Technology in Customer Service:

A study on the effects of the removal of humans in customer service at Amazon

CHAPTER 1

Introduction

Introduction

Artificial Intelligence (AI) has gotten the attention of businesses all over the world. This interest is due to the potential a company can save on overhead cost. The current AI Technology in 2020 includes but not limited to chatbots that are computer programs which trigger human speech or actual robots that can do the day-to-day tasks. The technology is still being developed and constantly improved but the current models are capable of replacing jobs such as call agents, concierge, client relations associate, customer service representative, etc. Robots have already taken over jobs in manufacturing plants and warehousing facilities. It’s inevitable that as technology advances it will start to affect every industry. According to Wald (2017, August 3), automation could save businesses $4 million annually. Four million dollars, an amount like that would persuade even fortune 500 business to make this switch and this technology is usable in almost any industry that makes the loss of millions of jobs a reality. Amazon is one of the biggest tech companies in the world and they have their hands in almost every industry in one way or another.  Amazon continues to lead the way in innovative business practices and continue to take huge chunks of market share in those industries. For that reason, the researchers will determine if removal of a human representative in Amazon’s customer service department will lead to loss of customers’ purchases and customers.

Problem Statement

The researchers argue that although these technologies are a great way to assist with mundane tasks and could save Amazon millions of dollars if used with Humans. However, the removal of the human component altogether will cause more damage to Amazon’s image than it will help. Money is important for businesses to think about, but the money isn’t as important as your customers. The customer is king especially for Amazon because removing customers from the equation will eventually put Amazon out of business.  According to Brown (2018, July 3), when you make external changes to your business, customers will usually react in one of two ways. First of all, they might be receptive to the changes, most often when the changes don’t require a huge shift in attitude or behavior and won’t take them out of their comfort zone. On the other hand, customers might be resistant to the change. This generally happens when they have to adjust how they interact with your business, whether it’s online, via an app, or in person. In this case, your changes will be more likely to have lingering negative effects. Amazon may not see the negative effects it may have on its customers when making this change to a fully automated customer service department.

Background and Need

For most companies, saving millions of dollars could be the difference in keeping the doors open or closing down for good. It almost impossible for companies to get ahead in any industry because of the amount of competition. So being able to have extra resources can give a company a competitive edge. Some companies competitive edge is the service that customers receive from the customer service representative. In essence, underestimating the need for human connection within customer service can have detrimental effects on the entire company as a whole.

Barriers and Issues to the Conducting of the Research

The barriers and issues that can occur while conducting this research is being able to go through the vast amount of information available on Amazon. Knowing what information to use and when to use it is a challenge because of the 8-week time frame. It’s almost impossible to conduct and create a compelling argument through a thesis in as little as 8 weeks. Also, a huge issue could be the accuracy of the information that’s obtained. Although the information is coming from credible sources, there is no way to know if the information is current because none of the researchers have insider connections. So, the majority of the information presented could be outdated or no longer relevant. These are all factors that the researchers will have to overcome to deliver a persuasive stance on this topic.

Barriers and Issues Related to the Research

The barriers and issues that the researchers are likely to experience during their research is gathering data that is available on Amazon. It will be difficult to determine what information is important and which one is irrelevant. Another issue is that the researchers are on different work schedules making it difficult to find a time to discuss changes or findings. However, the biggest barrier of this research is being able to understand the thesis and research it as a unison as possible. With many hands and thoughts into this research it will be difficult to ensure that the information delivered is not contradicting each other. The researchers will have to critically think and overcome.

Difficulty of Study

The main difficulty in this study is the fact that it is team-based which means that the team has to have a high team-work culture for the study to be effective. Group effort and efficient communication will be required to make the study a success. Another difficulty that is likely to be experienced in this study is that some sensitive and confidential information about Amazon might not be available to the public domain and this will limit the scope of the study.

Proposed Solutions

The researchers are proposing that human interaction in customer service is a crucial component to have a successful business. AI or other technologies should be used to enhance the customer service representative not replace them. The researchers also believe that it will be more effective for both consumers and Amazon to have the traditional customer service because it will help keep the company’s credibility. This will result in higher customer retention that will boost sales and keep the company revenue on a constant growth.

