Consumers’ Perception Analysis on Service Quality of Coffee Shops and Coworking Spaces in Yogyakarta, Indonesia

Every marketing organization, whether manufacturing or service provider company is required to create differentiation, positioning


INTRODUCTION
Culinary businesses in Yogyakarta range from small-scale stalls or Angkringan to mediumscale restaurants and cafes, as well as large-scale restaurants.Branding is essential for both large and new companies and the sustainability of businesses.Moreover, the strategy of building a brand is crucial for developing the cognitive aspects of SME actors.A strong brand-building process can transform a small business into a successful brand (Setiawati et al., 2019).Coffee shops are locally owned public meeting places that serve coffee to consumers, either as main or part of menus.Shops also provide spaces for completing assignments, reading books, and accessing the internet (Tumanan and Lansangan, 2012).Coworking spaces can be classified into six categories based on business model and user access level, namely public offices (free coworking spaces, such as libraries), third places (public spaces offering purchase of services, such as cafes), collaboration centers (public spaces fostering collaboration among workers), hotels or rentals (public spaces offering short-term rental contracts and service packages), incubators (public spaces focusing on entrepreneurship), and shared studios (shared office spaces with flexible rental contract terms catering to the community) (Weijs-Perrée et al., 2019).Coffee shops currently offer drinks and food as the main menu (Leonita and Tulistyantoro, 2017).The increase in demand has led to the emergence of various brands, cafes, and coffee shops in major cities.Despite the variety of options, the targeted market share varies.Therefore, the culture of coffee consumption can be practiced by people in both small and big cities (Kurniawan and Rosyid, 2017).Coworking spaces provide interiors that support informal social interactions more effectively than traditional office concepts (Bouncken et al., 2020).The physical configuration encourages the development of shared interaction, which is fundamental to building social relationships (Bhansing et al, 2020).
Coworking spaces are among the newer business concepts in Indonesia, with Bandung being the main pioneer in 2010.This business model offers a comfortable and economical workspace, supported by the rapid growth of new companies.The emergence of coworking spaces has prompted the creation of strong differentiation strategies (Damayanti, 2017).According to Creative Economy Agency (BEKRAF), the development of start-ups has been rapid, with 992 reported in 2018 and 2,193 in 2019.Therefore, Indonesia ranks fifth globally in terms of the number of startups, trailing the United States, India, the United Kingdom, and Canada (Irawan et al, 2020).In recent times, coffee shops offering coworking spaces are expanding, leading to the focus on strong service quality and consumers' satisfaction.Coffee shops providing coworking spaces aim to attract not just onetime visitors but to foster continuous and increasing patronage.Students and workers who do not require a traditional office setting often find comfort in working at shops with coworking spaces.The development of coworking spaces has increased significantly.According to the Indonesian Coworking Space Association, the number of coworking spaces increased from around 40 in 2013 to 200 in 2018, marking a 400% increase within five years (Surbakti and Danil, 2020).
Previous studies showed that community participation in coworking spaces is not optimal despite the continuous growth in coworking spaces.As a result, some people tend to perceive café or coffee shops as coworking spaces (Damayanti, 2017).A preliminary survey was conducted to determine whether there was a significant difference between coffee and coworking spaces in order to strengthen the study objective.The results were used to assess consumer's understanding of the differences.There were significant variations between the two, including marketing mix variables, atmosphere, and visiting motivations.Consumers could distinguish between coffee and coworking spaces, facilitating the adaptation to activities in coworking spaces.This concept was appealing and fostered collaboration among industry peers.When the concept side is standardized effectively across companies and consumers, perceptions of service provided are less biased.The facilities offered include private rooms, meeting rooms, outdoor spaces (gardens), indoor spaces, prayer rooms, toilets, printers, charging cables, and Wi-Fi.Various events are hosted at Coffee & Coworking Space, such as webinars for freelancers, Microsoft, and content creators.The private rooms are available for rent by individuals needing dedicated workspaces.All rented rooms are equipped with whiteboards, screens, air conditioning, and other facilities.Service quality is closely related to consumer's satisfaction, with good quality instilling a sense of trust in companies.Over the long term, trust enables companies to better understand consumer's desires and needs.Consequently, companies can increase consumer's satisfaction by maximizing pleasant and minimizing unpleasant experiences (Rangkuti, 2006).This study was aimed to determine the priority of service attributes requiring improvement and provide suggestions based on observations made at coffee and coworking spaces.

