What are Chatbots, and How Can They Improve Customer Experience?
Nowadays, consumption has become an indispensable part of our lives. We all hope to have smooth and pleasant purchasing experiences, but unfortunately, things don’t always turn out as we wish. We will inevitably encounter some problems during our purchases, and in these situations, we must rely on customer service to give us a hand.
In the past, customer service systems were all operated by human workers, and consumers can get the answers to their questions through communications. Nowadays, ‘robots’, or to be more specific, ‘chatbots’, have also been applied in the business industry in order to complement human service systems. In this article, we would like to discuss how chatbots can serve as useful tools to enhance customer experience.
The Background of Chatbots
Before we explain how chatbots can provide assistance to customers, we would like you to know a little background information about chatbots.
A chatbot, as its name suggests, is a kind of robot that ‘chats’ with people. Designed to stimulate human conversations, chatbots analyze inputs from users and generate responses either through typed texts displayed on the screen or voice commands.
The invention of chatbots can be traced back to the 1960s (Okuda & Shoda, 2018), when a computer scientist named Joseph Weizenbaum came up with a computer program named ELIZA. It is a program that can carry out conversations with humans by answering them with scripts designed in advance by the mechanist. It is known for its use of simulating a “psychotherapy session” (Dawson & Phelan, 2017, p.634) between a doctor and a patient. ELIZA plays the role of the therapist and responds to the patients’ questions. This breakthrough enlightened the whole technological industry, and more and more experts devoted themselves to the improvement of this tool. Since then, the development of chatbot has bloomed and expanded in various fields.
The Use of Chatbots
One common use of chatbots is virtual assistants, which are designed for personal use. A virtual assistant is “an application program that understands natural language voice commands and completes tasks for the user” (Rouse, 2017), such as Siri and Google Assistant in smartphones. The assistants can immediately offer information that the users inquire, and provide shortcuts to different applications on the phone.
There are also chatbots designed to offer users information directly from the professionals. For example, BotsCrew (n.d.), a chatbot company, introduced “Health Tap,” a healthcare chatbot from which users can get doctor’s suggestions immediately after they type in specific symptoms. Also, Bank of America releases a chatbot that provides financial related updates and tips on how to save money in order to help customers make reasonable decisions (Venkat, 2020).
In addition, there are chatbots designed for customer service, and this type of chatbot will be our main focus in the following paragraphs.
Chatbots for Customer Service
Perhaps you might wonder, why do companies need chatbots to improve customer service? Aren't human customer service systems effective enough?
Try to recall your memories of using the customer service systems. There should have been one or more experiences when you were frustrated with the service. It might have been the time when you ordered food delivery, but you still haven’t received your meal after waiting for a long time. You decided to report it to the customer service, but it turned out that no one answered you to assure that the courier would arrive as soon as possible. Or it might have been the time when you contacted the customer service for some problems you encountered during your purchase. You supposed that you would receive polite and informative responses, but unfortunately, the person behind the system replied in a bad attitude for no reason.
In these examples, you, as a customer, must have felt dissatisfied with the service, either the inefficiency or the inappropriate attitude. Fortunately, chatbots can be the solution to these problems.
Types of Customer Service Chatbots and How They Work
Before revealing how chatbots can improve customer experience, you might wonder how they actually work as customer service assistants. In fact, there are two types of differently designed chatbots that are able to carry out customer service in the market, which are rule-based chatbots and intent-based chatbots.
● Rule-based chatbots
As the term implies, rule-based chatbots deliver responses to users based on a series of determined rules that are designed beforehand. This type of chatbot makes anticipations of what kind of questions customers may ask. The chatbots’ reactions to the expected questions are programmed in advance.
In practice, most rule-based chatbots are designed only to respond to simple requests by providing buttons for customers to select from. When customers choose a button, they are immediately led to predetermined responses based on their selections. These responses might include information about the company’s recent sales campaign, links to certain products, or shortcuts to FAQs.