Hypothesis

Amazon will lower their purchase rates and lose customers if they fully automate their customer service department

Research Questions

• What type of technology is already being used? And how effective are they?
• How will customers react to this new system?
• What is the potential loss or gain if Amazon fully automates their customer service department?

Limitations

The researchers are only limited to the information that they can gather online or through surveys. No insider information will be available to them, which means that all of the information that will be gathered may not be as accurate as it is expected to be. In addition to this, the researchers are only limited to 8 weeks of class, therefore, the accuracy of results may be compromised due to the short time given to this study. Programming language or any kind of artificial intelligence technicality will also not be discussed in this study because the researchers cannot be able to secure and study the technical specification of this program and how it works. The researchers will only focus on the general information they can gather to the automation of customer service of the company.

Delimitations

The researchers cannot conduct a survey that can get the feedback of every single customer that Amazon may have, therefore, the researchers will only limit their number of target respondents and use a survey methodology that may help the researcher analyze the data gathered. Moreover, the researchers can only also analyze the automated system being used by the company through actual use of their system, and test how effective the automated system Amazon really is to address customer service. The researchers will then propose a recommendation to improve the services if possible.

Definition of Terms

The key terms used in this study have been defined below:

Artificial Intelligence: This is the development of intelligent computer systems that can think and work like humans. They perform tasks that would normally require human intelligence, such as decision-making, visual perception, translation between languages and speech recognition.

Customer service: This is the act of meeting the needs of the customer through high-quality service before, during and after doing business with them.

Chatbots: Chatbots are conversational agents used by software applications that mimic human speech in either spoken or written form. Their main purpose is to stimulate a conversation and interact with a real person.

Expert Systems: Expert systems are software that makes decisions and offer advice using expert knowledge databases. Expert systems are widely used in medical diagnosis.

Amazon: Amazon is a multi-technology company that deals with e-commerce, artificial intelligence, digital streaming, and cloud computing. It falls in the category of the Big Four tech companies worldwide.

Google: Google is a multi-technology company that deals with internet-related products and services such as online advertising technologies, software & hardware, search engine, and cloud computing.

Summary

In conclusion, Amazon should consider creating a healthy balance between automated customer service and the human aspect of customer service because there are some human attributes that Artificial Intelligence cannot replace. Artificial Intelligence has its advantages and disadvantages that Amazon should consider. The pros include a reduction in human error, availability at all times 24/7 with no human-related obligations and responsibilities, faster decision-making, helping in repetitive jobs that are boring to humans and other forms of digital assistants. The cons of AI include high costs of creation, implementation and maintenance, unemployment, lack of emotions and inability to think outside the box. In as much as Artificial Intelligence is more efficient and gets things done faster, the human aspect is very important to retain in business because it is what makes customers feel like there is an actual person that cares, and they are not just dealing with emotionless automated systems.

CHAPTER 2

Review of the Literature

Introduction

AI has been around for a while now, but the amount of AI technologies has grown in the past years and continues to grow. These advancements in AI are mostly due to the huge consumption of the internet and how cheap these technologies are today. The more information that is widely available, the more AI systems can continually grow and advance to new heights. Walker (2019) believes that artificial intelligence is currently being deployed in customer service to both augment and replace human agents with the primary goals of improving the customer service experience and reducing human customer service costs. The technology is not yet developed to where it can do all of the tasks, it can handle most simple requests without human input. As AI advances, so will the task it is able to do and soon humans will not be needed.

The researchers hope to identify technologies that are currently available, if and how customers prefer to communicate, and lastly, what Amazon competitors are doing and how effective their tactics are. To be able to answer all these questions the researchers must dive into several areas of research, but due to time restraints, this research will not include a complete list. This area presented in this research is the areas the researchers deemed the most important to get their points across. Those areas of research are as follows:

1. Technology
   1. Artificial Intelligence
   2. Expert Systems
   3. Chatbots
2. Amazon Customers
   1. Amazon’s Customers
   2. Amazon’s Customers satisfaction
   3. Amazon’s Customers preferences
3. Competitors
   1. Google
   2. Apple
   3. Microsoft

Research Area

Artificial Intelligence 

Artificial Intelligence, also known as machine intelligence, is the development of intelligent computer systems that can think and work like humans. They perform tasks that would normally require human intelligence, such as decision-making, visual perception, translation between languages and speech recognition.

