The Impact of Artificial Intelligence on Customer Satisfaction in China’s Online Retailing Industry, Case Study Taobao (Alibaba)

1.0 Researchable Topic Area

1.1 Background

AI is widely used in marketing field and projected to generate greater influences on competition landscape. It estimates that AI will replace 7 million jobs in the UK and create 7.2 million new jobs between 2017 and 2037 (Forbes, 2020). AI is promised to have far-reaching and significant impacts on economic, legal, and political environment (Forbes, 2020). The revenue of AI in China is promised to reach about $750 million with a 200% year-on-year increase rate (Yang and Goh, 2019). Chinese online retailers are using AI to evolve their retailing businesses (Noei, 2019). In China, AI is used not only for unmanned stores but also for an integrated and highly efficient retailing system featuring multiple technologies and implications: QR codes, Radio Frequency Identification Devices (RFID), machine vision, customer traffic recognition, digitalised offline retailing, and optimised supply chain (Noei, 2019). According to Noei (2019), Chinese online retailers are using AI to realise smart product search and recommendations, Chatbot, and pervasive personalisation. Meanwhile, Taobao is one of the largest online retailing platforms owned by Alibaba Group. It had over 299 million daily active users and reached 5.7 trillion (Yuan) in 2019 (DMR, 2020).

1.2 Business Problem

Chinese online retailing industry has accomplished an exponential increase with a CAGR of 58.58% between 2010 and 2019 (Allison, 2020). With the trend that the growth is slowing down, the competition among online retailers are becoming increasingly more intensive (Allison, 2020). Taobao had about 10 million stores in 2017 and it increased 30,000 stores per day averagely during the burst of COVID-19 (Shumin, 2020). With such large number of competitors, stores in Taobao have to identify effective approach to improve their customer satisfaction and thus competitive advantage to secure their survival.

1.3 Rationale of the Research

The research has significant contributions to the marketing management of stores in Taobao. It focuses on Taobao so as to provide the most applicable and fittest recommendations to these stores. It can help these stores to use AI to facilitate their customer satisfaction and thus competitive advantages.

Also, this research can make significant contributions to academic field by covering the research gap. This research finds that there is no academic research covering impact of AI on customer satisfaction in Chinese online retailing industry because the use of AI is new topic. Daqar & Smoudy (2019) found a positive relationship between AI and customer experience. Some studies investigated the impacts of AI on marketing (Davenport et al., 2019; Khokhar & Chitsimarn, 2019; and Jarek & Mazurek, 2019). However, none of these studies clarified the relationship between AI and customer satisfaction especially in online retailing industry.

2.0 Objectives for the Research

This research targets at the relationship between AI and customer satisfaction in the context of Taobao. The purpose of this research is to study the impact AI on customer satisfaction, identify how to use AI to improve customer satisfaction and thus make recommendations to the retailers in Taobao. It aims to identify how to use AI technologies to effectively improve customer satisfaction and then make recommendations for retailers in Taobao. The expected results can help the online retailers to strengthen their customer satisfaction.

Research objectives:

1) To develop a literature review related with the impacts of AI and customer satisfaction to form a theoretical framework

2) To identify what factors of AI can affect customer satisfaction based on the primary data collected by questionnaires and interviews

3) To discuss the findings with the literature review to make conclusions and recommendations for the retailers about using AI to improve customer satisfaction

3.0 Literature Review

3.1 AI for Marketing

AI tends to generate significant impacts on marketing strategy and customer behaviour (Davenport et al., 2019). Business models, sales process and services are changing because of the rise of AI. There is a trend that AI is becoming a mainstream technique for all retailers (Davenport et al., 2019). However, André et al. (2018) highlight that prevalence of AI is dangerous. Felming (2019) questions the implication of AI by revealing its problems including data privacy, algorithmic biases, and ethics. Given the contrasting viewpoints about AI, it is significant to study the impact of AI in marketing field.

With the help of AI, online retailers can analyse customer data and predict their behaviours and preferences (Agrawal et al., 2018). Furthermore, they can offer customers personalised recommendations and marketing messages based on the predictions of AI (André et al., 2018). AI is likely to replace salesmen or reduce their importance to sales processes. Consequentially, it can evolve the whole sales processes. More importantly, marketing strategy is moving toward personalisation with the help of AI (Davenport et al., 2019).

Davenport et al. (2019) summarise some AI tools which are related with online retailing. AI technologies such as Brichbox help retailers to have more accurate and wider predictions of customers including preferences and demands. It may lead to a business model that focuses on customer predictions. Sales AI technologies allow retailers to automate some sales processes by replacing human by robots such as Chatbot in order to reduce costs. Some AI robots help marketers to automate some customers services thus contributing to service quality.