The figure below is a simplified example of how a rule-based chatbot may work.
Figure 1
How a Rule-based Chatbot Works
Rule-based chatbots are really easy to operate, and they are able to provide customers’ desired information promptly and accurately. However, rule-based chatbots have a fataldisadvantage – they are not capable of answering questions beyond pre-set rules. They can only develop very limited conversations with customers, and are unable to handle complex problems and deal with scenarios they are not programmed for.
● Intent-based Chatbots
Intent-based chatbots, on the other hand, are able to generate answers to more complicated questions. These chatbots analyze the intent behind the customers’ input, so that they can provide appropriate responses. In order to achieve this function, they use a technology called Natural Language Processing (NLP). WhosOn (n.d.), a company that specializes in chat projects, defined NLP as “an element in AI that allows machines to derive understanding from language as it is used in natural conversation.” According to Academia Sinica (n.d.), NLP is composed of two stages. First, the intent-based chatbots look for keywords in the customers’ sentences to acquire important information. Next, they take the context of their previous conversations into consideration, and figure out what the customers intend to know. Finally, after following these steps, the chatbots can deliver the response to customers.
The figure below is a simplified example of how an intent-based chatbot may work.
Figure 2
How an Intent-based Chatbot Works
By using NLP, intent-based chatbots can understand texts and generate natural responses like humans. Also, they are able to recognize what customers want to ask even if the same question is written in different ways. Therefore, conversations with intent-based chatbots have less restrictions than those with rule-based chatbots.
Moreover, NLP introduces an algorithm named “Machine Learning” (Academia Sinica, n.d.), which can help build an algorithm model that allows computer programs toanalyze data, find patterns, and figure out the characteristics of different languages. With the assistance of the Machine Learning model, intent-based chatbots can keep learning from communicating with customers, evolve, and carry out conversations in an even more natural form.
How Chatbots Benefit Customers
Although rule-based chatbots and intent-based chatbots operate in different ways, they bring similar benefits to customers.
One of the biggest benefits of customer service chatbots is that they can provide responses for consumers right away. Consumers might have the experience of waiting for a long time without receiving responses from the customer service representative. It is reasonable because a human being cannot sit in front of the screen all day long without any rest, waiting to answer customers’ questions. Moreover, human workers are not capable of giving responses to every consumer immediately, especially during the peak season. Unlike humans, customer service chatbots can provide assistance for customers twenty-four hours a day. Also, they are able to respond to multiple customers at the same time. Therefore, consumers can get instant answers from chatbots anytime they wish. This can eliminate customers’ frustration that arise from long waiting time.
Another feature that differentiate chatbots from traditional service is that chatbots have the ability to always keep a consistent attitude towards customers. Chatbots have no emotions, so they never lose patience like humans do. Therefore, it is impossible for them to ruin the deal because of emotional issues. They will not fly into a rage just because you keep bothering them with endless questions; they will not blacklist customers. Instead, they will always treat customers politely and professionally, which guarantees the quality of conversation.
Chatbots can make up for the parts that human workers are not able to carry out. They can reply to consumers instantly, even during busy periods, and they can always maintain a patient tone and sincere attitude toward customers, because they will not be influenced by negative emotions. They show the profession as excellent customer service agents in various aspects. With these benefits, chatbots can boost customer satisfaction without doubt.
Customers’ Perspective on Chatbots
For decades, customer service systems in the business industry have been dominated by human agents. Compared to using chatbots, customers tend to be more familiar with human-human interactions. However, while chatbots can indeed benefit customers, do customers really know about and appreciate this new technology? To understand customers’ attitude towards the application of chatbots in customer service, we designed a questionnaire by using Google form . A total of 155 responses were received.
Figure 3
Preference Towards Customer Service Chatbots Versus Human Service Systems
This is a pie chart (figure 3) about consumers’ preference between customer service chatbots and human service systems. 53% of the respondents selected human operation as their choice when it comes to customer service, while 26% preferred chatbots. At the same time, 19% of the participants saw no difference between the service quality provided by either humans or chatbots, which means they could accept both of them as customer service. Still, we can see that over half of the respondents preferred human service according to their experience.