The main challenge of AI is that some people do not understand how it functions and can lead to trust issues where it is involved. Other people find it too complicated and therefore choose to ignore it. The other challenge of AI is that it cannot replace every task, some tasks still require human involvement (Gunning, 2017).

Artificial Intelligence has its pros and cons. The pros include a reduction in human error, availability at all times with no human-related responsibilities, faster decision-making, helping in repetitive jobs that are boring to humans and digital assistance. The cons of AI include high costs of creation, implementation and maintenance, unemployment, lack of emotions and inability to think outside the box.

Expert Systems

Expert systems are software that makes decisions and offer advice using expert knowledge databases. Expert systems are widely used in medical diagnosis, education, and engineering. The four main components of expert systems include a knowledge base, user interface, development engine and inference engine (Russell, & Norvig, 2016).

The Pros of using expert systems are that they are consistent, have a huge amount of memory, use logic and are always available when needed. The other pro is that they have no fixed age limit, unlike humans. The cons of using expert systems are that they require to be manually updated every now and then. This is because they have no ability to learn on their own. The other con is that they require a lot of time and cost input for the systems to run efficiently. They are costly to implement and maintain. Expert systems are also very specific and only have specific information. They have no ability to use commonsense and reason like humans. They solely depend on the data that is fed into them. The other con of expert systems is that they are emotionless and cannot assess a situation and act accordingly.

ChatBots 

Chatbots are conversational agents used by software applications that mimic human speech in either spoken or written form. Their main purpose is to stimulate a conversation and interact with a real person. Some chatbots are based on Artificial Intelligence while others just have fixed information fed into them.

The pros of using chatbots are that they provide quick customer service which leads to more customer satisfaction. They are multipurpose and reduce labor costs because they perform the work that would normally require several employees. Through the use of chatbots, companies can manage many different clients at the same time. They are also available 24/7. The cons of using chatbots are that they are expensive to create, implement and maintain, they offer limited responses to customers which can result in frustrations, and they are not to be used for all business. Some businesses require responses that are way too complex for chatbots to handle.

Research Area

Amazon’s Customers

As shown in Figure 2.1, by June 2019, there were an estimated 105 Million U.S. Amazon Prime Subscribers (Clement, 2019). Of those subscribers, the ages range from 18 years old all the way to 55 years old and above (Clement, 2017). However, to keep this research small and contained, this data does not include amazon’s international users. It shows that more than half of Amazon users are 35 and above and it is important to discover what the preferences of those users are before implementing a fully automated system.

Figure 2.1 Distribution of Amazon Prime Members in the US

Amazon’s Customers satisfaction

As shown in figure 2.2, over 60% of Amazon customers were very satisfied with the service they receive from Amazon and 28% were rather satisfied. It’s amazing to see that only 5% of amazon customers reported that they were dissatisfied with their customer service from Amazon. For nine consecutive years, customers have ranked Amazon number one in customer satisfaction with a customer score satisfaction score of 87 out of 100 (Forsee.com, 2020). 87 is the highest score amongst its competitors like Costco, Best Buy, and Walmart in terms of customer satisfaction.

Figure 2.2 How satisfied are you with the customer service of Amazon

Amazon Customer Communication Options

Currently, you can contact Amazon customer service in four different ways; Phone Call, Email, Live Chat, or through social media (John, 2019). When calling Amazon through a phone call, you are approached with an operator that routes you to a human representative, but it is not yet fully automated. All of their communication lines are still heavily reliant on the human touch. Amazon must be doing something write because they are number one in customer service satisfaction, but they are still being operated by humans not machines.

Amazon’s Customers preferences

A big challenge of AI technology is if the general public will accept and will they trust it. Investment into something that people will not want to use could be a huge hit even for a huge company like Amazon. So, it is important to consider the customer every step of the way in decisions that will directly affect them. When it comes to AI adoption, however, most survey respondents are still getting their feet wet, and are split in their use and opinions of AI in both their personal and business lives (Krogue, 2017). Users will ultimately have to trust the AI with their personal information and recent surveys have shown that the majority is not ready for that yet.