Study Object
This study was conducted at three well known Coffee & Coworking Spaces in Yogyakarta, Indonesia namely K, R, and S. The primary focus was on service quality, and identification was conducted directly to understand service system, including interviews with the owners to ascertain marketing strategies and mix.A non-probability sampling was used and included the distribution of closed questionnaires with questions related to service quality.
The questionnaires were distributed across the three Coffee & Coworking Spaces, with 100 respondents at each location.The variables assessed were in the dimensions of service quality (tangible evidence, reliability, responsiveness, assurance, and insurance), validated through CVR analysis.Each statement offered an alternative answer on a scale of 1-6.The Likert scale used comprised six options, namely SD (Strongly Disagree), D (Disagree), TD (Tend to Disagree), TA (Tend to Agree), A (Agree), and SA (Strongly Agree) valued at the range of 1-6.Questionnaires were distributed both directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents comprised 61 consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).

Dimensions of Service Quality
Five dimensions of Service Quality were used to measure service quality according to previous study by Tjiptono and Chandra (2016) as shown in Table 1.

Importance Performance Analysis (IPA)
IPA was conducted to identify the importance level of each service attribute and determine the priority of improvements.
The attributes were classified into the following quadrants based on the Cartesian diagram cited from Crowder et al. (2019): 1) Quadrant 1: The attribute has a relatively low level of performance and is considered essential to consumers, necessitating immediate attention and improvements; 2) Quadrant 2: The attribute has high performance and is considered very important to consumers, thereby representing major strengths and providing an opportunity to maintain a competitive advantage; 3) Quadrant 3: The attribute has  The attributes were classified into the following quad from Crowder et al. (2019): 1) Quadrant 1: The attribute h is considered essential to consumers, necessitating im Quadrant 2: The attribute has high performance and is thereby representing major strengths and providing a advantage; 3) Quadrant 3: The attribute has low level of p referred to as a minor weakness and does not require sig attribute has high performance level, necessitating compa quadrant requiring an increase in performance.low level of performance and importance.Therefore, it is referred to as a minor weakness and does not require significant improvement; 4) Quadrant 4: The attribute has high performance level, necessitating companies to reallocate inherent resources to the quadrant requiring an increase in performance.

Potential Gain in Customer Value (PGCV)
PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016), the steps for calculating the index value are presented in Equations 1, 2, and 3.

Potential Gain in Customer Value (PGCV)
PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016), the steps for calculating the index value are presented in Equations 1, 2, and 3.

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016), the steps for calculating the index value are presented in Equations 1, 2, and 3.

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents analysis, complementing the results of IPA.This is instrumental in determining priority decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahar steps for calculating the index value are presented in Equations 1, 2, and 3. Achieve Customer Value (ACV)iv  =  ̅ ×  ̅

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidenc responsiveness, assurance, and empathy.Each of these variables was represented statements based on the ServQual (Service Quality) framework for measurement ( Chandra, 2016).The variables used in the questionnaires were in the dimensions of s validated with CVR.The questionnaires were distributed through a combination of metho consumers who visited spaces (offline) and via Google Forms (online).The respo consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characte respondents obtained from preliminary questions included gender, age, occupation, dom frequency of visits, visits duration, accompanying partners, preferred coworking spaces, information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016), the steps for calculating the index value are presented in Equations 1, 2, and 3.

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016)

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an ad analysis, complementing the results of IPA.This is instrumental in determining priority impro decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (20 steps for calculating the index value are presented in Equations 1, 2, and 3. Achieve Customer Value (ACV)iv  =  ̅ ×  ̅

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, re responsiveness, assurance, and empathy.Each of these variables was represented by statements based on the ServQual (Service Quality) framework for measurement (Tjipto Chandra, 2016).The variables used in the questionnaires were in the dimensions of service validated with CVR.The questionnaires were distributed through a combination of methods, di consumers who visited spaces (offline) and via Google Forms (online).The responden consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics respondents obtained from preliminary questions included gender, age, occupation, domicile, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and information.PGCV index is a tool commonly used in marketing analysis methods.PGVC represents an advanced analysis, complementing the results of IPA.This is instrumental in determining priority improvement decisions of companies (Mastur and Matahari, 2016).According to Mastur and Matahari (2016), the steps for calculating the index value are presented in Equations 1, 2, and 3. Achieve Customer Value (ACV)iv  =  ̅ ×  ̅

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of physical evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.(3)

