The Impact of the Chatbot (AliMe) on Customer Service in Alibaba

Professor’s Name

Name

Institution

Date of Submission

Acknowledgments

My sincere and individual gratitude goes to my tutor, who helped me put these simple ideas into something concrete. Also, I want to give thanks to this institution for giving me this excellent opportunity to complete such a beautiful project on the (Impact of Chatbot (AliMe) on Customer Service in Alibaba). These not only enlightens us about the importance of the Chatbot in Alibaba Company, but it also gives us insight into the importance of artificial intelligence in e-commerce and the internet of things.

Table of Contents

Contents

1 Introduction 4

• Queries/Answers Knowledge-Based 14
• IR Model 14
• Generation-Model 14
• Attentive Sequence to Sequence Re-rank Model 15

16

16

16

16

16

16

• Experiment 16
• 1 Research Aims and Objectives 18

2 Literature Review. 18

3 Methodology Planning 19

• Research Procedures, Design, and Reliability. 21
• 1 Project Plan 23
• Risk Assessment and Ethical Issues…………………………………………………………………………………………………24

4 Critique of Previous Research Project…………………………………………………………………………………………….27

5 Implementations of Methods and Analysis Outline 27

• Evaluation and Reflection of Results………………………………………………………………………………………………29

6.1 Analysis of Discussion 42

6.1.1 Recommendations and Conclusion. 44

7 Reference 45

The Impact of the Chatbot (AliMe) on Customer Service in Alibaba

1 Introduction 

Chatbots are robotic online chats engines that use big data and artificial intelligence to interact with humans effectively. The Chat robot system causing Alibaba’s customer service growth is Ali-Me. It is a sequence-based sequence and re-ranks to chat robot system accepted by ACL 2017.  Chatbots are nowadays becoming more common e-commerce tools; they include but not limited to Alibaba (AliMe), Amazon (Alex), and Apple (Siri). These are just but a package of thousands of applications developed by digital organizations.

Alibaba is one of the biggest online shopping platforms globally, encountering millions of transactions in a day. Such operations include but not limited to; suppliers’ products display, a pre-sale guide for shoppers, in sale consultations, and after-sale services, among others. Traditionally, consumers would wait for a long time for feedback from the staff. Apart from the transactions mentioned above, Alibaba also offers product promotion, which does not affect the price of commodities.

If this were conventional, the promotion would affect products’ price significantly, and other customers’ requirements. Nevertheless, Alibaba would recruit a vast number of customer service staff, but instead, they resolved through Chatbot AliMe. Thanks to AliMe developers. In 2014 the hotline from consumers jammed, Alibaba could take up to 14 days to handle customers’ questions (Yikun Guo, 2017). It was hectic for Alibaba’s customer service delivery team.

Besides, the experience could lead to a confusing purchase process that led to lower customer satisfaction. Thus, there is a critical question of whether it is capable of creating a Chatbot to handle these challenges on customer service? Such a Chatbot would effectively improve the customers’ experience and lowers the labor cost (Fenglin Li, etc., 2018). Precisely, the development of Chatbot (AliMe) was timely and significantly valuable to Alibaba customers’ service improvement. AliMe is a set of robotic chats service, intelligently designed programs applied to serve in electronic-commerce companies under the Alibaba group.

The project research aims to see and express the impact of (AliMe) Chatbot as a tool on customer service in Alibaba ltd. With a consistent and sustainable advancement of artificial intelligence on technology, the use of e-commerce chat robots (Chatbot) is one of the most modern technological developments in China. Chatbot, such as AliMe, can effectively solve several problems related to social service provision.

For instance, AliMe Chatbot provides a 24/7 service to Alibaba’s customers by offering online chats; when a customer is on the Alibaba site, AliMe will recognize it and ask questions, give answers and allow a display of items. AliMe can quickly and intelligently understand what the customer may be looking for and filter such items very fast for the customer to see them with ease.

Concerning this discussion, the main research question for this project is how the Chatbot, like AliMe, affects customer service in Alibaba. AliMe plays a crucial role in customer service delivery in Alibaba; these roles are, for example. In 2017, the Singles’ Day shopping festivals on 11 ^th November AliMe responded to more than nine million questions to customers. The number covers 95% of customers’ services of electronic commerce platforms of Alibaba (Alibaba Tech, 2018.).Moreover, AliMe offers services to millions of users every day, according to Alibaba Tech, 2018.

AliMe not only allows Alibaba’s customers’ service system to explore perpetually, but it also satisfies the client’s needs in real-time. In summary, AliMe, as a Chatbot, is quite essential in Alibaba’s customer service delivery.

This research paper applies the use of qualitative methodology, which is an interview in this case, for collecting data and gathering of information. This type of research requires information obtained from people who use AliMe. For instance, students of Sheffield University from China and other regions. This method uses open-ended questions. The composition of the interview involved some suggestions for AliMe’s services, among others. These made the methodology most apt in this research project.

In this methodology, participants were not only given a chance to elaborate on their viewpoint but also they were able to divide information with researchers in their words and perspectives without any limitations.

Consequently, this methodology does not come without drawbacks. The methods as a qualitative interview depend on the respondent’s ability to give data honestly and accurately. Some interviewees were not available; nevertheless, the factor of social distance due to covid-19 posed a significant challenge to the process.

The Ali-Me system structure appears in several layers; the input layer supports voices as well as text-messages in multiple ends. Such ends are the mobile phone, pads, PC, among others. The other layer sketches the deliberate model that decides the path of each query, for instance, assistance product or customer service.

The third layer demonstrates the composition applied to process queries, and fourthly is the covering for the knowledge source that is question-answer and knowledge graph, from which results come. The process for customer queries is in flow figure.1 below.

Figure 1.Standard IHCI flow.