The mainstream view is that customers have negative opinions to AI (Castelo et al., 2018). Customers believe that AI is unable to feel, identify and respect their uniqueness and they are not willing to use AI (Castelo et al., 2018). However, Daqur & Smoudy (2019) find that AI is conductive to customer experience and their relationship is significant. This study focuses on two parts of customer experience: services and after-sale services. It finds that AI facilitates personalisation in throughout the whole consumer-buying process and thus improves customer experience. Thus, the study suggests the use of more personalised services and reduce waiting-time in after-sales services.

3.2 Customer Satisfaction

Customer satisfaction refers to a status that perceived quality meets or exceeds expected quality (Barboza & Roth, 2009). Customers evaluate their consumption by measuring the gap between their expectation and their experience (Rust & Anthony, 1993). If perceived quality meets or exceeds expected quality, they are satisfied. On the other hand, perceived quality under expectation causes customer dissatisfaction. Customer expectations are strongly related with customer satisfaction. These expectations are related with internal factors such as customers’ experience and external factors such as as prices.

Customer satisfaction has significant and far-reaching impacts on corporate performance (Anderson & Sullivan, 1993). It strongly correlates with customer loyalty and retention as well as sales performance and is viewed as an index of corporate performance (Barboza & Roth, 2009). High level of customer satisfaction is conductive to word of mouth, corporate reputation, and brand equity (Neupane, 2015). Strong customer satisfaction is regarded as a source of competitive advantage. Given the significance of customer satisfaction, studying how to use AI to develop it is prominent.

3.3 Theoretical Model

Customer satisfaction is related with service quality, price, and customer relationship (Oh & Kim, 2017). However, AI could facilitate these factors and then customer satisfaction.

Service Quality

Service quality directly relates with perceived quality and thus affects customer satisfaction (Carman, 1990). Oh & Kim (2017) prove that services quality positively affects customer satisfaction. Service quality can be measured by SERVQUAL dimensions including reliability, assurance, tangibles, empathy, and responsiveness (Carman, 1990). Reliability means the competency of offering the promised service in a dependable and accurate manner. Assurance refers to employees’ capability, knowledge and willingness to show their confidence and make customer trusts. Tangibles refer to physical environment including facilities, equipment, employees and communication tools. Empathy is the dimension of showing concerns and respects to individual customers. Responsiveness measures the willingness of supporting and assisting clients and offer them fast services.

AI may contribute to empathy, as it allows retailers to offer personalised services and personalised marketing messages (Jarek & Mazurek, 2019). The personalisation can make customers feel that they are important and being concerned. Also, AI can offer customers fast services by automation (Prentice et al., 2020), which can improve responsiveness. For example, chatbots can response customers in a fast way and 7 days for 24 hours. Furthermore, retailers can use AI to improve the reliability of this services (Prentice et al., 2020).

Hypothesis 1

H0: there is no correlation between customer satisfaction and AI-oriented service quality.

H0: there is a correlation between customer satisfaction and AI-oriented service quality.

Price

Price strongly affects customer satisfaction (Foster, 2016). Customers use price to make expectation on their purchase, and they tend to have greater expectation when they perceive that the price is high (Foster, 2016). When the expectation is high, it is harder to satisfy customers.

AI collects customer data and helps retailers to facilitate price optimisation. With the help of AI, the prices can be more dynamic and attractive (Jarek & Mazurek, 2019). Through automation enabled by AI, retailers can adjust its prices to each customer based on the purchase record (Jarek & Mazurek, 2019). AI improves pricing strategy which can contribute to customer satisfaction.

Hypothesis 2

H0: there is no correlation between customer satisfaction and AI-based price optimisation.

H0: there is a correlation between customer satisfaction and AI-based price optimisation.

Customer Relationship Management (CRM)

Customer relationship has a positive relationship with customer satisfaction (Swift, 2001). When customers perceive that they have a good relationship with service providers, they are more likely to satisfy (Swift, 2001). Therefore, CRM is an important task for all companies to develop customer loyalty and retention (Sin et al., 2005). Nevertheless, before the age of AI, CRM tended to be costly.

AI helps companies to automate CRM programs and improve their efficiency and effectiveness. Based on AI, companies can realise high level of personalisation featuring personalised marketing messages, personalised communication, and personalised purchase advises (Jarek & Mazurek, 2019). The personalisation makes customers feel that they are being concerned and bring convenience to them. Thus, AI-based CRM may be able to improve customer satisfaction.