Research Area

Competitors

Google

Google uses the conversational neural network, Meena, with 2.6 billion parameters that can chat with people better than any AI generator out there. The team trained the model with 40 billion words — 341 GB of text data including social media conversations — using the seq2seq model. Seq2seq is a variation of Google’s Transformer — a neural network that compares words in a paragraph to each other to understand the relationship between them (Metha, 2020).

As seen on the figure 2.3, Google has been putting in so much effort to make their artificial intelligence better than it used to be. Right now, Meena is at 79% when the sensibleness and specificity average is measured. This means that Google’s artificial intelligence is smart enough for it to understand and interpret humans. This may be the technology that can be used for customer support in providing customer service in the future. However, comparing it to humans, it is still less since the system is still under continuous improvement.

Figure 2.3 Sensibleness and Specificity Average

Apple

While Google is focusing more on improving its Meena as its automated chatbot for its business customers, Apple on the other hand is focusing on live person interaction type of customer service and calling it Business Chat. This customer service is encouraging businesses to avoid chatbots or bots. Apple believes that customer service is all about human interaction and not about automation. Apple thinks that it is faster if a live person can help the customer instead of using an automation system that other companies, such as Google and Facebook, use (Berger, 2018).

Figure 2.4 Apple’s Business Chat User Interface

Apple’s Business Chat user interface when using an iOS device is seen above. As show in figure 2.4, Apple is promoting customer service through messages in which, they will get a response from a human rather than a machine. This is to promote their belief that customer service is all about customer interaction.

Microsoft

Microsoft continues to tweak its tech support to deliver the best possible service. The company has greatly expanded its video offering on its YouTube channel. Microsoft has also improved the Virtual Assistant, enabling it to handle many of the general questions that consumers typically ask (Smith, 2019). If a customer wants to speak to an actual human being, Microsoft has 10 worldwide call centers, including centers in the United States, Latin America, China, and Europe. The researchers decided to put Microsoft’s tech support through its paces by grabbing a Surface Go, making calls, and pinging social media accounts. The researchers asked several questions, including how to disable the webcam, turn off or delay automatic Windows 10 updates, enable Cortana, and change pen shortcuts.

As seen in figure 2.5, Microsoft’s customer satisfaction is ranked 8th. This means that their web support and their phone support is outstanding since it entered the top 11 companies that provide customer service.

Figure 2.5 Technology Company’s Customer Support Ranking

Summary

It is undeniable that technology will continue to improve to the point that it may be able to do all the tasks that customer service representatives can do. Soon companies will be able to completely replace their human workers with Artificial intelligence and it is already happening now. Companies like Google, Apple, Microsoft, and Amazon are slow-moving the mundane tasks from their human representatives and letting the AI take charge of them. With this research, it is hard to determine if fully automating customer service assistance will lead to lower customers and customer purchases because these companies are using different approaches. To truly know if the hypothesis is correct, the research would have to be based on customer preference. Only after conducting surveys will the researchers be able to get a better understanding of what customers prefer when dealing with customer service.

CHAPTER 3

Research Methodology

Introduction

The Purpose of the research was to know with certainty whether or not transitioning to a fully automated customer service department would lower Amazons sales. Therefore, we explored on many levels how advance the AI is today, how the customers perceived the service provided by the company, and how customers prefer to communicate when it comes to customer service. This research was done through a mixed methods approach. The researchers began by collecting data from existing studies then by surveying people within their local community. The data gathered will then be compared and used to answer the research questions of this study.

Nowadays the race for the perfect AI is ongoing and this trend is evolving to a point where the fundamentals were broken when the AI was able to pass the test of Turing, meaning that a human will not be able to differentiate an AI from a human. Amazon will benefit from this advancement if they implement this type of AI as Samsung or Apple have done so already. But making gigantic advancements in technology will not necessarily be accepted by the customer. This research will show if customers will use this new service or if the lack of humanity in this service tends to be a real threat for Amazon and the other stakeholder that will use AI (Kaplan, 2019).