INPUT                                                     TEXT                                                  CONTEXT

Sematic reps

OUTPUT                                                         TEXT                                          CONTEXT

The user objective in AliMe is in three classes; first is to ask for assistance, for example, “I need a flight ticket.” Second, is to request for information, answers, or solutions, for instance; how to retrieve my password? The third is to chat, i.e., “I feel unhappy.” Then each class further is fine-tuned as per the support-biz scenarios, i.e., Assistance flight booking services, flower booking, and mobile credit recharging. Both contents that are; business parser rule and intention categorizer work hand in hand to capture the intension of every customer query. AliMe business rule is in hundreds of thousands of patterns retrieved from the recurrent set of items, mining, and trial based patterns matcher. The intention categorizer is fit on CNN.

Alibaba’s electronic-commerce bot (AliMe) serves a zillion of customers’ questions in a single day. Most of the items are in the Chinese language, but not limited to English. The queries are also highly conversational, calling for an open domain bot engine that offers an excellent user experience. Chatbots that are open-domain related, as AliMe applies the use of retrieved information (IR) or generation models. These models also have challenges, IR retrieve answers from questions, or they answer knowledge-based, and generation models gives feedback on a return answers based on the pre-trained sequence to sequence (seq to seq) models.

IR-model may fail in handling long-tail queries, which do not align the QA knowledge-based, while generation models do not always give feedback consistently.

AliMe integrates a blended approach as an open domain Chatbot, depending on IR and seq2seq generation models. AliMe applies sequence-sequence based re-rank models to maximize both feedbacks outstanding the IR and generation based Chatbots. When it encounters a query, the Chat robot asks the IR-model to dig out a series of QA in double for possible answers. Then, it ranks the candidate feedback again, using attentive seq to seq models. The highest scoring candidate answer gets the point above the set margin, then that becomes the feedback. In case this becomes inconclusive, then the solution is chosen via a generation-based model.

For queries and feedbacks of various lengths, the technical team from Alibaba used the bucket mechanism to fasten the training process of the Chatbot. The application of softmax used to set vocabulary samples instead of an entire set. The strategy for crucial sampling drives these. In the feedback (decoding) phase beam search was used to maintain top candidate-k (k=10) results in sequences at each time rather than greedy search. These make generation becomes more consistent and reasonable.

Having already found proof of AliMe success at the experimentation level, then it could be tested against standard Chatbot. Applying a selected group of reliable testing queries, AliMe Chatbot expressed an excellent result on 37.7% of the questions, and a worst-case scenario on 18.9%.

Among the most significant digital business platforms globally is Alibaba, with billions of activities happening every single day. These transactions include but not limited to customers shopping online. Suppliers offer products to display, pre-sale shopping guidelines, sale consultations, and after-sales services, among many more. These would not be successful without the applications of tools like AliMe as a Chatbot in Alibaba. These ensure that customers are satisfied immediately and in real-time, unlike in traditional transaction platforms where people had to queue waiting for solutions to their needs.

The popularity of e-commerce or online marketing, currently known as digital marketing, is nowadays changing traditional consumer behaviors of the Chinese population. The advantage of digital shopping platforms like Alibaba is that they are convenient and offer lower prices on commodities as they eliminate intermediaries from the supply chain. Young people or middle-aged populations embrace online shopping as well, and according to the Chinese electronic commerce research Centre, even the older people have embraced it.

CN monitoring data showed that the Chinese clothing industry grew in digital shopping to reach 943.3 billion Chinese Yuan by 2016. A growth of 25% per year while the market entrance rose to 37%. Being the most significant electronic commerce websites globally, Alibaba’s consumers have demonstrated preferences on almost all items, what kind of store meets their satisfaction, consumer clothing, online experience, among others.

This article shows customers’ experience and satisfaction level due to the application of AliMe in Alibaba. AliMe’s objective is to improve customer satisfaction experience and for marketing plans in the future of Alibaba’s digital industry.

In the recent past, the online marketing industry has turned out to be more successful through the development of commerce operation theory. Many researchers found that customer satisfaction is critical to the growth of trade. This research mainly deals with one-dimensional digital marketing, i.e., the influence of Chatbot (AliMe) on Alibaba consumers and its importance to their customer satisfaction.

The methodology is an interview process, the customers’ satisfaction information of Alibaba caused by AliMe Chatbot. The importance of this research is to evaluate the impact of AliMe on customers who use Alibaba online shopping. The purpose of this evaluation is that it can guide the sellers on how to improve the operational activities and help the Alibaba industry improve consumer loyalty as well as improve the overall image of the store. Besides, it will help reduce the cost of publicity for the sellers to increase profit.

Conducting the impact of AliMe on customer service in Alibaba research assists the seller understands the consumers’ evaluation of their store to succeed in the online competition; thus, an excellent theoretical and significant research analysis.

According to the findings on customer service and consumer satisfaction both locally and internationally, this paper responds to the literature data in-conjunction with the current situation on customers’ satisfaction in Alibaba using reliable information from the Alibaba industry. Consumer service index data comes via an interview interaction process, sample data collation, and analysis. The significant aspects affecting customer service in Alibaba corresponds to recommendations and measures that led to the development of Chatbot AliMe.

Alibaba, on a broad scope, e-commerce solution happens with similar leading cloud technology supporting double 11 as well as utilizing scenarios of advanced digital commerce. Such techniques include live streaming, cross border, and big data analysis. An Alibaba artificial intelligence tool (AliMe) allows online businesses to be outstanding over their competitors.

Alibaba Chatbot (Ali-Me), a data cloud technology using robotic-human online chats, has made e-commerce solutions for retail services businesses affected by the pandemic and many others at zero cost for a long time now. These solutions include almost everything the retail industry may require to connect with staff and customers.

These also make customers and businesses operate efficiently and with a high level of online security. Ali-Me harnesses services to the fast-launch of an e-commerce operation in just five days. It is also capable of identifying the shopping habits of customers and gives recommendations intelligently. Ali-Me is capable of driving growth by conveying personalized and hyper-targeted messages to clients. It enables customers to quickly upload or download their pictures, videos, and files at a lower cost. Nevertheless, it is possible to send a batch of emails fasts and efficiently without building your mail server.

Chatbots are robotic online chats engines that use big data and artificial intelligence to interact with humans effectively. The Chat robot system causing Alibaba’s customer service growth is Ali-Me. It is a sequence-based sequence and re-ranks to chat robot system accepted by ACL 2017.  Chatbots are nowadays becoming more common e-commerce tools; they include but not limited to Alibaba (AliMe), Amazon (Alex), and Apple (Siri). These are just but a package of thousands of applications developed by digital organizations.