Hypothesis 3

H0: there is no correlation between customer satisfaction and AI-based CRM.

H0: there is a correlation between customer satisfaction and AI-based CRM.

4.0 Details of your research

4.1 Realism Philosophy

Realism is the fittest philosophy for this research, which allows researchers to collect and analyse both quantitative and qualitative data. This philosophy stands on the assumption that reality is independent from people’s mind and the only way to develop knowledge is scientific way (Saunders, Lewis & Thornhill, 2012). Direct realism highlights that human senses can describe the world, which is consistent with qualitative analysis. Nevertheless, critical realism argues that human perceptions cannot describe the real world, which aligns with quantitative analysis (Novikov & Novikov, 2013).

Through direct realism, this research takes advantage of qualitative analysis to explore deep insights by making the researcher directly experience the research phenomenon. The research plans to perceive the phenomenon by engaging in data collection process. Qualitative research is designed to study feelings, behaviours, psychologies, experience and so on (Saunders, Lewis & Thornhill, 2012). Customer satisfaction in Chinese online retailing industry are closely associated with these elements. On the other hand, it is impossible to quantify all feelings and behaviours. Thus, it is suitable to apply qualitative research.

By critical realism, this research can conduct quantitative analysis in a scientific way and thus generate reliable findings. The results of quantitative researches are less arguable and more reliable as these researches have strict research framework and well-designed research methodologies (Saunders, Lewis & Thornhill, 2012). In this sense, researchers cannot affect their research in the process of data collection and analysis.

This research plans to use qualitative research to dig out deep insights and then implement quantitative research to generalise these insights. It gains insights from the stores in Taobao and generalises them by studying consumers’ responses.

4.2 Inductive approach

Given that there is no previous research studying on the relationship between AI and customer satisfaction, this research plans to use inductive approach to explore the relationship. It focuses on observations to dig out new insights about the impact of AI on customer satisfaction (Novikov & Novikov, 2013).

Deductive approach is not suitable for this research because it cannot help the researcher to study AI and customer satisfaction and gain new insights. It relies on existing knowledge and theories to explain a phenomenon, which is more suitable for descriptive and explanative researches (Novikov & Novikov, 2013). However, this research aims to explore the relationship and find out new insights. Thus, inductive approach is much more suitable for this research.

4.3 Survey

Survey has a variety of data collection tools such as questionnaires and interviews enabling this research to collect both qualitative and quantitative data. It is an effective strategy to build a larger size of samples especially within a relatively short term. The nature of a survey is to question a respondent about a topic and then record their responses (Jackson, 2011). Survey is easier to manage and implement than experiment and focused group. It can be operated in a more effective and economical way than other strategies such as observations (Denscombe, 2010). It is easy to analyse data collected by survey. Generally, survey is appropriate for students who are constricted by budget and time.

4.4 Questionnaires and Interviews

Questionnaire is widely used to gather quantitative data because of its high efficiency and effectiveness. It enables researchers to collect a huge size of samples in a fast way. Also, each questionnaire contains many questions which can quantify data and collect a huge amount of data (Monette, Gullivan & DeJong, 2010). By the internet, questionnaires can be spread widely and rapidly. The IT technologies can automatically record data. Meanwhile, the researcher does not need to involve in data collection progress, which enables him to do other research tasks (Albery & Munafo, 2008).

This research plans to collect 200 questionnaires from the users of Taobao by the internet. It will build its questionnaire-website in WJX.cn and share the link of the website to social media including WeChat, Weibo, and Zhihu. The questionnaire has three parts of questionnaires with different purposes: Part 1 examines the extent to which participants used Taobao; Part 2 collects data to measure relationship between AI and customer satisfaction; and Part 3 measures the participants’ demographics.

Interviews are applied by this research to collect qualitative data. By interviews, the researcher gathers data through a conservation with respondents and directly experiences the impact of AI on customer satisfaction in Taobao. The researcher can use interviews to dig out insights. Interviews are less constricted by traditions and methods and allow researchers to have a broad viewpoint to explore out details (Monette, Gullivan & DeJong, 2010). Interviews can discover variables which are neglected by research framework (Polonsky & Waller, 2011).

This research plans to use the internet to conduct 10 online interviews. It selects interviews and interviews them via WeChat. Semi-structured interviews are applied, so the researcher to have autonomy to ask personalised and tailored questions to respondents based on their response. The researcher observes the respondents’ facial and quickly analyses their answers to propose personalised questions to dig out deep insights.

