Assessing the significance of retail service quality on shopping frequency: an adaptation of retail service quality (RSQS) model

Purpose – This study aims to ascertain the significance of retail service quality dimensions on shopping frequency at supermarkets in Durban. This study also adopts the Retail Service Quality Scale (RSQS) to South African supermarket consumers. Design/methodology/approach – Primarydatawerecollectedfrom399consumersthroughmallintercepts using an adapted RSQS. Non-probability convenience sampling was utilised in selecting participants from different malls in Durban. To analyse data the SPSS software was utilised with multiple regression analysis to confirm relationships between variables of the study. Findings – Results indicate a significant linear relationship between retail service quality and shopping frequency. Two of the three dimensions of retail service quality (atmospherics and reliability) have a positive and significant influence on the shopping frequency of consumers whereas one dimension (policy) showed no significant influence on the dependent variable. Research limitations/implications – Further studiesare recommendedin ascertainingthe reasons behind aninsignificantrelationshipbetweenpolicyitemsofservicequalityandtheshoppingfrequencyofconsumers. Practical implications – This study highlights the managerial implications of retail service quality on improved shopping frequency of consumers. Originality/value – This study suggests a lesser emphasis on policy items specifically personal interaction amongst Durban consumers on their shopping frequency. This might be due to cultural differences as well as the importance of self-service and privacy in supermarkets. Furthermore, this study demonstrates the role of context in providing deviations in retail service quality measurement and conceptualisation.


Introduction
Retail customer service has advanced significantly over time.Physical stores are still threatened by South Africa's growing popularity of online shopping (Bu et al., 2020).Online shopping is said to have several advantages over traditional brick-and-mortar stores, including reduced costs, convenience, cost savings and personalised service (Moeti et al., 2021).To remain competitive for brick-and-mortar stores, recommendations still centre on enhancing the buying experience in actual stores (Angula and Zulu, 2021;Kotler and Armstrong, 2018).The goal of retail service has not changed with time.It is still about keeping customers to maximise revenue as it is more costly to attract new customers than to keep existing ones (Prabhu and Aithal, 2022).
It is important to note that consumers today are more hypersensitive than they were in the past when businesses tended to impose their service offerings on consumers.Nowadays, customers expect courteous, timely service that meets their standards anytime they visit a store.Customers have more leverage to demand value for their money because they have more purchasing power and a wide range of shopping options (Bu et al., 2020;Teoh and Gaur, 2018).Consumer behaviour today has changed from being passive to being vocal and outspoken about how they feel about service delivery (Ran and Zhou, 2019).Generally, South African customers are more demanding when making purchases due to increasing awareness of their rights as consumers and the desire for value for money.
There is a need to investigate service quality elements that affect consumers' frequency of shopping and moreover, a need to evaluate the importance of retail service quality on shopping frequency using an adjusted Retail Service Quality Scale (RSQS).Traditional methods of measuring service quality are insufficient to evaluate customers' unique service experiences with differing service environments (Palese and Usai, 2018).There is further strong evidence that retailing aspects of gauging service are still limited in South Africa.On the other hand, studies utilising models of retail service quality in developing contexts have yielded varied results that confirm at least two or more dimensions of retail service quality with reliability, problem-solving and policy considered insignificant (Adam et al., 2022;Leen and Ramayah, 2011;Das et al., 2010;Zia, 2020).Such differences represent a retail context gap as the majority of studies in these developing nations focussed on specialised retail settings such as apparel, departmental and discount stores over outlets offering essentials/staples.This calls for a generalised and compressed approach to measuring retail service quality.
Studies advise using a customer-centred perspective when conceptualising service quality because different values, contexts and metrics can accurately translate into behaviour (Prakash, 2018;Abror et al., 2020) and traditional service quality models are unable to accurately analyse and conceptualise unique service perceptions of consumers in retail environments due to their well-known distinctive behaviour.

