What is a chatbot?

A chatbot is a form of artificial intelligence (AI) that communicates with people in their natural language. These interactions take place mostly through mobile apps, websites or social media. Chatbots operate on the basis of various technologies. More complex chatbots use machine learning and Natural Language Processing (NLP) to better understand the context and requirements of the user.

Chatbots are generally divided into two types:

A) rule-based chatbots, where responses are generated based on specific keywords.

B) chatbots based on machine learning, which learn from interactions and become smarter over time.

History of chatbots

The history of chatbots dates back to the 1960s, when the first such ELIZA program was created at the Massachusetts Institute of Technology. This chatbot, created by Joseph Weizenbaum, was able to mimic a conversation with a therapist by transcribing users' questions. While ELIZA was a very simple program, it opened the door to further development and research in the field of artificial intelligence and machine conversation. Interestingly, although ELIZA was incapable of truly understanding or answering questions, many people who interacted with her felt that she truly understood their problems.

In the 1990s, there was another major advance in the field of chatbots. In 1995, a chatbot named ALICE (Artificial Linguistic Internet Computer Entity) was created. ALICE was much more advanced than ELIZA.

Another milestone was the release of a chatbot named SmarterChild in 2001. This chatbot was available on the AOL Instant Messenger and MSN Messenger platforms and was designed to provide useful information such as weather and news coverage.

There has been a huge boom in chatbots in recent years, especially with the rise of social media and increased use of mobile devices. An interesting example of current chatbots is, for example, OpenAI's GPT-4, which uses advanced machine learning algorithms to generate human text. This chatbot is capable of conducting complex conversations, writing essays or even poems.

Advantages of chatbots

Chatbots can improve call center efficiency by:

1. It automatically solves common queries, freeing up time for operators to do more complex tasks.
2. They are available 24/7, do not need holidays and do not go on sick leave
3. Always respond to the user immediately
4. They are able to talk to a large number of users simultaneously
5. Can easily generate statistics of user queries
6. On operating costs they are cheaper than live operators

What chatbots are not suitable for

Although chatbots can solve many different types of queries, they are not ideal for complex problems that require human judgment and empathy. It's simple to ask the chatbot what payment methods you provide to customers in your furniture eshop. However, if the customer asks about how a given piece of furniture visually performs in a small space without windows, then at that moment the chatbot will no longer advise him. In these cases, it is best to combine chatbots with live operators.

Combining a chatbot with a live operator

Typically this service looks like a user clicks on a chat symbol on the site where a chat window opens where the user has the option to simply interact with the chatbot. The chatbot takes all the necessary information from it and then connects it to a live operator, which then resolves the customer's request. The simple recruitment of information from the user is carried out by the chatbot and the complex request is subsequently made by the human. This is a cost-effective option that can provide a high level of quality customer care.

In what situations to use the chat with a live operator

In order to decide whether to use only a chatbot or its variant with a chat with an operator, it is important to know the nature of user queries and their expected quantity. Do users write hundreds of questions a day and ask the same questions all the time? The answer is simple: Get a chatbot! Are the queries diverse and are there in the tens to lower hundreds per day? Then a live operator makes sense.

The simplest decisions are those that can be translated into the speech of numbers. For the purposes of this article, let's talk about an example of a small to medium-sized eshop that has up to 20 employees and its customers have a reason to write requests for technical support and write, for example, 50 such requests per day. For this purpose, it is not economically worthwhile to have an employee in the company who will handle customer cabins. Such an employee will cost the company even with fees of at least 40-50 thousand CZK per month.

Nowadays, it is common to outsource the activity of chat support to a contact center. For example, in Peakforce, one customer check-in costs CZK 19, whether it is a chat or a call, no matter how long it took. With 50 requests per day for 20 working days, the operator handles 1,000 requests per month and charges the client 19,000 CZK. This is not half the cost for the client compared to the own employee, with operators available on the line up to 24 hours a day, professionally trained, regularly checked with artificial intelligence, as well as quality interviews and without the need to directly manage them, recruit them, pay holidays, etc...

If you are interested in outsourcing your chats with customers and would like to learn more, fill out the form below and we will provide you with more information to decide what form of chat to implement for your business.

‍