In contrast, the previous more rudimental versions where customers had to adhere to simple and complex structured languages. Current Chatbots make users apply original text and speech and, to some extent, images when interacting with such. In our research, Ali-Me may not have attained the levels of Samantha in the movie “Her,” but it is making a significant move to the realization of that level of AI and AI interactions.

Alibaba, as an e-commerce site, has experienced tremendous benefits to customer service delivery since the inception of Ali-Me as a Chatbot. The company is now able to offer a 24/7 automated support to customers in terms of customer service. Besides that, AliMe is also capable of analyzing previous data with its forecast functionality to understand the customers’ demand and give future results.

It has the highest scalability of peak demands. In case a retailer offers a discount for products, it automatically reminds the consumers and tells the customer about it at the same time. Besides, Ali-Me provides free statics and analyzed data to every seller on Alibaba platform. Alibaba research showed that more than six hundred thousand merchants have so far used AliMe. (Fenglin Li, et al., 2018.) Says that AliMe is an essential tool in Alibaba towards customer service delivery.

• Understanding AliMe as a Program.

AliMe is a sequence of Chatbot services, applied for online trade (e-commerce) under the Alibaba Group of companies. AliMe falls in the artificial intelligence service section in Alibaba.

A complete work of creating AliMe Chatbot, took several approaches as discussed here below:

• A sequence – sequence-based re-rank procedure

All these starts with the making of a QA (Question – Answer model), which is knowledge-based from the chat logs of the Alibaba customers’ service Centre. Using this knowledge-based model, AliMe team after that developed three other models:

• IR model
• Generation based model and
• Re-rank model.

It is worth noting that all the models are text-based; that is, segmentation of words is needed. IR input, are text features while generation model input characters and re-rank model are text embedded, this means that they pre-learned from the fast version (Bojanowski et al., 2016) then filtered within both models.

The releasing model, as well as the re-rank model for AliMe Chatbot, founded on the same sequence-sequence approach. Generation models release the output, while the re-rank model ranks candidates’ answers concerning input question. Presented with a query q, and threshold T, then the procedure would be:

• Using the IR model to reflect a set of candidate k, double QA (qkbi, ri) k i=1 (k=10).
• Next, q pairs with all candidates’ answer r~i and find a middle score o(r~i) = s (q, r~i) in all couples applying function equation two of the re-rank model.
• Finally, in consideration of answer r with a top score o(r) = max o(r~i): if o(r) ≥ T, taking r; output a reply on generation model. In summary, T results via experiential learning.

· Queries/Answers Knowledge Based

The construction of QA pairs comes from the dialogues of customers and staffs between a period of 1^st January 2016 and 6^th January 2016 as the original data .thereafter; each question paired with an adjacent answer. Then sorted out QA with business keywords to get the pairs.

· IR Model

As a retrieval mode, IR applied search methodologies to acquire the same queries of each input to get the paired answer. These Segments were resulting from an inverted index and then identifying each version to a set of such keywords from all sets of questions.BM25 formulae were then applied to work out the similarities of input and retrieve items. The combined feedback from the most similar one becomes the answer.

· Generation-Model

This part of AliMe is an attentive sequence to the sequencing process (Bahdanau et al., 2015).Assuming that Ɵ~i = {y1, y2… y~i– 1 c~i}, as the probability of text generation y~i present, then i=s~i-1 in a linear model determining scores of input present and ~i aligning the feedback in position ~i-1 (Bahdanau et al., 2015).

There were several interventions recommended as discussed below;

• Bucketing and padding. This approach deals with QA of varying lengths and answers to a specific symbol.
• Softmax over sampled texts. The approach that helps in speeding up learning (training) process by application of words that have been tested instead of a full set (jean et al., 2014).
• Beam search decoder. The beam search approach maintains top-k (k = 10) output sequences in every time instead of inconsiderate search that retains a single search at a time t. These improve generation mode and make it more reasonable.

· Attentive Sequence to Sequence Re-rank Model

Uniquely, the team in AliMe selected low probability, depicted as s^mean-prob.as a scoring function, and candidate feedback is a text sequence (w~1, w~2 … w~n). The inverse of average and mean expressed poor results.

Figure 2.below displays AliMe’s chatting flow module.

Figure 3.AliMe-Assist flow model,

· Experiment

Following a promotion activity of L’Oreal France, on 18^th June 2018, AliMe was applied by the T-mall store. The findings displayed that AliMe Chatbot covered 86% of consultations, differing, the previous month 49%.from the analysis, the agent of the brand expressed the benefits of AliMe that it helped their consumers reduce the waiting time primarily. The resolution rating accounted for 65%of the total consultations.

AliMe effectively improved customers’ satisfaction through the effective answering of questions and giving results to customers.

Summarily, to increase the consumer satisfaction of AliMe in Alibaba at a higher level, the Chatbot team can continually gather more data and knowledge from customers concerning the product. More use of the products and promotions will attract more data and therefore add to the knowledge base of AliMe.

• Illustration

The launch of AliMe Chatbot to the real-life industrial intelligence assistance has brought about service to millions of consumers. They are providing answers to millions of queries each day, addressing 85% of them.

The critical features demonstrated through AliMe-Assist include the following;

• Assistance services
• Users/customer services
• Chatting services

Example of customer service AliMe-Assistance chat.

1.1 Research Aims and Objectives

The main aim of this survey is to investigate the impact of Chatbot, such as AliMe, on users’ perceptions of customer service in Alibaba. The project tries to answer the question, how Chatbot, such as AliMe, does affect customer satisfaction in Alibaba.

Conducting the impact of Chatbot AliMe on customer service in Alibaba research assists the seller understands the consumers’ evaluation of their store to succeed in the online competition; thus, an excellent theoretical and significant research analysis.