Retailing in South Africa
Retailers are divided into two main groups in South Africa, those who sell food and those who sell general merchandise.In South Africa, full-service supermarkets, hypermarkets, warehouse clubs and convenience stores are the most common types of food merchants.The top five food retail groups, which together hold more than 72% of the market share for food retailing, are Shoprite Holdings, Pick and Pay Stores, the SPAR Group, Woolworths Holdings and Fruit and Veg City (Food Lovers Market) (Das Nair, 2018;Jere et al., 2021).South Africa's total number of supermarkets has been gradually increasing since the late 1990s.In the previous 20 years, supermarkets have advanced dramatically in the retail industry (Nickanor et al., 2021).Such development is seen in the retail industry, where major retailers contributed more than half of the country's sales of food, toiletries and confectioneries.This is because of the expansion of the economy, the rising of the middle class and changes in consumer tastes.In addition, supermarkets have expanded and changed in size as well as in design and location (Dhurup et al., 2005;Makhitha and Khumalo, 2019).In the South African context, supermarkets act as a conduit between consumers and the food system, influencing consumer food preferences and purchase behaviour through sophisticated marketing techniques.Today's supermarket offerings go well beyond a simple "basket of products" for a set price.Supermarkets give customers access to a variety of goods in a single location, are easily accessible, have flexible trading times and provide added services such as financial, cellular products and other services.Ultimately, the goal is to deliver an "overall consumer experience."(Das Nair, 2018;Mahlangu and Makhitha, 2019;Mielmann, 2019).

Literature review and hypotheses development
The concepts and the connections between the variables under research that serve as the foundation for debates and the formulation of testable hypotheses are reviewed in this segment of the literature.According to the study's null hypothesis, there is no meaningful correlation between retail service quality (an independent variable) and shopping frequency (a dependent variable).
Retail service quality conceptualisation Dabholkar et al. (1996) devised the RSQS, which was adapted from the SERVQUAL scale to enable the measuring of service quality in the retail setting, as significant efforts to study service quality originated from the services sector rather than the retail sector.The scale includes five factors: physical aspects, which refer to tangible and visible in-store aspects such as store layout, comfort, privacy, architecture, colour, materials and store design; interpersonal interactions, which refer to employees' willingness to provide service and inspire confidence in clients; problem-solving, which involves handling complaints and assisting clients in returning goods; and reliability, which is keeping to a service agreement (such as high-quality merchandise, convenient location and operating hours, adequate parking space and payment options).Boshoff and Terblanche (1997) made only a little contribution to the RSQS's validation in South Africa.Instead of supermarkets, the model was used in a hypermarket.According to Hassan et al. (2013), there is a size difference between the two contexts, with supermarkets predicted to yield fewer consumer insights due to their limited selection and selection of sports, houseware and home improvement products.Additionally, hypermarket customers typically need the aid of store employees, notably in the appliances and home improvement sections, whereas supermarket customers exclusively rely on self-service (Levy et al., 2019).Despite efforts to replicate the use of retail service quality in different contexts; results indicate partial success as service quality measurement is dependent on retail format and environment (Sirohi and Kumar, 2018).
Consumer service quality perceptions are anticipated to differ given the contextual deficiencies.It is necessary to alter the model because it is possible that some RSQS dimensions would be distorted in a supermarket scenario.This conceptualisation is consistent with the compressed and modified RSQS for South African supermarkets proposed by Dhurup et al. (2005), which included atmospherics, policy and reliable service quality elements.It also advances knowledge by recognising the cultural and economic variety of South Africa.

Atmospherics
The atmospherics of a retail outlet refers to the pleasant ambience inside the store that warrants a customer's convenience and enjoyment when shopping (Roschk and Hosseinpour, 2020).Similarly, these are tangible store aspects that facilitate convenience for patrons such as Retail service quality on shopping frequency equipment and fixtures, physical amenities in and outside the outlet as well as service materials (Zia, 2020).Besides the physical appearance of such amenities; atmospherics aim at providing shopping convenience to the customer with the layout of the store.Empirical evidence suggests positive perceptions of service quality result in even greater positive evaluations of retail service quality (Venkateswaran, 2021).Therefore, the following alternate hypothesis is developed to suggest an effect of atmospherics on increased shopping frequency.
H1.There is a significant relationship between the atmospherics of a store and improved shopping frequency.

Policy
Generally, policies are issues of service quality related to a retailer's policy such as quality of merchandise, trading hours, payment facilities and parking convenience (Zia, 2020).Briggs et al. (2020) substantiate the need for a holistic retail policy as it facilitates quick and convenient purchases or after-sale service.
In a South African context, when measuring consumer perception of retail service; distinguishing aspects of policy from personal interaction (i.e. with store personnel) and problem-solving issues proved challenging and prompted the possibility of having mixed findings.The RSQS concept was first used in a hypermarket setting rather than a supermarket setting.Despite being smaller than hypermarkets, supermarkets offer a broad and varied selection of goods.Limited quantities of sports, household goods, and home renovation products are sold in supermarkets.Contrarily, hypermarkets are bigger, conduct more transactions each year and are linked with infrequent purchases (Hassan et al., 2013).Additionally, hypermarket customers typically need the aid of store employees, notably in the appliances and home improvement sections, whereas supermarket customers exclusively rely on self-service (Levy et al., 2019).As a result, the three characteristics of the policy, interpersonal contact and problem-solving were combined into one category, called policy, in this study.
Personal interaction is how personnel treat customers when interacting with them and naturally customers opt for a friendly approach (Zhang et al., 2019).Moreover, virtual interactions in omnichannel retailing are equally important in augmenting positive consumer evaluations of service (Rhee and Lee, 2021).Furthermore, progressive interactions solve customer problems.Problem-solving ensures product returns are managed and complaints are addressed (Zia, 2020).An effective strategy that improves retail service quality encompasses problem-solving (Reynaers, 2020).In essence, a comprehensive retail service quality policy embraces operational issues (trading times, payment facilities and merchandise quality), positive personal interaction and problem-solving.An effective retail service policy is hence expected to significantly improve customers' shopping frequency.
H2.There is a significant relationship between the policy of a store and improved shopping frequency.

