H., & Setiawan, I. (2016). Marketing 4.0: Moving from traditional to digital. Hoboken, NJ: Wiley & Sons.
  33. Kristensen, K., & Eskildsen, J. (2012). The relationship between SERVQUAL, national customer satisfaction indices, and consumer sentiment. Quality Management Journal, 19(2), 47-61.
  34. Lin, J.S. C., & Hsieh, P.L. (2011). Assessing the self-service technology encounters: Development and validation of SSTQUAL scale. Journal of Retailing, 87, 194-206.
  35. Meyer, A., & Dornach, F. (1996), The German Customer Barometer: Quality and Satisfaction - Yearbook of Customer Satisfaction in German 1995, German Marketing Association e.V. and German Post AG, Dusseldorf.
  36. Munoz, C., Laniado, H., & Cordoba, J. (2020). Development of a robust customer satisfaction index for domestic air journeys. Research in Transportation Business and Management, 37, 519.
  37. O'Loughlin, C., & Coenders, G. (2004). Estimation of European customer satisfaction index. Total Quality Management, 15(9), 1231-1255.
  38. Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460-469.
  39. Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(4), 33-44.
  40. Pakistan Economic Survey. (2020-2021). Financial report.›survey_2021.
  41. Palacios, L. L., Lopez, R. P., & Redondo, Y. P. (2020). How situation affects frontline employees' competencies on customer satisfaction with the store. Journal of Retailing and Consumer Services, 52, 101905.
  42. Park, Y., Heo, P., & Rim, M. (2008). Measurement of a customer satisfaction index for improvement of mobile RFID services in Korea. ETRI Journal, 30(5), 634-643.
  43. Prasadh, R. R., & Suresh, J. (2017). Customer satisfaction index model for the Indian banking industry: A qualitative study. Asian Social Science, 13(1), 114-129.
  44. Prasilowati, S. L., Suyanto, S., Safitri, J., & Wardani, M. K. (2021). The impact of service quality on customer satisfaction: The role of price. Journal of Asian Finance, Economics, and Business, 8(1), 451-455.
  45. Pakistan Telecommunication Authority. (2020). Annual report 2020. Retrieved from Pakistan telecommunication authority:
  46. Raosoft (2004) Sample size calculator, Raosoft inc software. Available at: (accessed 31 January 2022).
  47. Shams, G., Rehman, M. A., Samad, S., & Rather, R. A. (2020). The impact of the magnitude of service failure and complaint handling on satisfaction and brand credibility in the banking industry. Journal of Financial Services Marketing, 25(1-2), 25-34.
  48. Shin, D. H. (2014). Measuring the quality of smartphones: Development of a customer satisfaction index for smart services. International Journal of Mobile Communications, 12(4), 311-327.
  49. State Bank of Pakistan. (2018). 2018 (Quarter-1, FY19), July - September). Payment Systems Review. Retrieved from state bank of Pakistan:
  50. Stevens, J. L., Spaid, B. I., Breazeale, M., & Esmark Jones, C. L. E. (2018). Timeliness, transparency, and trust: A framework for managing online customer complaints. Business Horizons, 61(3), 375-384.
  51. Tabaei, Z., & Fathian, M. (2012). Developing online customer satisfaction strategic maps: With Iranian online retailing case studies. International Journal of Electronic Customer Relationship Management, 6(1), 87-112.
  52. Ting, O. S., Ariff, M. S. M., Zakuan, N., Sulaiman, Z., & Saman, M. Z. M. (2016). E-service quality, e-satisfaction and e-loyalty of online shoppers in business to the consumer market: Evidence from Malaysia. IOP Conference Series: Materials Science and Engineering, 131, 1-11.
  53. Turkyilmaz, A., & Ozkan, C. (2007). Development of a customer satisfaction index model: An application to the Turkish mobile phone sector. Industrial Management and Data Systems, 107(5), 672-687.
  54. Virvilaite, R., & Daubaraite, U. (2011). Corporate social responsibility in forming a corporate image. Engineering Economics, 22(5), 534-543.
  55. Zarei, M. M., & Angel, F. (2019). Choosing the delivery and return method in purchases: the effect of situational factors in omnichannel contexts. Economic Research, 20,