Covid-19 customer experiences of self-service technology across pharmacies in South Africa

Abstract Using time-series data over a period of twelve months, Atlas.ti was employed to analyse the data extracted from Hello Peter, a South African online customer review and business platform, to assess Covid-19 customer reactions to e-services across South African pharmaceutical retailers. The focus was on the two largest retailers as they formed the bulk of the complaints. Multiple variables were grouped into relevant output variables. Content analysis was used as a guide in answering the research questions. Results showed that online pharmaceutical retailers had an attitude of no replies, equally ghosting customers. The study contribution is based on the understanding of customer reactions to the incidents of unacceptable online retailing services. The findings highlighted the profound implications on retailers to comprehend the importance of providing quality customer services to achieve customer satisfaction, leading to the establishment of profitable relationships, while also acknowledging that the implications were caused by the pandemic and could have been different.


Introduction
In December 2019, a severe acute respiratory syndrome (SARS-CoV-2) was first identified and reported in Wuhan City, China and later known as COVID-19. It is highly contagious, spread globally in a short period and was declared a global pandemic by the World Health Organisation (WHO) on 11 March 2020 (Koch et al., 2020;Lone & Ahmad, 2020). As a result of the COVID-19, customers, businesses and governments are still under intense pressure to reform and review the way businesses have been conducted across countries. To halt the spread of the virus, most countries imposed severe restrictions on social life, ranging from prohibitions on major events and the closure of schools and universities to a brief economic shutdown (Koch et al., 2020). In March 2020, due to the pandemic, South Africa was in a tough level 5 lockdown which left the residents confined to their homes with little proactive decision-making individual needs (Kollamparambil et al., 2021). Consequently, shake the South African economy in all aspects such as flow of goods and services, employment security including economic growth.
There is no doubt that the COVID-19 pandemic led service providers and customers to make regular use of online channels using self-service technology. According to Pham et al. (2020), consumers started recognising the benefits of online shopping, especially during the pandemic era. Due to the COVID-19 pandemic, eMarketer forecasted that e-commerce sales would grow by 18% (eMarketer, 2020). Therefore, the readiness of the online channel was heavily tested across the globe within the lockdown period. Makame et al. (2014) mentioned that African countries had a very low adoption of e-commerce, also known as online retailing (e-tailing). Mthembu et al. (2018) indicated that e-commerce was essential to the 4th industrial revolution (4IR) and countries were beginning to include it in their economic narratives.
When dealing with online purchasing platforms, the user's experience is influenced by a variety of factors, including the ease of navigation, aesthetic appeal, level of privacy and security, and the perceived usefulness of the platform (Wu et al., 2016). Online shopping starts when a customer interacts with the retailer's website, social media or software applications up to the correct products being delivered at the right place. Online business channels suffer from huge product return rates because of incomplete and inaccurate product information provided on their e-commerce websites (Vasic et al., 2019). This set of challenges has made it difficult to meet or exceed customer satisfaction. It has also led to a lot more customer frustration as well as complaints. It also exposed retailers' service levels through the internet (De Kervenoael et al., 2016). There seems to be a limited notion of the interface with the software and equipment used rather than the experience after the sale has been closed.
Due to the limitation of movement and risks of contracting the COVID-19 virus, consumers opted for online shopping. The pharmaceutical sector was no exception. The mortar brick retailers then saw a rise in the number of customers buying online. Businesses were looking for improved models of online shopping dimensions that might be used to adapt services to specific clients (Mofokeng & Tan, 2021). According to BusinessTech, 2021), since the initial lockdown, Dis-Chem quickly adapted to the e-commerce environment with the unforeseen benefit of speeding up its digital progress by as much as five years, with online sales increasing by 14.4% over the same period. The COVID-19 outbreak accelerated e-commerce growth in South Africa and beyond, enhancing online shopping exchange ties (Mofokeng & Tan, 2021). For instance, businesses such as retailers who wished to continue operating in the pandemic era were forced to expand their presence online. Today, businesses are in the race to improve themselves to gain superiority over their competitors in the global competitive environment (Akoğlu et al., 2022). These authors fall short of explaining and expanding on difficulties customers face, despite an acceleration in e-commerce.
A few scholars have written about e-service quality and how it affects or influences customer service from a South African perspective. Amongst these are Chinomona et al. (2014) who investigated the importance of e-service quality and contend that online shopping behaviour can be influenced by e-service quality, customer satisfaction, loyalty and perceived value outcomes . Cullen et al. (2022) investigated the factors that affected online shopping during the COVID-19 period exploring personal experience of online shopping as well as interaction with products and the impact of Covid-19 on shopping. Similarly, Rudansky-Kloppers (2014) investigated the factors which influence customer online buying satisfaction in the Gauteng province of South Africa and emphasised the importance for retailers to focus on security, logistics, convenience, price experience as well as product variety. They argue that retailers could use the latter in the design of their websites and advertisements, product descriptions, pricing policies, and delivery methods, resulting in increased online sales, online buying, and online customer satisfaction. Heyns and Kilbourn (2022) sought to understand online shopping behaviour and service quality perceptions of young people. They found that young online consumers consider order fulfilment aspects to be more important than the tangible considerations of online retailers.
Barring the latter study, criticism can be levelled against the previous studies in terms of sample size although generally they agree on the importance of e-service quality. Furthermore there is a clear absence of focus on particularly pharmaceutical retail. It therefore becomes important as it is a needed service during the outbreak of any pandemic.
The objective of the study was to investigate the satisfaction of customers who were shopping online across all lockdown levels imposed in South Africa. The study revealed the readiness of the mortar brick pharmacies, which were also offering online shopping, by analysing the online customer reactions found on Hellopeter platform, during South African lockdown from March 2020 to February 2021. The 508 complaints from Hello Peter were data-analysed, resulting in nine codes. These codes were coherent with the three main codes, namely the online ordering, not replied and customer care landline email. Similar to the above, the below authors have not expanded on the topic regarding non-replies as part of service levels related to customer care. This is despite dealing with user experience. There is an importance to investigate the use of online resources and customer behaviour during pandemics. The fact that many scholars have focused on retail in general and not on pharmaceuticals in particular makes it necessary to conduct a study of this nature. Through this the authors hope to make a tangible contribution to the research by highlighting important aspects of online pharmaceutical retail requirements as well as behaviour. The use of data as specific to a variety of complaints by online pharmaceutical retail customers can provide better insight into this research.
The composition of this study consists of a literature review, research questions, research participants, research design, findings and discussion as well as conclusion and recommendations.