Reliability
All store aspects relating to keeping task-related promises, meeting service expectations, error-free sales and being customer-friendly indicate how reliable an outlet is (Zia, 2020).
A reliable store has good rapport amongst customers.The reliability of retail service is the fulfilment of promises made, it is the result of any successful transaction and is key to the overall assessment of service quality (Zhang et al., 2019).Alternatively, non-fulfilment of promises made by the retailer results in negative responses from the customer.Hence, H3.There is a significant relationship between the reliability of a store and improved shopping frequency.

Research question
This study focusses on the importance of retail service quality in improved shopping frequency using an adapted RSQS model in Durban supermarkets.Consumer perceptions of retail service quality were measured across three adapted dimensions of the RSQS model namely atmospherics, reliability, and policy.To ascertain shopping frequency; the degree of shopping visits per identified outlet amongst customers was used.

Research methodology
This quantitative study employed a descriptive research methodology to assess the significance of retail service quality dimensions on shopping frequency in Durban supermarkets.Descriptive research, by Leedy and Ormrod (2019), entails either identifying the characteristics of an observable fact or delving into potential connections between two or more occurrences, which is in line with the goals of this research.In a similar vein, Schwandt (2014) interprets descriptive research as characterising, clarifying and interpreting the current state of the subject area.Descriptive research, in general, made it easier to assess a trend or occurrence broadly in the Greater eThekwini Metropolitan area, making it a suitable approach to achieving the objectives of this research by responding to the "what" questions (Churchill and Iacobucci, 2006).
In this study, all shoppers at supermarkets in the Greater eThekwini Metropolitan region were the target population.Demographic variables such as age, gender and income were taken into consideration because the researchers believed they might affect the findings.According to eThekwini Municipality (2013), a total of 3,442,361 people, comprising 1,679,040 males and 1,763,321 women, reside in the eThekwini Metropolitan area.Given the population's size of over a million, Sekaran and Bougie (2019) assert that a sample size of 370 participants should be sufficient to validate the research findings.As a result, non-probability and convenience sampling were utilised to select the study's participants.A sample size of 400 participants was chosen for this study since it was thought to be sufficient to gather data for analysis and compensate for non-responses considering the nature and characteristics of the sample.
Before data collection, the questionnaire underwent a pre-test to determine its suitability for collecting the required data.This process allowed the researchers to select the instructions to include in the survey as well as the structure and content of the questionnaire.In total, 400 questionnaires were self-administered to respondents at malls and supermarkets.A total of ten supermarkets were visited to administer a maximum of 40 questionnaires per supermarket.Using mall intercepts, shoppers were approached after shopping at supermarket outlets in the Greater eThekwini Metropolitan area.Primary data were collected using an adapted RSQS questionnaire.The RSQS model has been empirically confirmed across various contexts hence the adoption in this study.Using a Likert scale ranging from (1strongly agree to 5strongly disagree) the adapted RSQS questionnaire consisted of 29 questions measuring consumer perceptions of service in a store.
Following data collection, an analysis of the collected data was done to obtain pertinent knowledge that addressed the problem identified in this study.The SPSS software version 22 was used to compute validity, reliability statistics (factor analysis, Cronbach alpha statistic) and analyse data.Data analysis techniques using descriptive statistics, including graphs, charts, descriptive statistics and appropriate inferential statistics, were used.The Cronbach alpha statistic was computed to address internal consistency for reliability.Lower than 0.70 coefficients were disregarded and used throughout the instrument.