RESULTS AND DISCUSSION
Analysis of service quality comprised the dimensions of tangible evidence, reliability, responsiveness, assurance, and empathy.Each of these variables was represented by several statements based on the ServQual (Service Quality) framework for measurement (Tjiptono and Chandra, 2016).The variables used in the questionnaires were in the dimensions of service quality validated with CVR.The questionnaires were distributed through a combination of methods, directly to consumers who visited spaces (offline) and via Google Forms (online).The respondents were consumers of Coffee & Coworking Spaces residing in Yogyakarta (DIY).The characteristics of the respondents obtained from preliminary questions included gender, age, occupation, domicile, income, frequency of visits, visits duration, accompanying partners, preferred coworking spaces, and location information.
Based on the characteristics of the respondents, females tended to share information through Word of Mouth (WOM) regarding experiences felt by relatives or acquaintances.Consumers aged 20-30 years, classified as being in productive age, had high levels of mobility, required spaces to work, and tended to socialize frequently with colleagues.The majority of visitors were students, reflecting changes in lifestyle behaviors.Working outdoors offered new and comfortable atmosphere, stimulating enthusiasm, and generating many ideas.Furthermore, the strategic location in residential areas and near campuses attracted many students who prioritized working and socializing with colleagues.Respondents residing in Sleman were

IPA
Quality of services by coffee coworking spaces is determined based on conformity level, which includes comparing performance with consumer's expectations.Therefore, to investigate whether consumers accept or reject a product/brand, it is crucial to closely monitor the perspectives of products and services (Busono et al., 2017).Importance Performance Analysis (IPA) is an analytical method used to identify important performance factors that an organization should prioritize to meet consumer's satisfaction (Umam and Hariastuti, 2018).Degree of conformity value below 100% showed that the performance delivered did not meet consumers' expectations, and a value above 100% showed expectations were met, leading to increased satisfaction.
In Table 3, the level of conformity showed varying values, with each dimension having the highest degree of conformity.In the tangible dimension, the highest degree of conformity was obtained for the cleanliness of facilities such as toilets, kitchens, and meeting rooms at 95.54%.This showed cleanliness did not meet consumers' expectations and required further attention.In the reliability dimension, the highest degree of conformity was obtained in the accuracy of calculations during payment transactions, recording 101.33%.This showed accuracy met consumers' expectations and should be maintained.In the responsiveness dimension, the highest degree of conformity was recorded for waiting time when confirming a place booking at 97.57%.This showed waiting time did not correspond with consumers' expectations and needed to be addressed.The assurance dimension had the highest degree of conformity for employees' ability to use required work facilities at 97.48%.This showed employees' abilities did not meet consumers' expectations.In empathy dimension, the highest degree of conformity was obtained for the convenience of opening hours at 98.43%.This showed operating time of 09.00 a.m.-11.00 p.m. did not meet consumers' expectations.Generally, the performance of Coffee & Coworking Spaces did not fully meet consumers' expectations, evidenced by an average degree of conformity value of 94.76%.Attribute reliability specifically had a degree of conformity value of 101.33%, exceeding 100%, and attribute empathy had the lowest with 80.27%.
IPA is an analytical method for identifying important performance factors companies prioritize to meet consumers' satisfaction.Each attribute is represented in a Cartesian diagram, allowing for the determination of both the priority performance level of all attributes (96.864) and the average importance levels of all attributes (102.253).The center limit for the performance category or x-axis was 4.843, while the center limit for the importance category or y-axis was 5.113.
The results obtained from measuring the level of importance and performance using IPA could be analyzed for each attribute within the quadrant on the Cartesian diagram.Attributes in quadrant I are the main weaknesses that require immediate attention for improvement.This showed com panies' performance was unsatisfactory, necessitating the focus on the improvement of performance.These attributes were classified into quadrant I, there is one statement each on the tangible dimension, responsiveness, and empathy dimensions statement.Attributes in quadrant II included the main opportunities and strengths to maintain and achieve competitive advantage.Eight attributes were categorized into this quadrant.One statement in the tangible dimension and empathy, two statements each regarding reliability, responsiveness, and assurance.Attributes in quadrant III were not the priority for improvement.Six attributes were categorized into this quadrant, one statement in empathy, two statement in tangible dimension, and The results obtained from measuring the level of importance and performance using IPA could be analyzed for each attribute within the quadrant on the Cartesian diagram.Attributes in quadrant I are the main weaknesses that require immediate attention for improvement.This showed com panies' performance was unsatisfactory, necessitating the focus on the improvement of performance.These attributes were classified into quadrant I, there is one statement each on the tangible dimension, responsiveness, and empathy dimensions statement.Attributes in quadrant II included the main opportunities and strengths to maintain and achieve competitive advantage.Eight attributes were categorized into this quadrant.One statement in the tangible dimension and empathy, two statements each regarding reliability, responsiveness, and assurance.Attributes in quadrant III were not the priority for improvement.Six attributes were categorized into this quadrant, one statement in empathy, two statement in tangible dimension, and three statement in assurance.Attributes in quadrant IV should be reallocated to other attributes requiring increased attention.Three attributes were categorized into this quadrant, two statements in reliability dimension, and one statement in responsiveness.