Literature review
The section below will be focusing on a literature review that is done on the topic. It consists of the four sections. The first section provides an overview with a South African pharmaceutical retail angle and its importance during the Covid-19 pandemic. Secondly online customer service with a focus on the South African context, the last two subjects are online performance measures and the self-service technology, particularly in relation to the history and evolution of e-commerce.

South African pharmaceutical retailers and their importance during the COVID-19 pandemic
Retail pharmacies have seen tremendous changes over the previous decade, and these changes appear to be accelerating as retail pharmaceutical businesses face rising external and internal pressure (Wessels & Luiz, 2003). The local pharmaceutical business in a developing country like South Africa must compete in extremely dynamic international markets (Fatti & du Toit, 2013) in an attempt to advance its pharmaceutical industry (Adat et al., 2014). During the COVID-19 pandemic many South African pharmaceutical retailers paid attention to adopting online channels into their operations. All the world's economies are constantly changing and adapting to changes, risks and opportunities as their markets grow or shrink. This is how all global economies work (Fatti & du Toit, 2013). In recent years, South African online retail business has seen significant growth.
Most South African pharmacies (68.3%) are community pharmacies, all in the private sector. Private community pharmacies are classified as corporate (owned by large public or private companies) or independently owned (owned and managed by one or more pharmacists; Ward et al., 2014). Due to the onset of COVID-19 and the resultant regulations, the movement of people to places where they would normally gather in greater numbers or visit stores to shop was limited. This forced retailers to adopt online strategies, using their own in-house platforms as well as partnering services to ensure that clients could still source services from the comfort of their homes. This came with many challenges, as some retailers were not ready to perform tried and tested online retailing services. The restrictions imposed on customers at each level of the lockdown in South Africa had a significant impact on their buying behaviour; nonetheless, online shopping increased by more than thrice during the lockdown. Goga et al. (2019) found that despite the fact the South African online retailer channel was rapidly growing, it had lagged behind its BRICS counterparts and other developed countries such as the USA and UK. Despite the growth in these two leading markets, which are normally seen as a benchmark in terms of developed countries, particularly in the pharmaceutical sector, little to no research has been done on selfservice technology (SST) online retailers service levels in the sector, particularly in South Africa.
The pharmaceutical sector is critical to the delivery of high-quality healthcare services, especially during pandemics, when the supply chain for medicines may become overwhelmed or shut down for a variety of reasons (Tirivangani et al., 2021). Because of a rising demand for preventative medicine, the South African drugstore chain of the Clicks pharmacy group expects full-year earnings to grow (Dudla, 2021;BusinessTech, 2021). Virtual grocery shopping has also entered a commercial mainstreaming phase as retailers seek to expand their client base, particularly in light of the emergence of COVID-19 (Nguyen et al., 2021). This means that there is a need for the monitoring of customer behaviour in order to better understand the customer and explore patterns using the latest technologies such as artificial intelligence to analyse data and spot patterns. According to the Dis-Chem pharmaceutical group, they continue to experience a change in its sales and transaction gross margin mix due to altered consumer behaviour during the varying lockdown levels (BusinessTech, 2021). Consumers agree that if the virtual supermarket is a blend of online and physical supermarkets, it is unquestionably a shopping revolution; hence, they favour this method of purchasing (Nguyen et al., 2021). A shortcoming of this study is the fact that it does not look at the internal sources servicing the customer and how it influences the customer behaviour. The method currently used can be seen as a skewed way of looking at customer behaviour as it excludes looking at post-sale behaviour and what wrongs from the retailer's side influences it.
A customer begins to use an online channel from business websites (BusinessTech, 2021). The online shopping experience does not end with browsing a store's website, placing an order and paying; it is completed when the purchased item is delivered to the specified address (Aljarboa, 2016). Therefore, retailers involved in the supply chain must increase their ability to deliver value to final customers (Fawcett et al., 2014). For instance, more retailers are launching premium customer loyalty programs (Ankeny, 2016). Beside that, delivery speed is an important consideration for online customers (Garver et al., 2012;Gawor & Hoberg, 2019). As a result of these, retailers could be considered as that critical stage in the supply chain. Failure to adopt online retailer channel service can opt to outsource e-services to keep up with the space across the industry. Servicing and evaluating customers purely from a browsing, ordering and delivering perspective can be seen as a narrow way of looking at online customer service. The question that can be asked regarding the above is why retailers are ignoring the assessment of satisfaction levels of customers after the sale? The above authors all do not do so especially on the type of products ordered online as well as the quantity of the products offered and most importantly the speed of refunding the customer in the reverse logistics process to return and replace goods or refund the customer. Further, outsourcing e-services according to Hanafizadeh and Ravasan (2017) has dangers for a company and affects a company negatively from an adoption, cost and process/service perspective. Of importance is the loss of organisational knowledge particularly with regards to the customer when purchasing behavior is known to the retailer. Therefor, the suggestion by Garver et al. (2012); Gawor and Hoberg (2019) to outsource has its inherent risks and is not as simple as the authors suggest, particularly in relation to online customer service.
Pandemics like COVID-19 are expected to become more common as a result of rising globalisation and migration which are linked to high rates of sickness and mortality, with a detrimental impact on communities' socioeconomic position and livelihoods, particularly in resource-limited nations. South African pharmaceutical retailers are therefore also not insulated from the previously mentioned internal and external pressures, and they need to be aware of this reality while actively looking to find ways to improve in this online shopping area.