The research was conducted to meet the following four objectives as outlined below;

• You are describing Chatbot functionality and operations in customer service delivery to e-commerce like Alibaba.
• To identify how customers apply the use of Chatbot, such as Ali-Me.
• To learn how Chatbot, such as Ali-Me, influences user’s satisfaction, and finally.
• To determine the pros and cons of customers’ service for Alibaba by applying this robot chats.

2 Literature Review.

Since the mid-nineties, Tsinghua University introduced the aspect of consumer satisfaction because of customer service delivery in the local market. It is after this that many academicians actively started to construct Chinese customer service models to suit their national market conditions (Hou, 2014). According to (Wang C.2003), customer satisfaction is a critical determinant of customers’ loyalty, and customer service is a necessary condition for their commitment.

Liang Yan built the first consumer satisfaction index model in 2007, including value-perception, company image, quality perception, quality service delivery perception, customer relationship, and customer loyalty, and eight structural variables.

Consumer happiness and satisfaction are a beneficial measure of customer recognition to the business, the products and services, and thus repurchasing tendency. From a digital marketing viewpoint, the main task of emerging companies is to attract more customers, but more importantly, to sustains those customers when the business matures. He noted that losing a loyal client is synonymous with losing eight prospectuses; thus, customer satisfaction equals customer loyalty.

3 Methodology Planning

Even though the real field survey did not take place due to the current crisis of COVID-19, the original research methodology plan recommended the use of a qualitative method, that is, an interview. The intention here was to apply the use of intensive discussions that is an optimal way of collecting data from individuals’ personal experiences, histories, perceptions on this sensitive topic ‘impact of AliMe on customer service in Alibaba. The composition of the discussion designed set for different people to make them enjoy the process, as they express the benefit of AliMe Chatbot in Alibaba. Besides, the interview was semi-structured, in that some queries were predetermined for all the interviewees. Similarly, several questions were asked during the process to make it clear and explore other queries (Creswell, J.W., 2013).

• Research Materials

The following were the qualitative and in-depth interview materials as planned to complete the project. The methodology required the prior preparation of the following resources as materials for the survey.

• Letter of invitation
• The information sheet for the participants
• A consent paper (form)
• Interview plan(schedules)
• Three Interview renaissance guide or topic guide for the different participants’ groups(data set)
• An extra sheet for any other relevant information

• Approach

Bearing in mind the danger of the crisis caused by the covid-19 pandemic, the interview was done and implemented online via platforms as We-chat. These showed that online meetings could save more time and reduce human contact than one on one interview. It also made the process more effective. However, the online discussion comes with their limitations; for instance, the space of conversation cannot be determined or controlled.

The interviewer is also required to confirm the platform they will use, be it a computer internet or other gadgets. These became some of the restricting factors to the interviewees (Creswell, J.W., 2013). The research target people were Chinese since most of them use Alibaba. It was easier to obtain some answers that I needed to collect, for the research. Any time consumers purchased from Alibaba, they have to consult the customer service system. Therefore, every consumer has used AliMe.

The methodology also included people from varying ages and careers; these are; children, senior citizens, or students, among others. The plan was to interview twenty people of varying ages, gender, and career. The collection method was simple random sampling because the number of persons in We-chat was static. These were the requirements that the interviewer needed to select 20 interviewees to research random sampling from the We-chat.

· Research Procedures, Design, and Reliability.

These are the blueprint for the data analysis, geared towards answering specific questions or testing specific research hypothesis; in this case, ‘how does AliMe Chatbot pose impacts on customer service in Alibaba?’ The design included the following three processes:

• Sampling: in the survey, the data was to be selected randomly from the participants based on age, sex, income status, designation, shopping methods, amongst others. The sample is taken from the Chinese population because this is the origin of Alibaba. Still, most of the participants targeted would be the working class, young and middle-aged persons, as well as university students. This sample is believed to be using Alibaba online shop for the purchase of their items hence reliable and valid.
• Data collection: this process is planned to entail the following steps
• Interview with audiotape, videotape, and transcripts.
• Direct non-participant observation from the researcher,
• Secondary data from previous field notes, internet materials, and journals supporting the topic in question.

The interview procedure was conducted online we-chat, and there were no face-face researcher-participant allowed because of covid-19.the information gathered is supposed to be stored in the form of recorded conversations and written transcripts in the tools mentioned above. These would, therefore, support the validity and reliability of this investigation as well as the researcher’s experience in the subject. Besides, active participation in the field during the whole process supported the authenticity of this data as laid down in the process.

Nevertheless, the validity and reliability is supported by the following other factors;

• Presence of an open context,
• Researcher’s non-preconceptions,
• Researchers focus on awareness and energy on the topic.
• Total concentration and absolute determination of this topic,

The comparison of this data and the several previous types of research on a similar topic is also supportive of the current data analysis.

• Process for scale establishment (data analysis): text data analytic method was suitable for our research, analyzing the level of satisfaction and experience after using AliMe Chatbot in Alibaba. The process entailed the following steps;
• Leaving with the data. Taking the data to the designated place for analysis.
• We are classifying and grouping the data for further interpretation.
• Testing the concepts and themes
• Identification and description of relationships between the concepts,

3.1 The Project Plan

The table below is an illustration of the plan for this research paper indicating the number of tasks involved to complete the project. The schedule runs between May and August 2020, and the time duration for each job is determined through beta probability distribution. It is equally distributed to each task based on the length of the activities to be taken from the start to the completion. The events are arranged in order of priorities bearing in mind that some tasks were to be completed before the others while others were depending on the information from others. Therefore, some jobs were preceding others, while others succeeded in others.

• Risk Assessment and Ethical Issues

It is worth noting that checking of risks is a necessary factor in modern research and current society (Beck, 1992, 1994; Giddens, 1991, 1993, 1994). Which has turned to be a “risk society” traditionally; the examination of risks has been limited to the only assessment of risks to participants.

Currently, both parties’ risks need to be assessed, and migration measures put in place before the project kick starts. The following risks are laid down and evaluated according to their intensity of danger.

The risk determining factors level was low, and therefore the methodology was suitable for this research. All factors placed constant; the study was worth proceedings.