Research method(s) employed

To have a better understanding of the situation the researcher have to understand directly what customer think about AI. The best way to figure this out is by using a quantitative approach as a hard review on a base of Pegasystems research, a leader in customer engagement software. Pegasystems conducted a global study to measure consumer attitudes toward AI and, more specifically, what they think of AI used in customer experience. In total, 6,000 adults were surveyed in North America, EMEA, and APAC. This study can be used as a hard review regarding the importance of the sample of the person selected, in fact it will reduce the error concerning the trues of the information needed (Pegasystems, 2019).

The second method employed will be utilizing a quantitative approach through google forms to survey Amazon specific users. The researchers aim to reach amazon users ranging from ages 18- 55 and older. The survey will include questions as shown in Appendix A. The main purpose for this survey will be to determine whether the findings from Pegasystems also pertains to Amazon customers. The survey conducted by Pegasystems was a global survey whereas the survey the researchers will be conducting will reach mainly Amazon customers in the San Diego area. The researchers hope to find patterns and connections between these two surveys to support the hypothesis.

Specific procedures employed

The survey mentioned in Appendix A will be shared on many avenues such as Facebook, Instagram, Twitter, Emails, and Text messages. Each question within the survey has a specific purpose to assist the researchers determine if the hypothesis is valid. The purpose for each question is as followed;

• Questions 1: What is your Age?
  • Knowing the fillers age will allow the researchers to what age demographic they are reaching. More importantly, the researchers aim for a representative sample of each population for every generation. Knowing how each age group responds to the survey will give great insight into the preferences when it comes to communicating with customer service. If the results show that all age groups, feel the same way this will provide us the strong evidence we need. On the other hand, if the results show different feelings than then hypothesis will be harder to prove.
• Questions 2: Do you purchase items from Amazon?
  • The researchers are looking to survey only Amazon customers due to the fact that the thesis is based around Amazon and their customer service. Knowing whether the participant is an Amazon customer or not will show us how exactly Amazon customers feel about Amazon’s current customer services. Also, it will provide insight into how Amazon customers specifically prefer to communicate when contacting Amazon’s customer service. With that being said, if the participants answer this question with no the survey will end and submit. The reason for ending the survey so early is to prevent gathering of unnecessary information.
• Question 3: Have you ever needed to call Amazon’s Customer service?
  • This question is also to prevent unnecessary information. If the participant answers no, the survey will end and submit. The ultimate goal for this survey is to target those who contact and communicate with Amazon’s customer service. If the participant has never needed to contact Amazon’s customer service department before then their input is of no significance to this research.
• Question 4: On a scale from 1-5, how would you rate the service you received?
  • It is important to see how customers feel about the service they are receiving from Amazon’s current customer service practices. This will show whether or not Amazon’s current practices are effective or if they can be improved by fully automated systems. However, if a participant is not satisfied it does not necessarily mean that automation would fix the issues. Which is why this question is so important, it will provide insights into what areas could be improved and investigate how to improve them.
• Questions 5: When contacting Amazon’s customer service department, do you prefer communication with an automate machine or a human representative.
  • This is a follow to question four. This will allow the participant to directly inform us what about their experience they did not like. Also, what about the service could be improved but more importantly if that issue could be solved by full automation.

According to survey monkey, 80% of responses to surveys are gathered within seven days of the survey being released and only 11% of responses are collected during the second week.  So, for the purposes of this research project, the researchers will keep this survey open for 10 days. This will give the researchers about 80% – 90% of responses to have enough data to make an educated decision on the hypothesis.

The survey conducted by Pegasystems starts with questions related to the knowledge that people have about the AI. Then the survey continues on specific questions regarding the way that people are using AI and where they will be more confident with it or not by the type of use. This study has been realized globally, in fact it has been done in North America, Europe, Middle East, Africa, Central and South America. The goal was to have a global approach on this subject because it will be the future of everyone that will have access to the internet or a phone.