Online customer service
Customer service is essential for any business growth, survival or even success. Failure to ameliorate its service will have the organisation face consequences of shutting down its operations. During the current COVID-19 pandemic, most businesses, if not all, have rapidly reacted to an online customer service demand as a new or existing market. The development of technologies enables customers to shop anytime, anywhere, and it has increased retailer interaction by providing an additional means of presenting information and content (Brynjolfsson et al., 2013). Customers adopting any form of information communication technology (ICT) to connect with any business may be considered an online customer. These online customers are more likely to express dissatisfaction with and admiration for brands that either fail to meet expectations or offer memorable experiences (Deloitte, 2019). Therefore, a satisfied customer will remain loyal based on the services experienced. Deepening the conversation on the above assertion and challenging it, the following question can be asked: What happens within the company when the customer expresses dissatisfaction? Does the company even care to address the levels of dissatisfaction? In this area there is also no reference to the customer or consumer watchdogs or consumer rights groups in the entire debate on service levels in online shopping and their roles. Hellopeter is but a consumer complaint and commending website but well worth taking into account as an online retailer. There is a lack of incorporating the legality in terms of e-commerce and particularly with regards to dissatisfaction and due process for legal recourse in the e-commerce debate and is not dealt with in the literature. Understandably, the laws of all countries are different, but this aspect can and should be expanded upon in the literature.
A common description of a customer is considered as a party that patronises an organisation such as retailers by purchasing its products (Eresia-eke, 2022). Benton (2020) describes customer service as a fraction of customer demand that is satisfied without delay or received as promised, also known as service level or fill rate. This satisfaction depends on the extent of service that a business is able to offer its customers (Eresia-eke, 2022). It includes quality, cost, flexibility, fast delivery and innovation (Fawcett et al., 2014). According to Benton (2020), customer satisfaction is subject to change. This means each customer has their own customer satisfaction features to achieve the desired end. That would be a sum of activities resulting in the delivery of a product or service to the customer (Eresia-eke, 2022). The author further elaborates that customer service has the potential to clearly differentiate retailers from their competitors and create an unmatchable customer service experience. These customer service experiences in retail chains have become a fundamental component to attract more customers, gaining customer loyalty and determining online retailers' success.
Gaining customers is the lifeblood of retail as it increases sales and boosts the bottomline of any company. Fawcett et al. (2014) point out that loyalty is important because it leads to profitable repeat business. Failure to consider "customer service" as a center stage in retail chains creates customer dissatisfaction. Today, businesses focus more on maintaining and retaining existing customers rather than on attracting new customers (Zare & Mahmoudi, 2020). Research shows that it is 5 to 25 times more expensive to attract new customers than to retain current customers. It can therefore be argued that when not paying attention to the customers that could be lost due to dissatisfaction, it could even be more costly as a result of losing customers and not paying attention to them since retaining costs 5 to 25 times more. Therefore, it becomes crucial to improve service for any retailer's growth and survival as a pillar of succeeding in the online arena. Van Weele (2014) considers this service as a process that consists of a series of activities between customers and systems of the service providers which are provided as solutions to customer problems. Figure 1 shows various elements of customer service that retailers need to cover in their day-to-day interactions with customers. This applies to the brick and mortar as well as online sales.
An e-commerce transaction begins when a seller promotes products on a website and customers express interest in the products, analyse their features, prices and delivery options, purchase the products of their choice, and then check out (Mofokeng & Tan, 2021). According to Meuter et al. (2000), research on e-services was still in its infancy in the early 2000s and no generally accepted theoretical conceptualisation had existed up till then. The two proposed methods to evaluate services were service quality theory as well as a more empirical approach to self-service technology (SST). Service quality has been a topic of extensive inquiry for decades that has now emerged in the form of self-service technology (SST) which has profound effects on the way customers interact with firms to create positive service outcomes, namely customer satisfaction, loyalty and behavioural intentions (Mofokeng & Tan, 2021). Companies are beginning to employ self-service technologies (SSTs) instead of interpersonal encounters during service delivery, since society has transformed and become increasingly characterised by technology-facilitated interactions (Saeid & Macanovic, 2017). Van Riel et al. (2001) have argued that there is a need to better understand consumer evaluations of e-services and to identify suitable determinants of the e-commerce operating environment. They proposed five components in e-services, namely: (1) The core service (2) Facilitating services (3) Supporting services (4) Complementary services (5) User interface through which the customer accesses the services Lin and Hsieh (2006) further support the above by positing that it is necessary to investigate how customers assess the attributes of SST services and how service outcomes are affected. Their results reinforce the fact that technology readiness (TR) plays a role in consumer-SST interactions, Source: Eresia-eke (2022).
Customer service and that firms should therefore pay special attention to TR to improve favourable outcomes of SST implementation. In the context of developing nations, technology readiness is a problem not addressed by authors. Access to data and added to it is the problem of access to technology. These two factors are key to technology readiness not addressed in the literature thus assuming all things being equal between the online retailer and the customer. It is a wrong assumption further deepened by the fact that fewer people have access to online platforms in developing countries, further skewing the picture as the majority do not have access or simply cannot afford it.
E-services, in this context, are therefore seen as supplementary services and associated with the high cost of developing and maintaining a business. Online retailers must guarantee that their shopping websites run properly and adhere to all areas of system quality in order to flourish (Hung-Joubert & Erdis, 2019). The latter authors further stress the importance of companies having to find out if they are positively valued by customers, and how that affects customer satisfaction and e-loyalty. Retailers have exciting and innovative opportunities to respond to changes in consumer behaviour that have resulted in exponential growth in online browsing and purchasing (Pentz et al., 2020). By doing so, online retailers in the South African market can capitalise on the potential loss of R34 billion per year (representative of online purchase abandonment), as indicated by the 2019 South African Digital Consumer Report (Pentz et al., 2020). A direct comparison between e-services and brick-and-mortar services is drawn wherein the customer perceives the e-service as superior to the latter. Santos (2003) also argued that service quality is an increasingly important part of e-commerce. He contends that best quality is a key determinant of successful e-commerce (Santos, 2003). This view is further supported by Yang (2001) who states that service quality in online environments has become recognised as an important factor in determining the success or failure of electronic commerce. According to Pentz et al. (2020) and Pham et al. (2020), online shopping has been a popular modern channel for consumers in recent years, powered by high internet penetration and increased online time spent by consumers. Furthermore, e-service quality may potentially increase the attractiveness, hit rate, customer retention, stickiness, positive word-of-mouth and maximise the online competitive advantages of e-commerce.
The research into self-service technology conducted by Meuter et al. (2000) assessed more than 800 incidents. They found that positive incidents were rare and that dissatisfying incidents were often caused by process failures such as lost orders. Further, Parasuraman (2000) proposed that flexibility, convenience, efficiency, and enjoyment were examples of major positive themes in the online environment, while negative themes included security concerns, risk of obsolescence, personalisation and lack of control. Yang (2001) further highlighted that companies had to facilitate the searching, retrieving and integrating of information if they were to respond efficiently to consumers' enquiries. This was especially necessary for the handling of online complaints.
A number of scholars viewed online purchasing within the pharmaceutical industry from a service quality perspective. Most notable amongst these were Chen et al., 2014), they looked at service quality from the perspective of using electronic platforms as well as the ability to be able to inform customers of their order processing status. This had an important aspect to these factors, as they could go some way toward eliminating unnecessary enquiries in the form of phone calls or emails, as the client could be fully aware of where they were in the ordering process.
Furthermore, the customer is informed of their shipment a day before delivery. Even when looking at the distributor to retail interface in the pharmaceutical supply chain, Mehralian et al. (2016) argue that pharmacy managers are highly affected by qualitative and quantitative accuracy. They further recommend that the implementation of an online ordering system be a major breakthrough for the pharmaceutical supply chain. Holloway and Beatty (2008), in turn, have looked at service quality from an online customer satisfaction and dissatisfaction point of view, with the latter featuring particularly strongly. They argue that less attention has been paid to the drivers of online service encounters and even less attention has been paid to the drivers of online dissatisfaction encounters. A further shortcoming highlighted by these authors is that the literature focuses less on the full purchase process and limits itself to evaluation and interaction with the website (Holloway & Beatty, 2008). They make a contribution through the identification and sub-categorisation of incidents related to satisfaction and dissatisfaction (Holloway & Beatty, 2008).
During the use of SST, the evaluation from four different levels of interaction, namely the information phase, the agreement phase, the fulfilment phase as well as the after-sales phase. One very important element reviewed under these different categories is the return policies. This is particularly important, as online customers are also quick to return whatever they deem to be an incorrect order in terms of product, quality, quantity etc.