Any factor that was predetermined to pose harm to both parties was mitigated, and necessary measures were put in place, these include;

• Avoidance of face-face meetings to maintain social distance due to covid-19, the interview was predetermined to be conducted online.
• The researchers were encouraged to use secondary data instead of primary data.
• Prior sharing of the risk assessment form with participants,
• Training
• Self-awareness
• Considering risks
• Inform the third part about your research and all involvements.
• Ensure all materials required for research are ready and available with you in the prior period.

• Ethical Issues.

The possible advantages of research need to be looked into both participants and researchers, community, or society on a broader manner. The reason for this is to examine the risk factors that might occur during the process. The potential risks must be weighed against the potential advantages of the research. The benefits need to be looked into and articulated to all parties hoped-for results instead of one guaranteed for fidelity for any uncertainty in many types of research. Incase harm to participants is higher than minimal, and no evident benefits or little then no justification of carrying on the project. Participants are suitable for making that conclusion.

• Critique of Previous Research Project

One of the main limitations of the article is that it failed to address the risk factors assessment touching the researchers, rather concentrated on the participants. The danger posed by such failure is that in case of a harm risk factor affecting the researcher affected the quality of data collected or causing incompletion of the project. These factors hence are critically addressed by this current research project to mitigate the future occurrence of such.

We used secondary information currently accessible, organized text data, records that were collected from various references to set new conversations as well finalize on the survey, for trade clusters after the qualitative and substantial examination (Mayring, 2000). The qualitative and actual analysis is the most suitable strategy for new research subjects like Chatbot (AliMe) business models. The fact by which Alibaba set such a vast online-based shopping industry is the basis of our motivation.

Reason being that Alibaba has several challenges and factors affecting the flow of research at this point and related analysis (Tan, Pan, Lu and Huang,2009; Wei, Zhu, and Lin,2013).Alibaba is the most preferred B2B digital marketing site this aimed at expanding its market and service globally. Researchers, therefore, are required to analyze the reliability of such issues in research. Validity and dependability are the key features in quality research findings.

It is the bridging point for testing and evaluating qualitative research (Golafshani, 2003). Only a few investigators argued that qualitative analysis would apply both aspects while designing a test, dividing findings, and checking the success of the research methodology, such as its reliability, validity, capacity as well as the authenticity of its data.

In summary, these findings are the qualitative online interview with reviewed private research information. The research starts with a plan and a vision, which looks into the impact of Chatbot on customer satisfaction in online businesses like Alibaba. The results are authentic written work, journals, and articles. The strategy for action and application of the testing process is shared hereunder. The sources of information and data of our research, may it be contextual case analysis or qualitative data, was from authentic internet sites, articles, and journals.

5 Implementations of Methods and Analysis Outline

These aims at reporting the findings and observations arrived at each level. Description of customer satisfaction, customer experience, and market segments is provided with an adjacent discussion of the data, market structures, and respective graphics. This is a presentation of results from persons who do digital shopping and marketing experiences on the Alibaba website and have interacted with AliMe as a Chat assistant. In total, twenty questionnaires were randomly tabled via the online We-chat platform. Fifteen out of them were, effectively, and precisely received with 95% active recovery. This data was collected, accepted, and stored as notes, recorded audios, and videos and would remain valid for reference in any case for more than six months and less than a year.

• Evaluation and Reflection of Results.

Q1. Sex (digital shopping users of Alibaba with AliMe experience, structure, and analysis)

• (Figure1.self inserted)

It is noted from the findings that; among the persons in Alibaba with Chatbot experience were both male 48.5 % and female 47.5%, a 1:1 ratio between male to female. These results showed that both women and men love online shopping and digital marketing in Alibaba. The data set also revealed that they could enquire for services via AliMe-Assist Chatbot, just from either a mobile phone, PC, or any other internet device. They could quickly pay for services through Chinese digital payment systems. Men who do not prefer purchasing items are now changing that trend.

For the customer, convenience and price are a fundamental principle. Most of the people who had used this platform said it is convenient, and commodities prices were low. This means that the Chatbot (AliMe) had significantly improved customer service in Alibaba, reduced labor costs, and other expenses, thus lowered the price of goods and services. This made any consumers loyal to the company. It saved their time, energy, and they obtained a discount.

Q2. (Digital shopping users of Alibaba with AliMe experience, structure, and analysis)

• (Figure2.self inserted)

Of all the respondents, most of the Alibaba users are aged between 26 years and 30 years, accounting for 38 %, followed by age bracket 31-40 years by 28%, then 18-25 years old accounting to 22%, 41-50 years 7% and finally under 18 years 5%. It is worth noting that, Alibaba users are between ages 18-40 years but majorly middle-aged population.

This depicts that Chatbot (AliMe) significantly affected most of the young and middle-aged people. Thus, Alibaba should improve toward targeting young people. This analysis is not only important to see who are the users of Alibaba but also to express the population left out by the digital age. Therefore, Alibaba should also design strategies to reach the aged people on the use of online shopping, marketing, and other services.

Most young people said that AliMe has simplified the work of online shopping, saved their time, and the experience was terrific. They said they now more loyal to Alibaba than ever before. Most of them also were B2B users; this Chatbot has promoted their business.

Q3.Title (Digital shopping users of Alibaba with AliMe experience, structure, and analysis)

1. (Figure 3.self inserted)

From the findings, for digital shopping, employees accounted for 43%, followed by business people 21%, senior managers 18%. These three categories summed up to 80% of the total Alibaba users. This displays that the nature of occupation has a significant influence on digital opportunities. Since most of the employees and senior staff are busy most of the time, the development of Chatbot as AliMe meant a lot to them concerning online shopping in Alibaba. Most of them were impressed by the easiness of Alibaba services caused by AliMe-Assistant chat. Students are busy with academic work; they are not earning, thus accounted only for 5%. Government institutions could be hindered by regulations and procedures, and therefore, accounting only 7% .alibaba should target more of the working class and business class. Some may not have learned the presence of AliMe calling for a more online campaign promoting the Chatbot.

Q4.Monthly earnings. (Digital shopping users of Alibaba with AliMe experience, structure, and analysis).