4.5 Convenience Sampling Technique

Convenience sampling is the most suitable technique for this research. It collects primary data from those participants who are the easiest for researchers to access, so it can build a huge size of samples in an efficient and effective way (Albery & Munafo, 2008). This research only has about 3 months for data collection, which means that it needs a highly efficient way to collect data. With the help of convenience sampling, the researcher can collect data from his friends and motivate these friends to further spread the questionnaires. By fully using his social networks, the researcher is able to complete data collection within regulated time.

On the other hand, probability sampling does not fit in this research. The pre-condition of implementing probability is the mechanism ensuring that each of research population has the same chance to be chosen (Albery & Munafo, 2008). Obviously, the researcher does not the capabilities to access all store owners and customers in Taobao. Therefore, the use of convenience sampling is realistic and manageable.

4.6 Data Analysis

For quantitative analysis, statistical analysis tools are implemented, including frequency, correlation, and regression analysis. Frequency analysis is used to analyse each question and demographics of participants. Correlation analysis examines the strength of the relationship between AI and customer satisfaction. Furthermore, the research uses regression analysis to measure the significance of the relationship.

For qualitative analysis, this research adopts thematic analysis to address qualitative data. Through thematic analysis, the researcher analyses primary data from the interviews based on theories but beyond words to explore out ideas.

5.0 Research Ethics

This research fully complies with the University’s Ethical Code. This research will inquire and record the identity and contact information of respondents. It is frank to all respondents and reveals its real aims and objectives by a consent letter. This letter also informs respondents what kind of questions they will be asked. It notices them that they have right to quit this research or withdraw this data within one month. So, they can understand this research before they decide to offer data. The data will be locked in the researcher’s computer and accessible only to the University. It will be destroyed after 8 months. The questionnaires and interviews contain no sensitive questions that may arouse negative feelings of respondents. Furthermore, this research involves no commercial secrets and debriefing to avoid controversy.

6.0 Conclusions

This research aims at the correlation between AI and customer satisfaction in the context of Taobao. It has the purpose to investigate the impact AI on customer satisfaction, identify how to use AI to improve customer satisfaction and thus make recommendations to the retailers in Taobao. Previous studies show the customer satisfaction has significant and far-reaching influence and AI is becoming prevalent. This paper uses previous studies developed a theoretical framework including price, CRM, and service quality. AI can improve the three factors and thus customer satisfaction. The research hypotheses have well justified. Then, the paper designed research methodologies including realism, quantitative & qualitative research, inductive approach, questionnaires & interviews, convenience sampling technique, and statistical analysis & thematic analysis. The research ethics is well planned, and timetable is presented in the following section.

7.0 Timetable for your research

Month	Dissertation Activity (parts)
Month 1	Collecting journal articles and academic books related with the impact of AI on customer satisfaction Revisiting the research proposal
Month 2	Developing literature review Designing research methodologies Designing questionnaires and interviews
Month 3	Collecting data by questionnaires and interviews
Month 4	Continuing to collect data and focus on interviews
Month 5	Organizing and analyzing data Discussing the findings with the literature review Writing conclusions
Month 6	Writing-up the whole dissertation Making modifications

Table 1

Reference

Agrawal, A., Gans, J. S., & Goldfarb, A. (2018). Prediction machines: The simple economics of artificial intelligence. Harvard Business School Press

Albery, I. & Munafo, M. (2008) “Key Concepts in Health Psychology” SAGE Publications

Allison (2020) Key Growth Engines of Internet Retail in China, Retrieved from: https://daxueconsulting.com/internet-retail-in-china/

André, Q., Carmon, Z.,Wertenbroch, K., Crum, A., Frank, D., Goldstein, W., et al. (2018). Consumer choice and autonomy in the age of artificial intelligence and big data. Customer Needs and Solutions, 5(1–2), 28–37.

Anderson, E., & Sullivan, M. (1993). The Antecedents and Consequences Of Customer Satisfaction for Firms. Marketing Science, 12(2), 125-143.

Barboza, G., & Roth, K. (2009). Understanding customers' revealed satisfaction preferences: An order probit model for credit unions. Journal of Financial Services Marketing, 13(4), 330-344

Carman, J. (1990). ‘Consumer perceptions of service quality: an assessment of the SERVQUAL dimensions’, Journal of Retailing, 65(1), 33-55.