Online performance measures
A key feature in evaluation is the incorporation of various types of quality measuring instruments, mainly SITEQUAL to measure internet service quality (Yoo & Donthu, 2001), PIRQUAL to assess perceived internet retail quality (Francis & White, 2002), WebQual to measure website quality (Barnes & Vidgen, 2002;Loiacono et al., 2002), eTailQ to measure eTail Quality (Wolfinbarger & Gilly, 2003), and more recently, E-S-QUAL to measure electronic service quality (Parasuraman et al., 2005).
A major finding made by Holloway and Beatty (2008) using the latter techniques so that the presale dimension (website design/interaction) dominates as a driver of satisfaction (59.4% versus 20.3%), while the post-sale dimension(fulfilment/reliability) dominates in the case of dissatisfaction (64.2% versus 29.2%). This information reinforces the need to distinguish satisfaction and dissatisfaction drivers from one another in evaluating important marketing outcomes. Despite the in-depth analysis of the topic, the authors concede that the sample is not representative in that it includes a very low percentage of minorities in their study. Meuter et al. (2000) define self-service technologies as technological interfaces that enable customers to produce a service independent of direct service employee involvement. They further highlight examples of SSTs such as automatic teller machines (ATMs), automated hotel checkout banking by telephone and services over the internet such as package tracking and online brokerage services (Meuter et al., 2000). This may further be extended to recent developments such as online ordering of food, clothes and gadgets as well as pharmaceutical items. While highlighting certain important scholars, emphasis is placed on further understanding the encounters involving SSTs to identify sources of satisfaction. The categories and examples in Figure 2 are in use and focus on mainly two aspects, namely telephone interactive voice response and internet or online activity.