PGCV
The attributes in quadrant I of IPA cartesian diagram should not be considered top priority when recommending improvements.Therefore, re-analysis should be conducted using PGCV to determine priority improvements based on PGCV index value.A higher index value showed the attributes were crucial and should be prioritized for improvement.The index value was determined by ACV and UDCV values.

PGCV
The attributes in quadrant I of IPA cartesian diagram should not be considered top priority when recommending improvements.Therefore, re-analysis should be conducted using PGCV to determine priority improvements based on PGCV index value.A higher index value showed the attributes were crucial and should be prioritized for improvement.The index value was determined by ACV and UDCV values.
Based on Table 3, ACV was obtained by multiplying expectation value with the highest perception value.The highest ACV was obtained in attribute responsiveness at 24.556, followed by tangible at 22.206, while empathy had the lowest at 21.608.UDCV value, obtained by multiplying expectation value, had the highest level of performance in attribute tangible with 31.900,followed by attribute empathy at 31.218, with responsiveness recording the lowest at 30.70.PGCV index was used to determine which attributes had the highest potential to provide added value, serving as reference to identify the top priority for improvements (Mastur and Matahari, 2016).The first priority for improving companies' performance was attribute tangible supporting facilities such as tables, chairs, wifi, power outlets, and air conditioning, with a PGCV index value of 9.694.Given service function, the facility was prioritized as it was the first attribute consumers noticed when visiting Coffee & Coworking Spaces.The second priority was attribute empathy, namely employees' willingness to help without being asked, with a PGCV index value of 9.610.Lastly, the third priority was attribute responsiveness, namely the provision of recommendations to consumers confused when ordering, with a PGCV index value of 6.224.

Recommendation for Improvement
Achieving consumers' satisfaction can be challenging when companies fail to understand what is expected from product or service (Lubis and Andayani, 2017).In a related study, consumers' evaluation of Keda Kopi Mandheling in Medan showed satisfaction.The attributes requiring improvement included brand image, decoration, and barista presentation.Developing Mandheling Coffee within the trade sector presents a promising opportunity for every coffee farmer, trader, and entrepreneur to leverage both the national and international markets (Leonita and Tulistyantoro, 2017).Recommendations for improvements should focus on attributes influencing consumers' satisfaction.Attributes in quadrant I of IPA with highest PGCV index value were prioritized for improvement recommendations.The suggestions were based on field observations at K, R and S coffee & coworking spaces.The issue with attribute tangible originated from inadequate attention to the provided facilities (old and deteriorating), and a lack of power outlets.Therefore, regular checks on all facilities were necessary, with unsuitable facilities replaced, and power outlets (electrical terminals) installed accordingly.Deficiency in attribute empathy occurred due to employees' lack of attention to consumers, necessitating training to improve service skills.Human resource development should educate employees on proper and consistent consumers' interaction methods (Thomas, 2022).Examples of employees training included offering excellent service that exceeds consumers' expectations, prioritizing time, cost, accuracy, quantity, quality, safety, comfort, process, and satisfaction, commonly referred to as Service Excellence.The issue with attribute responsiveness was that employees were not fully acquainted with the entire menu and lacked effective consumers' interaction skills.Therefore, employees should be able to interact with consumers by offering all available menus to prevent queues at the cashier's desk and reduce waiting times.

CONCLUSION
In conclusion, PGCV index showed the order of priority for improvement was the tangible evidence variable, empathy, and responsiveness.Recommendations for the improvements of attributes are included regular checking and replacement of damaged facilities, required training to facilitate employee's abilities, while necessitated training of employees to interact with consumers effectively by offering available menus.

Figure 1 .
Figure 1.Cartesian diagram for determination of improvem

Figure 1 .
Figure 1.Cartesian diagram for determination of improvement priority :  ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is 6.Potential Gain in Customer Value (PGCV) ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is 6.Potential Gain in Customer Value (PGCV)  =  −  (3) :  ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction le Potential Gain in Customer Value (PGCV)  =  −  (3) ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is 6.Potential Gain in Customer Value (PGCV) , the steps for calculating the index value are presented in Equations 1, 2, and 3. Achieve Customer Value (ACV)iv  =  ̅ ×  ̅ (1) Where:  ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is 6.Potential Gain in Customer Value (PGCV)  =  −  (3) ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is Potential Gain in Customer Value (PGCV) ̅ = Average level of importance and  ̅ = Average performance level.Ultimately Desire Customer Value (UDCV)  =  ̅ ×   (2) Where:  ̅ = Average level of importance and   = The highest value of satisfaction level is 6.Potential Gain in Customer Value (PGCV)  =  −  (3)

Figure 2 .
Figure 2. Cartesian diagram results measuring the level of importance and performance using IPA

Figure 2 .
Figure 2. Cartesian diagram results measuring the level of importance and performance using IPA

Table 1 .
ServQual dimension for service quality measurement