Analysis and results
To ensure this study's instrument measured what was intended and if the results are dependable to be generalised to appropriate contexts; reliability and validity tests Retail service quality on shopping frequency were conducted.A Cronbach alpha score exceeding 0.7 was deemed reasonable to confirm reliability.Table 1 depicts an overall 0.877 Cronbach alpha value which is considered good for confirmatory purposes.This indicates the high dependability and stability of the scale used in this study.Hence, the dimensions adopted in this study are of relevance to retail service quality.
It was necessary to conduct Bartlett's test of sphericity and Kaiser-Meyer-Olkin (KMO) measure of sample adequacy to assist factor analysis and assess the eligibility of the data for structure discovery.Naturally, the acceptable ranges for KMO and Bartlett's test are larger than 0.500 and less than 0.001, respectively.
Table 2 indicates KMO values greater than the minimum standard of 0.500 and Bartlett's test values of less than 0.05.These values imply many acceptable results for conducting Factor analysis.KMO values way above the 0.5 standards as depicted in Table 2 indicate that there is sample adequacy in this study.P-values less than 0.01 for the Bartlett test in Table 2 confirm that data can be compressed (i.e., using factor analysis) in a meaningful way.
Factor analysis was considered helpful in data reduction since it identified dimensions that may have a significant impact on the dependent variable because all p-values with less than 0.05 and more than 0.500 were met.Table 3 lists the 20 variables that were determined to be significant enough to affect the dependent variable (i.e.shopping frequency) and warrant further investigation.Both items with loadings lower than 0.500 and those loading on multiple components were removed.Overall, it was determined that the 20 items listed in Table 3 may deliver reliable results for this investigation.
The purpose of the study was to assess the significance of retail service quality in improving shopping frequency.Multiple linear regression analysis was utilised to determine significant relationships, hence testing the hypothesis.The null hypothesis of this study proposes no significant relationship between retail service quality and shopping frequency.
In using multiple linear regression analysis to confirm the hypothesis; p-values of less than 0.05 were acceptable to reject the null hypothesis and confirm the alternate.Secondly, multiple R-values were considered to ascertain the nature of the relationship (positive or negative) between the variables of the study.Thirdly, R 2 regression values were considered to confirm variance in data (i.e., change in dependent variable due to independent variables).4 depicts a significant p-value of 0.00 between atmospherics and shopping frequency.This indicates a significant relationship between the variables thus rejecting the null hypothesis.Furthermore, multiple R and R 2 values from Table 5 of 0.152 and 0.023 indicate a greater than zero positive relationship between variables yet not a perfect linear relationship.A 0.023 R 2 value is small and does not explain much variation in the data.
Table 6 shows an insignificant 0.07 p-value between policy and shopping frequency.This represents an insignificant relationship between variables, hence confirming the null hypothesis.Moreover, from Table 7 multiple R and R 2 values of 0.091 and 0.008, respectively indicate a positive relationship greater than zero yet not much variation in the data.
Table 8 illustrates a significant p-value of 0.01 between reliability and shopping frequency.This demonstrates a significant relationship between variables, therefore, rejecting the null hypothesis.In addition, Table 9 shows multiple R and R 2 values of 0.132 and 0.017 respectively.These indicate a positive greater than zero relationship between variables although there is not much variation in the data.
All three predictors of this study (atmospherics, policy and reliability) deduced low R 2 of 0.023, 0.008 and 0.017.Generally, this indicates a very low measure of explanatory power by a model.However, small R 2 values can be significantly different from 0 indicating a statistically significant explanatory power of the model.A low R 2 in this study is attributed to the limited number of predictor variables.This study falls under behavioural sciences and as such numerous predictors of shopping frequency are expected outside retail service quality dimensions which tend to lower R 2 values and are excluded.Other studies in consumer behaviour acknowledge the high possibility of generating low R 2 values (Hair et al., 2021;Moreno et al., 2021).