Self-service technology (History of the evolution of e-commerce)
It is most important in studying SSTs for a company to understand how customers feel about them, how they use the product/s to grow and if they will use them in the future. Understanding of SSTs, therefore, goes hand-in-hand with understanding customer experiences by focusing on sources of dissatisfaction or satisfaction as customers interact with the technology. The most important aspect, therefore, in understanding SSTs has to do with ensuring a balanced sample in terms of customers by including those who use different types of technologies and comparing them. Meuter et al. (2000) also stress that effective management of SST delivery options may be an excellent means of creating a competitive advantage.
By reviewing an internet portal site, Van Riel et al. (2001) also followed customer perceptions of value and satisfaction as well as loyalty. They emphasised how easy it was for customers to evaluate and compare the benefits of competing services and switch to lower costs (Van Riel et al., 2001). This then stressed the importance of ensuring that customers were retained through the development of loyalty by the service provider (Van Riel et al., 2001). While evaluating the quality of e-service in the use of a website, they found that most likely the customer would not evaluate each sub-process in detail during a single visit to a website but would perceive the service as an overall process and outcome unless one of the service elements stood out, especially if positive or negative (Van Riel et al., 2001). This even went beyond a customer's lived experience but also their mood, which could be affected when evaluating and using SSTs and might cause effective reactions such as excitement or irritation (Van Riel et al., 2001). Two of the most important aspects in the US highlighted in this study were e-loyalty as well as e-service quality (Van Riel et al., 2001).
It therefore goes without saying that companies in the online trading space have to pay close attention to the technologies to which they expose their customers, but more importantly, understand the lived experiences of their customers while interacting with the technologies. Providing high-quality services has become a critical priority to gain customer trust and build long-term relationships for all online selling platforms (Shafiee & Bazargan, 2018).

Research questions
(1) Does self-service technology meet the criteria of the core service?
(2) Does self-service technology facilitate customer service in e-services?
(3) Does self-service technology ensure support in pre-transactional and post-transactional e-services?
(4) Do the e-service customers experience the complementary services with the self-service technology?
(5) How effective is the user interface through which the customer accesses the services?

Research participants
HelloPeter is a South African most trusted online review platform connecting South African customers and businesses, on which customers can share and express their satisfaction or grievance from one to another. It was founded in 2000 as a complaints platform for customers to share service experiences to hold business accountable and improve customer service level. Therefore, data was collected from HelloPeter specifically relating to pharmaceutical customers' experiences during the first South African lockdown due to Covid-19 Pandemic. The customers

Figure 2. Categories and examples of SSTs in use.
Source: Adapted from Meuter et al. (2000) reviewed the online service of two of the biggest pharmaceutical retailers in South Africa. This type of data is readily available for the public online and required to filter based on the research scope.

Research design: qualitative approach
The initial aim of the study was to find out how brick-and-mortar businesses were handling online shopping during Covid-19. Having found more than 1,000 online reviews of the pharmaceutical industry, closer to half were complaints about online shopping. According to Chun Tie et al. (2019) grounded theory study, qualitative and quantitative data collection procedures can be utilised to establish, uncover, or construct a theory from systematically collected and comparative-analysed data. Ground Theory became the most suitable analysis method for the study until the authors realised that the main challenge was the application (APP) that the customers used to do shopping. The philosophy and fundamental concepts of techniques, study objectives and questions, and designs and data collection criteria distinguish qualitative from quantitative methodologies, (Vaismoradi et al., 2013).
Furthermore, Vaismoradi et al. (2013) emphasise that the objective of qualitative approaches is to gain an understanding of a particular occurrence from the perspective of individuals experiencing it. The ineffective and inefficient nature of the APP caused several other challenges which the online customers accounted for. The main characteristics of qualitative methodologies are a belief in multiple realities, a commitment to identifying an approach to in-depth understanding of the phenomenon, a commitment to participants' perspectives, conducting inquiries with minimal disruption to the natural context of the phenomenon, and reporting findings in a literary style rich in participant commentary, Vaismoradi et al. (2013). The researchers realised that there is an SST framework which aims at ensuring that the APP and other customers' service experiences are effective and efficient. Ground Theory would have been an option for the study, but because there was no need to come up with a new theory, content analysis became the better option. Through initial codes and categorizations generated from existing literature or a theoretical framework and applied deductively to categorise the data, Content Analysis seeks to add credibility to or conceptually extend a theoretical framework or theory (Leung & Chung, 2019).
This study falls short in epistemology because it was not yet known that APP was the causal mechanism for most of the customers' complaints before the study was conducted. Axiology also falls short in supporting this study due to the fact that what is valued was not known at the time of conducting the study other than the value of what it costs to attract and keep the customers (5 to 25 times), as previously discussed in the above literature. Ontology was then adapted for the study. In order to understand the relationships between online customer applications (APP) and other complaint elements found in the reviews (Figure 1.4), we must reflect on the reality of what really exists. The philosophical stance in this study is critical realism, due to the APP being the independent variable that causes the existence of the other elements in Figure 1.4. Further, this stance is also exacerbated by the attitude shown by the pharmaceutical retailers in their refusal to compromise their stance and extreme position by not answering or responding to the online complaints of their customers.
Quantitative content analysis is a research method in which features of textual, visual, or aural material are systematically categorised and recorded so that they can be analysed (Coe & Scacco, 2017). 508 reviews were extracted from the Hello Peter website, where the customers' reviews were mostly in the form of complaints about online customer service. Content analysis has been defined as a systematised, repeatable technique for condensing numerous words of text into a smaller number of content categories based on clear coding principles (Stemler, 2000). The objective is to achieve a concise and comprehensive explanation of the phenomenon, and the result of the analysis is concepts or categories that describe the phenomenon (Elo & Kyngäs, 2008). The study was using SST as a phenomenon to understand the complaints which are made by online customers and how the SST impacts both the retailers and the customers.
The study followed the quantitative content analysis steps presented by Coe and Scacco (2017).