The bar graph below (figure 4) shows the possible reasons for this result.
Figure 4
Drawbacks of Customer Service Chatbots
The bar graph (figure 4) showcases the drawbacks of chatbots that the respondents agree with. Over half of them viewed ‘unable to deal with complex problems’, ‘unable to give accurate answers’, and ‘incomplete functions’ as the disadvantages of chatbots service, standing at 86.45%, 67.74%, and 50.32% respectively. These drawbacks might decrease consumers’ preference towards customer service chatbot.
Even though more of our participants favored human service systems over chatbots, it is also true that the majority of them did regard chatbots as useful tools. We asked our participants to rate customer service chatbots’ degree of use usefulness on a scale of 1-5, 1 being the lowest, and 5 being the highest. The results are shown in the graph below (figure 5).
Figure 5
Usefulness of Customer Service Chatbots
From figure 5 we can see that 68% of the respondents, including the ones that vote for ‘4’ and ‘5’, viewed chatbots as quite practical tools. On the other hand, less than 10% of the participants considered chatbots as extremely useless devices. It turned out that the consumers who had the experience of using customer service chatbots actually gave quite a positive review of their usefulness.
Based on this result, it can be inferred that chatbots still possess some features that can bring benefits to consumers. In our questionnaire, we listed some advantages of customer service chatbots and asked our participants to select the ones they agreed with, and the following figure (figure 6) shows their answers.
Figure 6
Top Six Benefits of Customer Service Chatbots
Figure 6 lists the top six benefits of customer service chatbots according to our participants. Here, we are going to focus on ‘24-hour availability’, ‘prompt response’, and ‘lack of personal emotions guarantees quality of conversations’, which are the three top-ranked benefits. Their percentages were 78.81%, 67,10%, and 43.87% respectively. It turned out that most of the participants identified with the benefits we mentioned in the previous section. Thus, it can be seen that chatbots are generally agreed to be useful tools which give customers easy access to prompt and friendly services.
We acknowledge that at the current stage, customer service chatbots are far from perfect. They have limited functions, and lack the ability to really think and solve problems like humans. However, it is an undeniable fact that chatbots can and do serve as very useful tools to improve customer experience.
Case: Kentucky Fried Chicken (KFC) Taiwan’s LINE Chatbot
In order to give you a more concrete idea of how chatbots help improve customer experience in the real world, we would like to introduce a case, which is Kentucky Fried Chicken Taiwan’s LINE chatbot.
You have probably heard of Kentucky Fried Chicken (KFC), one of the world’s most popular fast food restaurant chains. Its branch in Taiwan (KFC Taiwan) launched its Line chatbot in September, 2018 (Cheng et al., 2019). Their chatbot is a rule-based chatbot named Xiao K.
The following points are some features of Xiao K that help KFC enhance customer experience.
● Convenient Shortcut Menu
On the bottom of Xiao K’s chatroom, there is a menu that provides shortcuts to certain information, such as updates of the company’s flash sales, new products, and popular meals. By clicking the buttons on the menu, customers can easily acquire the information they want to know without even typing a word. This function can effectively reduce customer’s time spent on browsing through the company’s website, looking for the information they desire.
● “Order Food” Function
KFC Taiwan was the first fast food company in Taiwan that allowed consumers to order food on its LINE chatbot (Cheng et al., 2019 ). Customers can start to order their meals by clicking on the "Order Here” button in the chatroom, or by typing in sentences such as “ I’m hungry”, or “I want to order some food.” Xiao K will then guide the customers through some simple steps to choose the food he or she wants to order, the payment method he or she prefers, and the branch store he or she would like to go to pick up the meal when it is ready.