Castelo, N., Bos, M., & Lehman, D. (2018). Consumer adoption of algorithms that blur the line between human and machine. Graduate School of Business: Columbia University Working Paper

Daqar, M. A. M. and Smoudy, A. K. (2019). The Role of Artificial Intelligence on Enhancing Customer Experience, International Review of Management and Marketing, 9(4), 22 -31.

Daqar, M. A. & Smoudy, A. A. (2019) The Role of Artificial Intelligence on Enhancing Customer Experience, International Review of Management and Marketing, 9(4), 22-31

Davenport, T., Guha, A., Grewal, D., & Bressgott, T. (2019). How artificial intelligence will change the future of marketing, Journal of the Academy of Marketing Science, 2(1), 1-19.

Denscombe, M. (2010) “The Good Research Guide for Small-Scale Social Research Projects” fourth edition, Butterworth-Heinemann

DMR (2020) 30 Taobao Statistics and Facts (2020), Retrieved from: https://expandedramblings.com/index.php/taobao-statistics/

Fleming, P. (2019). Robots and organization studies: Why robots might not want to steal your job. Organization Studies, 40(1), 23–38.

Foster, B. I. (2016) The Effect of Price and Service Quality on Customer Satisfaction in Mutiara Hotel Bandung, American Research Journal of Humanities and Social Sciences, 1 – 12.

Forbes (2020). What Is the Impact of Artificial Intelligence (AI) On Society?, Retrieved from: https://bernardmarr.com/default.asp?contentID=1828#:~:text=When%20AI%20takes%20over%20repetitive,increase%20happiness%20and%20job%20satisfaction

Jackson, S. L. (2011) Research Methods and Statistics: A Critical Approach, 4th edition, Cengage Learning, p.17

Jarek, K. & Mazurek, G. (2019) ‘Marketing and Artificial Intelligence’, Central European Business Review, 8(2), 46-57.

Khokhar, M. P. & Narang, C. (2019) ‘Evolution of Artificial Intelligence In Marketing, Comparison With Traditional Marketing’, Our Heritage, 67(5), 375-389.

Monette, D.R., Gullivan, T.J. & DeJong, C.R. (2010) “Applied Social Research: A Tool for the Human Resources” Cengage Learning

Neupane, R. (2015). The Effects of Brand Image on Customer Satisfaction and Loyalty Intention in Retail Super Market Chain UK, International Journal of Social Sciences and Management 2(1).

Noei, S. (2019). AI in New Retail in China: Not Just Unmanned Stores, Retrieved from: https://daxueconsulting.com/artificial-intelligence-new-retail-china/

Novikov, A.M. &Novikov, D.A. (2013). Research Methodology: From Philosophy of Science to Research Design. CRC Press

Oh, H. & Kim, K. (2017) "Customer satisfaction, service quality, and customer value: years 2000-2015 ", International Journal of Contemporary Hospitality Management, 29(1), 2-29.

Prentice, C., Dominique-Ferreira, S., & Wang, X. (2020). The impact of artificial intelligence and employee service quality on customer satisfaction and loyalty, Journal of Hospitality Marketing & Management 1(3), pp. 204 - 216

Polonsky, M.J. & Waller, D.S. (2011) “Designing and Managing a Research Project: A Business Student’s Guide” 2nd edition, SAGE

Rust, R. T. & Anthony, J. Z. (1993) ‘Customer Satisfaction, Customer Retention and Market Share’, Journal of Retailing, 69, Summer, pp.145-156

Saunders, M., Lewis, P. & Thornhill, A. (2012) “Research Methods for Business Students” 6th edition, Pearson Education Limited

Shumin, L. (2020). Taobao Gets 30,000 New Shop Owners a Day Amid Shift Online Due to Coronavirus, Retrieved from: https://www.yicaiglobal.com/news/taobao-gets-30000-new-shop-owners-a-day-amid-shift-online-due-to-coronavirus

Sin L.Y.M., Tse A.C.B., Yim F.H.K. (2005), CRM: conceptualization and scale development. European Journal of Marketing, 39(11/12), 1264-1290

Swift, R. (2001). Accelerating Customer Relationship using CRM and Relationship Technologies. New York: Prentice Hall Inc ISBN 0-13-088084-9.

Yang, Y. and Goh, B. (2019). EXCLUSIVE-China's SenseTime expects $750 mln 2019 revenue despite U.S. ban -sources, Retrieved from: https://www.nasdaq.com/articles/exclusive-chinas-sensetime-expects-%24750-mln-2019-revenue-despite-u.s.-ban-sources-2019-1

The Impact of Artificial Intelligence on Customer Satisfaction in China’s Online Retailing Industry, Case Study Taobao (Alibaba)