unitizing (segmenting the texts for analysis),
Data was collected from two pharmaceutical brick-and-mortar stores, which are also selling products online. The data collected was the reviews that were made from 2 March 2020 to 23 February 2021. When it came time to count, there were 1557 reviews. When it was boiled down, only 508 reviews were related to the online customer service. 16 codes were created using the reviewing and "search and code" commands on Atlas Ti. The website or customer application had 18 quotations. In the 18 quotations of APP, code 10 codes have a relationship with the App. Kim et al. (2018) also recommend simple random sampling when deciding on reducing or choosing a certain sample size of social media text data. Therefore, they refer to Cavazos-Rehg et al. (2016), where they did a keyword search on tweets collected Between a certain time frame, they randomly selected a certain number of tweets related to what they were researching.

Sampling (selecting an appropriate collection of units to analyse),
The study used simple random sampling to select an appropriate data set from 1557 reviews, and 508 reviews were randomly selected as related to shopping online. This article tests the efficiency of simple random sampling and constructed week sampling by varying the sample size of Twitter content related to 2014. Simple random sampling is more efficient than a constructed sampling plan. According to Kim et al. (2018), the two types of analysis for social media data appropriate for this study are thematic and content analysis. The thematic analysis was not used because it analyses qualitative data. In this study, content analysis was employed because the data was analysed quantitatively even though it was following the SST framework. The method of content analysis is typically defined as the "systematic, quantitative investigation of message features." Ullah et al. (2018) The two main content analysis sampling techniques according to Kim et al. (2018), are the constructed week sampling technique and simple random sampling.
Constructed week sampling evaluates or analyses data on a weekly basis, whereas this study has done analysis over a longer period from 2 March 2020 to 23 February 2021 simple random sampling was more appropriate.

Data reliability
Reliable data are those that remain constant despite alterations in the measuring technique, according to Bolognesi et al. (2017). The APP code is an independent variable, while the 10 codes are dependent variables. The study is analysing the relationship of these 10 codes with the App code. The coefficient was conducted using ATLAS.ti to show the relationship between the independent variable and the dependent variable. The document was read manually for review and the authors also did manual coding using the ATLAS.ti coding. Further, machine learning did "search and code" automatically from the entire document in the ATLAS.ti. The "Find Redundant Coding" feature of ATLAS.ti was then used to ensure that there were no duplicate quotations. Therefore, the data coded, all the coefficient values are below 1. There is no duplication, and its relationships were clearly interpreted using the exact quotes from the customer's complaints. Where the customer's complaints are related to both dependent and independent variables.

Data validity
In the event that researchers have a theory or conceptual framework in mind, they may employ a deductive procedure in which the current theoretical constructs are compared to the data observed from a group of participants in a particular situation; this approach validates the theory's credibility (Leung & Chung, 2019). The constructed elements of the SST framework were compared with the coded data to check if the APP is able to meet all the requirements of the framework. In this way, researchers were able to see what was lacking in the APP of the pharmaceutical retailers in order to meet the efficiency and effectiveness of the SST framework.
The study has combined content analysis and ATLAS.ti qualitative tool, following the steps for data collection: (1) 508 reviews were collected from the Hello Peter app.
(2) The reviews were turned into PDFs and downloaded into ATLAS.ti.
(3) A Word Cloud was developed from the uploaded PDF documents.
(4) With the assistance of Word Cloud, words were constructed.
(5) Familiar constructed words were then put into a "Search and Code".
(7) Following the SST theme of this study, the online shopping application "APP" became the main code.
(8) The APP code was then measured for co-occurrence with other codes.
(9) A coefficient analysis was conducted between the APP code and other corresponding codes.
The study used deductive content analysis as it followed the SST framework. The critical case sampling. Here, individuals, groups, or settings are selected that bring to the fore the phenomenon of interest, such that the researcher can learn more about the phenomenon than would have been learned without including these critical cases (Omona, 2013). Coding is a way to conceptualise data by comparing incidents and concepts over and over again (Walker & Myrick 2006). The study used selective coding because the central category, in this case, the APP code, correlated with the selected codes. The study identifies the category and is related to other categories.