After gathering information from both surveys, the researches w

Formats for presenting results

The results of the Pegasystems findings will be displayed in forms of figures and infographics such as figures 2.7 and figure 2.8. All of the questions from the Pegasystems survey will include their final results. However, to give a meaning to the result in respects to this research the researchers will be utilizing the Pegasystems as a systematic review. The information on the comparison will be displayed in side by side graphs and figures to show the correlations within customer perception with AI and customers preference in regard to customer service. With these two segments the researchers will be able to show that the hypothesis is indeed true.

Figure 2.6 Customers feelings towards AI

Figure 2.7 Customer preference when connecting with customer service

Resources Used

The main resources of this research are the study made by the company Pegasystems, in January 2019, named What Consumers Really Think about AI. The researcher will use this study because it is relevant the study they are conducting based on a smaller sample and based only in America. Another reason that makes this study more relevant as a research method for the researcher is, the fact that Pegasystems is a major actor in the development of the use of AI as a customer service, that makes their study relevant to this study. The secondary resource that researchers will be using is their own survey

Summary

The data reveals that there are confused consumers who are split on their feelings toward AI, in fact, over one-third say they are comfortable with AI used by businesses, while just under one-third say they are not. And then there is another third who say they just do not know yet. To be sure, many consumers are excited to embrace the benefits of AI and see a promising future ahead. But some harbor deep-rooted fears about AI and most still prefer the familiarity of the human touch over a faceless machine when given the option. Amazon has the choice to implement it for its customer relationship but as this study showed it will not be as preferred by the customer as human interaction and they might lose a considerable number of customers because of that. The ideal way will be the mix between the human and the IA (Pegasystems, 2019).

CHAPTER 4

Results

Introduction

During this step the researchers took over the data generated in the purpose of getting information to confirmed or infirmed the following hypothesis; Amazon will lower their purchase rates and lose customers if they fully automate their customer service department.

The result that the researcher has been use are coming from two distinct studies. This is allowing a better approach for the collect of information. Indeed, after being global with the first study that the researcher used from the company Pegasystems, consisting on getting a global sample, they have realized a more local focuses study based in San Diego on a local sample. The result will be shown in a way that the two studies will allow the researchers to be able to compare the two different type of sample and making correlations that will highlight the similarities and the differences. It is important to notice that the two studies have been executed on two different purpose, the first one, to give more credit to a service that Pegasystems want to develop, however it is a very serious study and the result will have to match the market reaction on this service. The second one, the researcher made it for the special purpose of this study.

After collecting all the result from both studies, the researcher highlighted and showcased the result as following.

Findings

The findings of the study indicate that transitioning to fully automated customer service will increase Amazons sales. Data collected from the study shows that customers are embracing Artificial Intelligence even though there are still some who have mixed feelings about it. Even though the human touch is stiff important in various aspects, Artificial Intelligence is exactly what Amazon needs to increase its sales. Being a company that operates mainly online, it would be unrealistic to have actual humans performing the different tasks and responding to every customer query. That would slow down the services substantially because humans are slower at performing tasks compared to Artificial Intelligence (Wald, 2017).

Customer service is just one aspect of a business. There are other aspects such as seamless services, quality of service and pricing that affect the smooth running of a business. By investing in automated customer service systems, Amazon will be able to achieve customer satisfaction through quick, accurate and standard service. Artificial Intelligence is more reliable and has limited errors compared to humans. Artificial Intelligence is also able to detect what the customer wants before they even ask.

Customers have become smarter and they know what they want when they do business with a company. The company, therefore, has to meet these needs and expectations to ensure the customer is satisfied. Customers expect to get feedback as soon as they need it and they need to get it with respect, empathy, and courtesy. Having a fully automated customer service ensures that a customer gets all these in real-time. With advanced technology, chatbots can detect customer behavior and respond accordingly. Businesses like Amazon need to take advantage of these AI-powered solutions to increase customer engagement, reduce dependability, improve productivity and pave way for greater customer experiences.