Discussion
The null hypothesis of this study states that there is no significant linear relationship between retail service quality (independent variable) and shopping frequency (dependent variable).Specifically, predictor variables of retail service quality such as atmospherics, policy and reliability have no significant impact on the shopping frequency of consumers.The null hypothesis was rejected on two predictor variables out of three.Atmospherics and reliability have a significant impact on shopping frequency with p-values less than 0.05 whereas policy showed no significant impact on shopping frequency with a p-value greater than 0.05 (Table 10).
The study's objective was to determine significant relationships between retail service quality dimensions and the shopping frequency of consumers in Durban supermarkets.Consumers' purchasing frequency was found to be significantly and favourably correlated with two aspects of retail service quality (atmospherics and reliability).These results are congruent with the literature that store atmospherics are crucial in achieving retail shoppability (turning shoppers into regulars) (Chatterjee and Shukla, 2020;Behera et al., 2021).Similarly, Dokcen et al. (2021) confirmed retail atmospherics as determinants of store patronage in an emerging economy.
On the other hand, one dimension of retail service quality (policy) has no significant relationship with the shopping frequency of consumers.Explicitly, items relating to personal interaction and principles guiding the conduct of store personnel deduced no significant relationship with the shopping frequency of consumers.Likewise, studies by Ali et al. (2017) and Suhartanto et al. (2019) found personal interaction in form of the helpfulness of personnel of no impact on consumers' overall assessment of retail service.Moreover, Zia (2020) from a Saudi context discovered an inverse relationship and concluded that patrons were uncomfortable with increased personal interactions with store personnel.The study also concluded no significance of policy aspects on the behaviour of consumers.
This research emphasises the importance of store atmospherics and reliability in achieving improved shopping frequency for consumers.It implies that Durban shoppers value store atmosphere, lighting convenience (atmospherics), as well as problem-solving, fair pricing, and security (reliability) in becoming regular patrons of a store.Alternatively, friendliness, helpfulness and return policy (policy) are less valued by Durban shoppers in becoming regular patrons of a store.These inferences indicate the contextual significance of supermarkets in consumers' evaluation of the service delivered.Consumers in a supermarket, therefore, prioritise convenience and problem-solving (Zia and Hashmi, 2019;Zia and Khan, 2018).

Conclusions
According to the findings, shoppers seem to place the greatest emphasis on the atmospherics and reliability components of retail service quality.Therefore, reliability and atmospherics are stronger indicators of high-quality service in a supermarket to increase customer satisfaction.The atmospherics and reliability of service should be the primary management priorities in supermarkets.Establishing and maintaining aesthetically pleasing, contemporary store designs that encourage shopping convenience is important.Fresh produce and high-quality merchandise should always be guaranteed by a consistent approach.Fair pricing of products should be the norm.
The management should guarantee that customers have access to attractive and adequate in-store lighting.Changing lighting affects the mood of customers, improves sales which can be planned or impulsive, boosts customer visits and provides visual attraction.From a virtual

Hypothesis testing results
Retail service quality on shopping frequency reality perspective (Sina and Wu, 2023) discovered cool lighting to be arousing customer interest more than warm lighting.Management should maintain aisle cleanliness as it is important for creating a favourable customer perception.The findings show that aisle hygiene has a considerable impact on consumers' frequency of shopping, and it makes it easier to find merchandise on the shelves.Additionally, enough room must be provided adjacent to checkout counters.This reduces crowding in the store, enhances how customers perceive it and raises customer satisfaction.Managing aisle hygiene and crowding in stores correspond with post-Covid 19 consumer needs and findings of better hygiene and safe distancing behaviour (Nguyen et al., 2021;Itani and Hollebeek, 2021).
The frequency and amount of customer spending can be changed by using appropriate signage.Visually clear and concise in-store signage is important for customers.This makes it possible to save time and decrease the annoyance of customers who must frequently ask store employees for assistance.By using appropriate signage there will be ease of customer movement in-store.Consumers' shopping experiences can be complicated, and an unorganised setting discourages them.However, attractive digital signage in-store waiting areas result in reduced perceived waiting time and increased overall satisfaction with a store (Garaus and Wagner, 2019).
The findings also indicate that Durban consumers do not prioritise policy factors in the quality of retail services and, consequently, in their patronage.Therefore, supermarkets should concentrate less emphasis on interpersonal connections and promote customer autonomy, control and privacy when they are shopping.

Limitations
The study used a purely quantitative approach which provided conservative responses from consumers.A mixed approach could have given consumers more room to express their opinions of retail service quality in supermarkets in Durban.Despite using a representative sample, this could have been further improved by including other cities in South Africa and generalising findings to a broader context.The study focussed on Durban metropolitan areas; the inclusion of lesser developed areas could have provided diversified and more representative results.
Further investigation is recommended on the policy dimension to ascertain the insignificant influence on consumer shopping frequency.Developing complex relationships that include a mediator or moderator could have further improved the study.

Future research focus
This study confirmed an insignificant relationship between policy aspects and the shopping frequency of consumers.Policy aspects were adapted to cover issues of store operating times, return policies and personal interaction.Further examination to justify the reasons for such an insignificant relationship is required beyond this study.
Replication of this study in an online context is recommended to investigate significant dimensions of retail service quality when shopping online versus consumer patronage.There is also the potential of developing indirect relationships with mediators or moderators.
Finally, future research efforts should also include security as part of retail service quality.

Table
Study are original works by the authors