Data analysis
The 508 complaints were extracted from the Hello Peter app. These were the complaints received from 2 March 2020 to 23 February 2021. ATLAS.ti is a strong workstation for qualitative data analysis, particularly for vast portions of text (Smit, 2002). The complaints were downloaded onto a Microsoft Word document and then turned into a PDF document, which was then uploaded into ATLAS.ti. The first step was to create a Cloud from the complaints from the customers who had been buying online. In the Cloud, the two main words that were removed were the names of the two pharmacies about whom the customers had complained. The threshold of the words started at 111 and went up to 1 107. The Figure 3 on Word Cloud was used for the selection of coding. Terms like "missing order", "bad service", "missing items", "poor service", "customer service", "no delivery" and "no response or feedback" were searched from the document search of ATLAS.ti as illustrated below.
One of the most visible terms in the ATLAS.ti Cloud was "not replied" because the data was collected from the Hello Peter website. "Not replied" is the term taken from the Hello Peter app which shows that the pharmacies do not reply to complaints from their customers. "Not replied" became one of the codes due to its visibility out of the 16 codes in the ATLAS.ti Cloud. The overall groundedness of the "not replied" term was 432. This was because of all the complaints to which the businesses did not reply. Figure 4 shows the coefficient of the online shopping app and the "not replied" code, the leading code before the customer service. The customer care email is the third leading code. This is because there is always an email address which the business offers its customers to send their enquiries. Unfortunately, the link between the customer care email and the online shopping app shows mainly that customers do not receive replies to the emails and when they do receive replies, they are not quite welcomed by the customers. In fourth place, the customers simply refer to the customer service as "bad service". In Figure 5, the "bad service" code is clearly visible, and when Source: Authors compilation matching the term on Search & Code, overall, there are 54 quotations and six coefficients with the online application app. Such experiences deem customers blind to the complementary services that come with the whole experience of buying a certain product. The bad service on the comments quotations shows that customers are unhappy even after they had to experience the challenges with the app. There are further services where they report the app is still not helpful. Figure 5, which is called the Sankey diagram, shows the codes which have the coefficiency with the online application.
The APP refers to the online shopping app, the interface of self-service technology software which customers use to make sales interactions with the retailer. The APP was selected because it is the main component used by customers to buy products online. APP is the main aspect that connects the retailer with the customer. The online application APP is the main device between the business and the customer. It is used in Figure 5 to look into its integration with other codes or complaints. The Sankey diagram is the result of analysing the code "co-occurrence" in the ATLAS.ti table. The online shopping APP is the main code with which the other codes co-occur. The leading codes, which correspond, are "not replied" and "customer services". The aim is to see if this selfservice technology meets the core service criteria of the customer and the effectiveness of the user interface of the technology provided.
Through the APP, the customers mainly complain about "customer service", "stock-out" and "bad service". These are the three main codes which interact with the APP. Of the 18 APP quotations, eight relate to "customer services" as a code in which customers express their disappointment with the online app, specifically on app failure, tracking systems, confirmation, and price discrepancies. Figure 6 shows the exact comments made by the customers. The shaded wording are concepts which may link the reader to the actual name of the pharmacy. It is shaded so that the actual name is not revealed. The customers show that there are several challenges with the online shopping app which results in their being dissatisfied with the available customer services. The figure also shows that the APP, as a code, has a groundedness of 18, while "customer services" has a groundedness of 105. The groundedness reveals that there is other customer dissatisfaction relating to shopping online other than on the online shopping app.
The "not replied" code is the highest, with 13 quotations that correspond with the online shopping app. Most quotations are from the Hello Peter app, which shows that with each complaint made by a customer, the business (pharmacy) has not replied to the customer's query. There are also a few comments from customers which show that even with the use of the customer care line and email, there was still no reply from the business. It is unfortunate that customers have to experience challenges during the transactional process. Post-transactional, the D2 to D6 in Figure 6, the comments, show that there are other codes which are linked to the same comments. Therefore, in one quotation, you will find that customers complain about the app and not getting the favourable response they expect when trying to follow up on the challenge they have come across.
The online application shows the coefficient codes which are the main complaints of the customers. The leading codes are "customer services", "bad service", "no feedback", "returned", "no order", "no delivery", "no stock", not replied" and "incomplete orders" in Figure 7. Amongst all of these elements, "customer services" has 508 quotations, meaning it is the most highlighted code with quotations. Customers mostly complain about how the pharmacy businesses handle the online customer services which include all the other codes when a complaint is lodged. "Customer services", with a coefficient of 0.07, leads in the online shopping app. The second-placed leading complaints are the three codes of "bad service", "no feedback" and "returned" which are all at a coefficient of 0.06. Figure 7 shows the four quotations referring to "bad service" as no fruitful feedback from the business and matters being dragged out for a month, with the customer care service line also not working. The "no order" and "no delivery" complaints have a 0.05 reading as a result of the customers successfully ordering and making payments on the app, but then the products are not delivered to them, or later in a follow-up, the customers realise that the order has never even reached the business.