Summary

In conclusion, and looking at the results, it can be seen that there is 72% of the population surveyed that knows about artificial intelligence. On the other hand, 17% have responded that they do not know anything about artificial intelligence. Moreover, 34% of the population mentioned that was surveyed acknowledged and they are certain that they have interacted with artificial intelligence technology while 34 percent said the other way. Although there are 32% of the people that are not sure if they are interacting with an artificial technology when they contact the customer service department. The researchers are not sure whether they are not aware that they are dealing with artificial technology when they contact the customer service department. As an addition, the researchers also found out that the biggest age group that participated on our survey was at the 18 to 34 years old range. 70 percent of the sample says that they have used Amazon to purchase goods, while 28% mentioned that they have not used Amazon to purchase goods online. 60% have said that they have called the customer service department to ask for customer support and they have satisfactory opinion about how good the service was. 70 percent among within the sample that was surveyed still wants a “Human Representative” to assist them on their needs while 28% mentioned that they really do not mind if AI is the one assisting them.

CHAPTER 5

Conclusion

Introduction

Conclusions

With this study, the researchers sought to prove that Amazon will lower their purchase rates and lose customers if they fully automate their customer service department. As mentioned before, Amazon is currently utilizing a combination of human representatives and some AI systems. Amazon’s current set-up has brought got them to the number one seat in customer service, and they continue to stay there for nine consecutive years. Fully automating Amazon’s customer service could increase or decrease their customer satisfaction ratings, but it is determined by what their customers want. The way the researchers wanted to prove this is by answers three specific research questions. The responses to the research questions are as follows:

• What type of technology is already being used? And how effective are they?
• How will customers react to this new system?
• What is the potential loss or gain if Amazon fully automates their customer service department?

What type of technology is already being used? And how effective are they? 

            In this study, the researchers looked at Google, Apple, and Microsoft to see what they were currently using to service their customers when needed. The trend was that Google, Apple, and Amazon is still using human representatives but with the assistance of AI technologies. The only company that is trying a fully automated AI system is Microsoft, but the system they wish to implement is still improving. So the consensus is that utilization of human representatives with current AI technologies is the best way to assist their current customers. Although this proves that human representatives with AI tools are useful, it does not help us answer if the AI advancements will be more effective. However, as it currently stands, the utilization of human representatives with AI technology is the most effective way to service customer service needs.

How will customers react to this new system? 

According to the results of the study, the majority of Amazon customers still prefer human representatives over automated AI systems. Also, the global study on consumer thoughts on AI shows that only 35% of people are comfortable interacting with AI. When paired with the survey findings that 70% of Amazon customers prefer interacting with human representatives, the researchers conclude that switching to a fully automated AI system would affect 70% of their customer base. The study conducted by Pegasystems also shows that even among the people who use AI, only 55% are comfortable interacting with the AI. This finding is significant because even among those who understand and use AI, there is still 45% of people that are still not comfortable. With so much unease and distrust with AI, implementing a fully automated system as your only way of communication will indeed lower customer sales and purchases.

What is the potential loss or gain if Amazon fully automates their customer service department?

The study conducted was not able to answer this last research question. However, based off of the preference of current Amazon customers, 70% could be affected. Even if only half of the 70% were genuinely frustrated with the AI system, that is still 35% of Amazon customers. Although loss and grain can not be reported, there is a substantial potential loss if they implement a fully automated system. Knowing the exact loss and gain would be another area for further research for those interested in the effects of fully automated systems on customers.

Through the research and the study, the researchers conclude that our hypothesis is indeed correct. Amazon will lose customers and have a lower rate of purchase if they fully automate their customer service system. The people domestics and globally is just not ready to put their full trust in artificial intelligence. However, the researchers believe as artificial intelligence advances, the less need for human representatives become. Also, as the older generations die off, the upcoming generations will live in a world of AI. So in time, fully automated systems will be utilized but just not right now.

Implications

Type Here

Recommendations for Future Research

• Quality product over quality service could be an area for future research. There could be a correlation between costumer’s preference on quality of customer’s service and how they rate their service that hey are receiving without relating it with the quality of the product.

Summary

Type Here

Appendix A: Amazon Customer Survey

We are students of the California School of Management and Leadership program from Alliant International University. We are conducting a Customer Satisfaction Survey on Amazon Customers through Google Forms. This Survey will be distributed in varies platforms such as Facebook, Instagram, Twitter, Emails, and Text Messages. You will find a list of questions below for all participants of this survey. The surveys would be taken privately and individually.