Usually, in order to enjoy a meal at KFC, customers would have to wait in long lines before it is their turn to order, especially during lunch hours. Xiao K’s “Order Food” function provides a more convenient and efficient way for customers to order food. Just by clicking on some buttons and typing a few inputs in the chatroom, customers can select what they want to eat in advance, and pick up their meals at their desired time. This successfully eliminates customers’ waiting time, and thus boosts customer experience.
● Friendly Conversation
Speaking to Xiao K is like speaking to a friend. It is designed to talk in a friendly and perky style, and may sometimes give responses that would lift the corners of your mouth. For example, if you ask Xiao K, “Are you male or female?”, Xiao K would answer something like, “My boss told me to not to say.”, and add a playful emoji at the end of the sentence. If you tell Xiao K that you want to eat McDonalds, Xiao K will reply, “What in the world is wrong with you?? Don’t think about other foods when KFC has the world’s most delicious fried chicken ^^”. This design can spice up the conversations and lighten the mood of customers.
Data released on Business Next (2019) indicated that the percentage of customers that used chatbots were two times more than the percentage of people who ordered on the official website. It showed that consumers are quite willing to use KFC’s chatbot compared to the original online order platform. With well-designed functions, KFC’s chatbot successfully improved the quality of customer service, and has become a new practical tool for its consumers.
Video 1
Case: KFC’s Line Chatbot!
.jpg)
Conclusion
At the present time, it is true that the technology of customer service chatbots still has room for improvement. However, there is no denying that chatbots can indeed deliver a higher quality service than human agents in some aspects. By using customer service chatbots, customers can receive responses instantly, which is unlikely to be done by human workers. Moreover, customers will no longer have to face the unpredictable attitude of customer service representatives, because chatbots will always reply in a polite manner. With these outstanding advantages, chatbots can complement the insufficiency of human service systems, and serve as convenient and practical tools that can enhance customer experience.
References
Academia Sinica. (n.d.). Duan Kai Jhong Wun De Suo Lian! Zih Ran Yu Yan Chu Li
[Break the Chains of Chinese! Natural Language Processing]. Yan Jhih You Wu.
https://research.sinica.edu.tw/nlp-natural-language-processing-chinese-knowledge
-information/
Cheng, Y. H., Gao, J. Y.,& Siao, M. Y. (2019). Jhen Ren Chih Bu Sia De Ke Fu Liang,
Ta 30 Miao Gao Ding! 3 Ge Cheng Gong An Li, Kan Chatbot “ Liao” Chu Da
Shang Ji. [The great amount of customer service which human workers can not
digest can be handled by ‘them’! 3 successful cases that demonstrate how
chatbots bring big business opportunities.] Business Next.
https://www.bnext.com.tw/article/54610/chatbot-kfc-skyscanner
Dawson, H. C., & Phelan, M. (Eds.). (2017). Language files: Materials for an
introduction to language and linguistics (12th ed.). Columbus: Ohio State
University Press.
Five industries that benefit from chatbots already. (n.d.). BotsCrew. Retrieved
December 19, 2020, from
https://botscrew.com/blog/industries-benefit-chatbots/
Okuda, T., & Shoda, S. (2018). AI-based Chatbot Service for Financial Industry.
Fujitsu. Retrieved December 19, 2020, from
https://www.fujitsu.com/global/documents/about/resources/publications/fstj/archi
ves/vol54-2/paper01.pdf
Rouse, M. (2017).Virtual Assistant (AI Assistant). TechTarget.
https://searchcustomerexperience.techtarget.com/definition/virtual-assistant-AI-as
sistant
Venkat, K. (2020). Examples and Best use cases of chatbots in Banking Industry. Voice
Tech Podcast.
https://medium.com/voice-tech-podcast/examples-and-best-use-cases-of-chatbots-
in-banking-industry-1a46cebdb0cd
What is an intent-based chatbot? (n.d.). WhosOn. Retrieved December 19, 2020, from
https://www.whoson.com/chatbots-ai/what-is-an-intent-based-chatbot/
影片看這裡：What are Chatbots, and How Can They Improve Customer Experience?