Findings and discussion
According to literature and industry research reports, the large pharmacy groups did exceptionally well in terms of forecast and actual profitability margins. However, customer complaints were evident in-store. The Google review website and Hellopeter platform showed that companies could do much better considering their online customer services.
The core service in self-service technology is the retail products the pharmacies sell online to the customers. There are issues regarding this technology, as online customers experience poor service due to stock-outs, for example. Despite the latter, in most cases, the products are still advertised on the business website. Customers may even go as far as purchasing the products Source: Authors compilation online without realising that they are out of stock. Self-service technology now has an impact on the core service of the pharmaceutical business. Most of the retailers in mortar and brick stores always ensure that there is no stock-out of products, as this may have a bad impact on both the customers and the store. The same applies to online stock-outs, and it could have even been better if, when the customer makes the purchase, they are aware of the remaining stock and/or that the products are out of stock. This will enable customers to stop the final process of purchase, as the current situation, according to the findings, causes customers to further experience the challenges of refunds and very late deliveries.
The facilitating services pay more attention to the services offered on the online shopping application and delivery. These two are crucial. The APP needs to be user-friendly. It needs to be quick in detecting that the products are no longer available in the store, specifically the ones the customer intends to purchase. It also needs to have an efficient tracking device for the delivery of the products. Several customers who have complained have also experienced no delivery and delays in delivery. This shows that there is a lack in the facilities' services if the customers are in a position to buy unavailable products and only when the product is about to be delivered it becomes known to be out of stock. The functionality of the online shopping application is crucial in that it needs to integrate the pre-transactional, transactional and post-transactional. It needs to be fully functional at all times. In the case where the customers have gone through all the processes of buying and interacting with the application but the result is "no order", even when they have already paid for it, this simply shows the dysfunctionality of the application.
If it happened that the customer was physically in the store, the consumer was going to be assisted by the employees. The facilitation support would be readily available; so, in the case of the SST, the customer would have to be assisted by the features of the website or the online shopping application and other post-transactional activities. In the absence of physical employees, the most efficient communication is in the form of instant messaging and telephonic conversations. Even in the case where the customers do not get any response from the retailer, they become frustrated and end up using other external self-service technologies such as the Hello Peter website to lodge their dissatisfaction. Even in these cases, the customers do not receive any assistance because the business has not replied. In cases where the supporting services are not fully helpful to the customers, they do not only result in dissatisfied customers but also lead to helpless customers. The online support services must be efficient and reliable, and communication should be key as well as ensuring the user-friendliness of the services.
The complimentary services are crucial to a business, as they offer the business the opportunity to leverage its reputation and advertise more products on offer. The complimentary services may be in the form of discounts, community and health service information, and competition draws. But if the customers came across challenges when using the online services, they would not even be able to notice the complimentary service or all the good which is experienced when using the application will be deemed by the final services, in cases where refund could not be done timeously without the effort of the customers and the cases where delivery is delayed and the is no prompt communication with the reason it is delayed. The point is that the complimentary services are only seen and acknowledged by the customer when the customer is satisfied with the entire purchasing process.
User interface is a tool through which the customer accesses the services. The user interface is that tool that allows you to interact with your computer. The customer must interact with the technological application in the same way that the customer will interact with the store's support staff. In this case, the web application informs the customer that a specific product is available, only to discover later, particularly during the anticipation of delivery, that the product is out of stock. The user interface should be able to provide self-service customers with the same level of service as if they were in a store. If a customer enters a store looking for a specific product and is unable to locate it, he or she will most likely seek out a store assistant to inquire about the product and, before leaving the store, will be able to determine the state of the product the customer is seeking. The same should be applied to the user interface application in order to know the status of the product before making an online payment.

Conclusion and recommendations
The implication of online customer services, particularly with the two main pharmaceutical retailers, was motivated by the pandemic when most customers who were using brick-and-mortar shops opted for online shopping. The pharmacies reported success in their online shopping during the period of the hardback in South Africa, meaning that out of the online shoppers there were several who had experienced the challenges and found themselves having to use external SSTs to showcase their concerns. This was due to the emergence of the whole matter where customers were either using the online service for the first time and the retailers receiving a higher volume of online requests than usual with new issues coming in place like stock-outs, deliveries, customer support and other technical challenges which the customers had experienced.
It is really unreliable to wait for days on products or services due to the fact they have requested from an online platform of the service provider located within the surrounding areas of the customer. However, it is understandable to wait for a long period on products or services procured from other parts of the globe. South African retailers need to demonstrate to customers that it is better to shop online than walk-ins in the stores, while also ensuring reliable service delivery and adopting social distance regulations. The need to adopt a sense of emergency service across pharmaceutical retailers is necessary based on the nature of its business.
As major South African grocery retailers have understood their business nature, they are now competing on delivery speed and affordable cost. One of the giant retailers has ensured quality delivery within sixty minutes. It is recommended that South African pharmaceutical retailers monitor their websites and apps in order to root out all pitfalls for a smooth logistics process, pretransaction, transaction and post-transaction, regarding online customer services; this will prevent customers from seeking alternative ways to voice poor service on online platforms such as Google review, Hello Peter and other social media. The data collected was during the South African first lockdown period due to COVID-19, it could be that if there was no pandemic, such number of reviews particularly the complaints could be less. It could be that the pharmacies were experiencing a higher number of sales and not fully prepared to serve such high numbers of online customers or neglected the online channel. Future research can look into the online customer reviews post COVID-19 or lockdown and could also research on South African pharmaceutical reverse logistics particularly of online returned products. From a managerial point of view the study has value and significance in that it can help managers in pharmaceutical retail to plan stock levels accurately as well as sort out product correctness to reduce the number of return products and create online customer satisfaction.