• Questions 1: What is your Age?
  • 17 or Younger
  • 18-34
  • 35-50
  • 55 and Older
• Questions 2: Do you purchase items from Amazon?
  • Yes
  • No
  • Other:
• Question 3: Have you ever needed to call Amazon’s Customer service?
  • Yes
  • No
• Question 4: On a scale from 1-5, how would you rate the service you received?
  • 1 Dissatisfied
  • 2
  • 3
  • 4
  • 5 Very Satisfied
• When contacting Amazon’s Customer service department, do you prefer communicating with an automated machine or a human representative?
  • Automated Machine
  • Human Representative
  • Other:
• Questions 5: When contacting Amazon’s customer service department, do you prefer communication with an automate machine or a human representative?
  • Your Answer:

References

Wald, J. (2017, August 3). How Automation Could Save Your Business $4 Million Annually. Retrieved January 22, 2020, from https://www.forbes.com/sites/waldleventhal/2017/08/03/how-automation-could-save-your-business-4-million-annually/#607cb6823807

Brown, G. (2018, July 3). Managing Your Customers Through Change. Retrieved January 25, 2020, from https://www.entrepreneur.com/article/316161

Clement, J. (2019, September 3). U.S. Amazon Prime subscribers 2019. Retrieved January 28, 2020, from https://www.statista.com/statistics/546894/number-of-amazon-prime-paying-members/

Clement, J. (2017, August 21). U.S. Amazon Prime user age distribution 2017. Retrieved January 28, 2020, from https://www.statista.com/statistics/304940/amazon-prime-us-age-distribution/

Kunst, A. (2019, September 3). U.S. Amazon customer service satisfaction in 2017. Retrieved January 28, 2020, from https://www.statista.com/statistics/705173/satisfaction-with-the-customer-service-of-amazon/

About the Author As a pioneer in customer experience analytics, As a pioneer in customer experience analytics, & Subscribe to the ForeSee Blog hbspt.forms.create({ portalId: ‘2278447’. (n.d.). Why Amazon Ranks Number One in Customer Experience. Retrieved January 28, 2020, from https://www.foresee.com/blog/amazon-ranks-number-one-customer-experience/

Mehta, I. (2020, January 29). Google claims its new chatbot Meena is the best in the world. Retrieved February 5, 2020, from https://thenextweb.com/artificial-intelligence/2020/01/29/google-claims-its-new-chatbot-meena-is-the-best-in-the-world/

Apple’s Business Chat Has a Golden Moment to Dominate Customer Service. (n.d.). Retrieved February 5, 2020, from https://fonolo.com/blog/2018/04/apples-business-chat-has-a-golden-moment-to-dominate-customer-service/

Smith, S. L. (2019, April 24). Is Microsoft Customer Service Good? 2019 Rating. Retrieved February 5, 2020, from https://www.laptopmag.com/articles/microsoft-tech-support

Walker, J. (2019, January 31). Use Cases of AI for Customer Service – What’s Working Now. Retrieved February 5, 2020, from https://emerj.com/ai-sector-overviews/ai-for-customer-service-use-cases/

John, S. (2019, June 13). The best ways to contact Amazon for help when you have a problem with your account. Retrieved February 6, 2020, from https://www.businessinsider.com/how-to-contact-amazon-customer-service-phone-email-chat

Krogue, K. (2017, September 11). Artificial Intelligence Is Here To Stay, But Consumer Trust Is A Must for AI in Business. Retrieved February 23, 2020, from https://www.forbes.com/sites/kenkrogue/2017/09/11/artificial-intelligence-is-here-to-stay-but-consumer-trust-is-a-must-for-ai-in-business/#33ad8064776e

Kaplan, M. H. (2019, July 17). A Brief History of Artificial Intelligence. Berkeley Hass, 61 (4), 5-14.

Pegasystems. (n.d.). What Consumers Really Think About AI: A Global Study. Retrieved February 23, 2020, from https://www.pega.com/system/files/resources/2019-01/what-consumers-really-think-about-ai-study.pdf