1. (Figure 4.self inserted)

During the survey, we found that monthly income was also a factor that influenced online shoppers in Alibaba.48percentage of all the interviewees earned CNY 5000 and above, 23% of them received an income between 3000 and 5000 Chinese Yuan. More so, 17% made between 1500 and 3000 Yuan while 12% earned 1500 Yuan and below. This shows that the majority of persons earn high monthly revenue, thus able to purchase goods and services online; they can afford internet and smart gadgets like smartphones and computers.

Using this information, the respondents also expressed their interests on Alibaba and said AliMe Chatbot impressed them. Most of them said they buy almost 90% of their products on Alibaba now that it has become more convenient and users friendly due to the Chatbot.

Q5.which online website do you know?

1. (Figure 5.self inserted)

49% of the respondents said they knew Alibaba, and 44% said they knew Jingdong while only 7% of the respondent knew other websites. From the findings, we can say that most of the people know Alibaba and have experienced shopping or seeking their services. Nevertheless, not forgetting its close competitor Jingdong, this calls for Alibaba to do more campaigns and promotions to ensure they are well known in China and globally. Even if they have the best customer service globally and Chatbots like AliMe, it can only be fully active when many people who need such services have learned about it.

Alibaba will have to do many campaigns to reach many people in China and make known of their services and existence. They have to gear up their market strategy to improve their visibility.

Q6.how did you learn about Alibaba?

• (Figure 6.self inserted)

About 48% of the population from the research knew Alibaba through online advertisements. And 40% of them learned about it through television advertisements. Only 5% of the interviewees learned about Alibaba through friends’ social networks, and 2-3% through other channels such as magazines. This means that Alibaba’s advertisement medium and strategy were impressive but could consider putting more energy in different ways.e.g.focusing on social relationship marketing, customer loyalty, good reputation, and many more. Make more knowledge and understand who Alibaba is, and it features anytime, anywhere.

By utilizing this information, Alibaba’s key features like AliMe will get more introduced to the people, and they will learn its benefits. Thus, increase the Alibaba customer base. Customer service and customer satisfaction is key to any business, but more so, letting people know how you serve them is more important. Alibaba can tap social media, media, and other platforms to advertise more about its new services and features.

Q7.Have you used Chatbot (AliMe) when shopping in Alibaba.

• (Figure 7.selt inserted)

From the data analysis, 57% of the respondents often use AliMe as chat assistance or other services at any time they are shopping in Alibaba. This shows that many people in China are happy to use the Chatbot; thus, many are preferring to use Alibaba to buy goods and services.34 % of the people used it occasionally, and only 5% never used it. This also means that more than 90% of the respondents use Alibaba services and have experienced with Alibaba features such as Chatbot (AliMe). It showed that many Chinese people had embraced Alibaba, they buy items on Alibaba, sell on Alibaba, and enjoy other services from Alibaba. Thus, Alibaba’s customer base increased drastically from the time Chatbot AliMe got installed. Showing the significant impact it had on customer service delivery in Alibaba.

The use of this Chatbot assistant resulted in Alibaba’s single-day sales turnaround; on 11th November 2018, a single day shopping bonanza by $30.8 billion total sales volume. This was 27%more than any other time there before. The figure was more than double that of the previous year black Friday or cyber Monday.

Q8. Reasons why you do not shop on Alibaba?

• (Figure 8.self inserted)

According to this survey data, it is clear that there are reasons that people do not use Alibaba services. Between 22 and 23% of respondents are worried about the quality of the products sold online and were concerned about the delivery time; it seems that delayed delivery or extended time between the order date and time of goods arrival. Alibaba can use AliMe to alert and keep the customers informed on when they will receive their products, how long it will take, and the reason for any delay.

It would be wise to keep customers alert in case any change occurs. Besides, the means of delivery can be improved. At the same time, AliMe could be used to build customers’ confidence and to assure them that the security of their money and product is paramount. Let them know that now most of the services, activities are automated and are high tech for interference.

This is because 16% of the respondents were worried about internet security. Still, 12% of the people were concerned about the price, and they said that the price was not low enough in comparison to other websites like Alibaba. Between 8% and 10% of the respondents perceive that online shopping is a complicated process, and to get the service done would take a lot of time. Therefore, Alibaba should work more on its; product strategies, price strategies, and consumer service strategies. Doing this will attract more potential customers and improve potential revenue as well as customer satisfaction.

Having developed a Chatbot like AliMe, Alibaba is set to improve on these areas smoothly as it is designed to enhance such services as customer service, customer satisfaction, communication through chat services, service booking, among others. Thus, AliMe is the critical tool for improving customer satisfaction in Alibaba and, as a result, tapping on the potential customers and revenue above the competitors.

Q9.if these aspects were improved, would you consider using Alibaba services?

• (Figure 9.self inserted)

The findings show that 95 % of the respondents were ready to reconsider using Alibaba services once the improvement completed. This means that Alibaba has some challenges that needed to be improved further. AliMe Chatbot was the solution to these problems; working on internet safety, product quality, efficiency, and attitude was key to the company turnaround.

After-sale, services through AliMe would bring about a turnaround on the brand reputation. AliMe is a tool for Alibaba to make customers remain assured that shopping online and will bring a turnaround in performance. Alibaba sales will grow; market share will also increase as well as customer satisfaction.

Q10.When shopping online, what factors do you consider?

1. (Figure10.self inserted)

From the findings, when respondents are shopping online, they consider several factors.26 % look at the brand reputation, 24% like quick delivery of products. Also, 16% of them are concerned about transaction security and will only use a site that they feel well secured.  13 % of the respondents consider the fair price of products as critical to them, and websites with favorable prices are attractive to them. 9 % of the people preferred a simple process and one that is easy to use.

Customers’ experience carried 7% of the respondents. Good customer experience, as we have seen with Chatbots like AliMe, is key to attracting more business in Alibaba.5% of the customers, were concerned about product variety and product display, where they can be selected. It is, therefore, worth noting that different consumers consider varying factors. Such factors influence online transactions.

Q11.main reason for choosing Alibaba.

1. (Figure11.self inserted).

From data analysis and findings, it is clear that 27% of the respondents choose Alibaba’s services because of the pleasing quality of products as well as the attractive costs of goods and services. Of them, 26 % use Alibaba because it is convenient and efficient when doing any of the transaction they offer and thus, saves them time and energy. Still, 16% of people prefer to pay cash when goods are delivered to them.

It is making them feel more secure about their money. About 14% of the respondents liked the Alibaba website design and features. It is showing that website display is critical in attracting more customers. The reason behind this is that an attractive website display or appearance pleases the eyes of the user; it expresses the business art of being organized, innovative, and smart. Thus, it plays a role in product quality and product attractiveness. It seems that digital shopping is convenient for many users. This is because shopping on Alibaba saves time, it has a variety of products and does not run out of stock.at the same time, and it is a one-stop-shop where all goods and services are available. More so, deliveries are done house to house, thus the leading cause of online purchasing.

It is always crucial that Alibaba put these factors in mind to maintain the highest level of customer service and satisfaction.in doing so, Alibaba sales revenue will keep scaling high in china and globally even amid rampant digital competition. By the use of features like chat assistance by AliMe, Alibaba is good to go.

Q12.where do you think Alibaba needs to improve as a website?

• (Figure12.self inserted)

In a world of digital competition and revolution in the marketing industry, Alibaba has to retain its competitive benefit, needs to work more on marketing strategy and research for improvement and change. From the analysis of the findings, customers expressed varying points of view; 24% said Alibaba needed to further improve on promotion method. Those who expressed the need to improve after-sales services were 21%, consumers’ service needs accounting to 19%, website appearance improvement was 14%, and change on the information was transmitted accounting to 12%. The objective of any business is to improve its sales and performance.

For performance to grow, Alibaba has to attract more customers by improving on factors such as product price, customer services, and satisfaction, among others. Using AliMe Alibaba can quickly develop on these areas and create a healthy and conducive online shopping environment.

6.1 Analysis of Discussion

Alibaba is one of the globally famous digital marketing companies known in many years now. It is well known for its product diversification. In the recent past, Alibaba developed one of the key chat assistant called AliMe. Before this Chatbot, Alibaba’s customer service and operation were tedious, hectic, and confusing. Many customers were not happy and, thus, unsatisfied. These showed that business was going well with Alibaba. According to the research findings, a good number of factors such as,

• Quality,
• Website design
• Customer service
• After-sales services
• Information transmission
• Delivery
• The price of products and many other factors needed some degrees of improvement.

It is by these challenges that Alibaba got the idea of developing and using Chatbot (AliMe); this tool has solved almost all of these problems.

AliMe made Alibaba sell $30.8 billion in a single day, after its emancipation on 11^th November 2018. The highest sale in its history, this means that its customer base had grown significantly. It also means that the customers were happy and satisfied.as the global largest, online shopping website, Alibaba is now having above five hundred and fifty million active users. Sellers are in millions generating zillions of transactions via different, Alibaba promotional tools as well as activities per day.

Alibaba’s environment involves millions of online operations merchandise. Almost all of Alibaba vendors, are 3^rd party logistics vendors through Cainiao network logistics. A million of offline mom-pop shops through Alibaba retail and wholesale business. Besides, Alibaba has eight hundred and seventy million customers finance consumers through Alipay of Ant Financial.

Alibaba relied traditionally on such inside customer services to support purchasing activities and other related after-sales activities.

The intensity and increase of Alibaba groups’ business growth lead to more problems for consumer service staff. For instance, since the 1^st singles’ day purchasing bonanza in 2009 until 2017, gross activities volume increased from 50 million to more than 168 billion transactions. This number of sales and sellers kept on growing and has increased from 27 to 180,000 retailers. That means a high increase in customer service demand, and the consumer service team cannot tackle it alone. Since Alibaba has built the world’s complex consumer service systems, artificial intelligent consumer service is the only long-term solution for the future of business. The development of this AI Chatbot (AliMe) was the solution behind this growth and success.

It is evident that two years after its launch in 2015, during a 2017 singles’ day purchasing bonanza, AliMe served nine million customers questions. These were 95 % of the consumers’ service on Alibaba online platforms. AliMe further builds more than 56 billion various personal recommendations, purchasing lists for consumers.in 2018, AliMe has empowered the customer service team, increased pro-active advertisement and promotion, reminders to customers to finish a transaction.

AliMe can generate invoices, change address, alerting deliveries, and appointing logistic vendors. As well, AliMe can make a purchase transaction information dashboard for retailers and offer intelligent purchase assistant. For example, AliMe can learn a consumer’s psychology or emotion and prioritize signal staffs from the customer service section to intervene. It is capable of alerting retailers to boost their buffer stocks to meet the rising demand.

6.1.1 Recommendations and Conclusion.

Alibaba can expand the Chatbot (AliMe) customer service robot beyond its online platform. Because Alibaba has also invested in social media like (Weibo) and video website (Youku), payment tool (Alipay), and many other millions of offline sellers and consumers, Alibaba can connect all these with the full consumer journey.

7 Reference

1. https://www.marketscale.com/industries/retail/an-inside-look-at-how-alibaba-uses-technology-to-improve-customer-experience/
2. https://www.customerthink.com/alibaba-customer-experience-a-study-in-how-not-to-do-it/
3. Tan, B., Pan, S. L., Lu, X., & Huang, L. (2009).Leveraging digital business ecosystems for enterprise agility: The tri-logic development strategy of Alibaba. Com.
4. com (2014). Alibaba Group. Retrieved 23rd October 2014, from http://www.alibabagroup.com/en/ir/financial
5. Qing, H. (2008, December). A Model for Value-added e-Market Provisioning: Case Study from Alibaba. Com. in Future Generation Communication and Networking Symposia, 2008.
6. Second International Conference on (Vol. 1, pp. 47-52)
7. Osawa, J. (2014). How Does Alibaba Make Money?
8. The Wall Street Journal Retrieved 23rd October 2014, from http://blogs.wsj.com/digits/ 2013/09/09/ how does-Alibaba-make-money/
9. WANG, G., & Lim, Y. T. (2011, August). Research in China’s Alibaba’s Development. In Artificial Intelligence, Management Science and Electronic Commerce (AIMSEC), 2011 2nd International Conference on (pp. 2295-2301).
10. com. (2014, September). Linked In. Retrieved 20th October 2014, from Business Development and Marketing Director: https://cn.linkedin.com/jobs2/view/17317067.
11. Doland, A. (2014, 24th September). Branding Lessons from China’s Alibaba. Retrieved 21st October 2014, from Adage: http://adage.com/article/global-news/Alibaba-teach-Chinese-brands/295115/
12. You Fei, J. (2012). “E-marketing strategic management at Alibaba.com” (pp. 1-12). De Monfort University. MBA Library.
13. Clasen, M., & Mueller, R. A. (2006). Success factors of agribusiness digital marketplaces. Electronic Markets, 16(4), 349-360.
14. McGovern, G. (2014). Customer-centric and easy-to-use is the new business model (The Alibaba story). New Thinking. Retrieved from http://www.gerrymcgovern.com/new-thinking/customer-centricand-easy-use-new-business-model-alibaba-story
15. Puglisi, D. (2014). Alibaba Group Holding Limited. (2014) (1st Ed.). Hong Kong. Retrieved from http://www.sec.gov/Archives/edgar/data/1577552/00 0119312514362772/d794345ds8.htm.
16. com. (2014). Alibaba Group. History and Milestones. Retrieved 23rd October 2014, from http://www.alibabagroup.com/en/about/history.
17. Leong, T. (2013). Introduction to Alibaba.com.Retrieved 16th October 2014, from Alibaba.com:http://www.alsma.org/wp-content/uploads/2013/10/
18. Timothy-Leung-Introduction-to-Alibaba.pdf
19. Sau-ling LAI, L. (2014). Chinese Entrepreneurship in the Internet Age: Lessons from Alibaba.com (1^st, pp. 1-7).
20. International Science Index: World Academy of Science, Engineering, and Technology. Retrieved from http://waset.org/publications/15135/
21. chinese-entrepreneurship-in-the-internet-agelessons-from-alibaba.com
22. Global Journal of Management and Business Research Volume VII Issue IX Version I Year ( ) E 2014 © 2014 Global 1 Journals Inc. (US)
23. The Review of Alibaba’s Online Business Marketing Strategies which Navigate them to Present Success
24. Batjargal, B. (2007). Internet entrepreneurship: social capital, human capital, and performance of internet ventures in China. Research Policy, 36(5), 605-618.
25. Essays, UK. (November 2013). Market Analysis and Strategy for Alibaba Marketing Essay. Retrieved from http://www.ukessays.com/essays/marketing/
26. market-analysis-and-strategy-for-Alibaba-marketing essay.php?cref=1
27. Yang, N. (2010, August). The strategy of search engine marketing for high-tech enterprises. In Advanced Computer Theory and Engineering (ICE), 2010 3rd International Conference on. Doi:10.1109/ICACTE.2010.5579626 (Vol. 5, pp. V5-

318). IEEE.

28. Mourdoukoutas, P. (2014). Alibaba is the Five Advantages. Forbes. Retrieved from http://www.forbes.com/sites/panosmourdoukoutas/2014/04/15/alibabas-five-advantages/
29. Essays, UK. (November 2013). The Alibaba Group Company Analysis Marketing Essay. Retrieved from http://www.ukessays.com/essays/marketing/thealibaba-group-company-analysis-marketingessay.php?cref=1
30. Schepp, D., & Schepp, B. (2009). The official Alibaba. Com success guide: insider tips and strategies for sourcing products from the world’s largest B2B marketplace. John Wiley & Son.
31. Shih, G. (2014). Alibaba affiliate Alipay rebranded Ant in new financial services push. Reuters/China Daily. Retrieved from http://www.reuters.com/article/2014/10/16/us-china-Alibaba idUSKCN0I50KJ20141016
32. Chen, L., Solomon, B., & Mac, R. (2014). Alibaba’s IPO Pop Makes Jack Ma The Richest Man In China. Forbes (Forbes #400). Retrieved from http://www.

forbes.com/sites/briansolomon/2014/09/19/alibaba’s-ipo-pop-makes-jack-ma-the-richest-man-in-china/

33. Doland, A. (2014, 24th September). Branding Lessons from China’s Alibaba. Retrieved 21st October 2014, from Adage: http://adage.com/article/global-news/alibaba-teach-chinese-brands/295115/
34. Gang-lan, H. U., Xiang-hua, L. U., & HUANG, L. H.(2009). Research on E-business Ecosystem and Its Coordination Mechanism——A Case Study of Alibaba Group [J]. Soft Science, 9, 003.
35. Walraven, P. (2009). A Brief History (and Future) of Alibaba.com – TechNode. TechNode. Retrieved 23rd October 2014, from http://technode.com/2009/01/22/a-brief-history-and-future-of-alibabacom/
36. Tisdell, C. (2009). Economic reform and openness In China: China’s development policies in the last 30 years. Economic Analysis and Policy, 39(2), 271.
37. Osawa, J. (2014). Softbank, Yahoo Support Alibaba’s Partnership Structure. The Wall Street Journal. Retrieved 23rd October 2014, from http://blogs.wsj.com/digits/2013/09/26/softbank-ceosupports-alibabas-partnership-structure/
38. Cheung, T. C. H. (2013, December). Investigation on the relationship between China’s social media and the purchases made on the e-Commerce platforms for better promotion strategies. In SIGBPS Workshop on Business Processes and Services (BPS’13) (p. 99).
39. (2014). Billionaire Jack Ma teaches you how to be successful in life and business – Vulcan Post. Vulcan Post. Retrieved 23rd October 2014, from https://vulcanpost.com/5407/billionaire-jack-teaches-you-how-to-be-successful-in-life-andbusiness/© 2014 Global Journals Inc. (US)
40. Global Journal of Management and Business Research Volume VII Issue IX Version I

Year ( ) E 2014 The Review of Alibaba’s Online Business Marketing Strategies which Navigate them to Present Success

41. Lin, R., & Xie, J. (2014). Understanding the adoption Of third-party online payment: An empirical study of user acceptance of Alipay in China.