Chatbots are increasingly common on websites and in messaging apps. They will be in customer service, marketing or sales used. Before integrating a chatbot, companies should ask themselves how the chatbot interacts with users.
There are two main types of chatbots:
1. AI chatbots enable free text input and are therefore usually based on natural language processing
2. Rule-based chatbots that offer defined click structures so that the user cannot enter free text but “clicks through the conversation” using buttons.
The rule-based chatbots in question are not based on AI, while chatbots that understand and process free text using NLP can also do so with the help of AI.
What is NLP anyway? NLP stands for Natural Language Processing and enables machines to understand written or voiced human language. What seems simple at first glance is a highly complex technical undertaking. Because even though it seems self-evident — understanding words and semantic structures is not easy in the case of machines, because natural language and grammar are much more complicated than you might initially think. In addition to the words, the meaning and context of the sentence must be understood, which can be very complex when natural language is ambiguous. In contrast to humans, the computer also cannot draw on experiences and experiences — actually. Because thanks to AI and machine learning, computers are becoming ever smarter. The aim of NLP is to enable communication between humans and computers. The main areas of application include communication between humans and chatbots or between humans and digital assistants such as Alexa or the Google Assistant.
Natural language understanding (NLP) vs. guided chat dialogs
The strengths and weaknesses of both types of chatbots mentioned above are listed below based on many years of experience and practice in the area of chatbots.
NLP chatbot: Using free text inputs to the right answer
Free text input allows users to write their concerns directly to the chatbot. Using modern NLP technology, this analyses the inputs and assigns them to so-called intents (question intentions). If the chatbot has a suitable answer to an intent, it is sent back to the user.
The big advantage for users is that a concern can be reported quickly and easily.
However, it is difficult for the company to answer every question correctly from the start. With every user input, there are three possible results:
1. The chatbot correctly understands the user and gives a correct answer.
2. The chatbot misunderstands the user and gives an incorrect answer
3. The chatbot doesn't understand the user at all and cannot send back an answer.
The last two cases quickly lead to frustration on the part of the user. Only in the first case is there a good user experience (UX). However, the last two cases have an advantage for companies using the chatbot. You can find out what your users/customers would like to ask or know from the chatbot and, if necessary, can react correctly to such inputs in the future or adapt existing intents and expand the answers. By the way, with moinAI, this works automatically, i.e. the AI automatically clusters the questions it did not understand and suggests that the relevant company address these topics. This AI feature is called Dreaming.
The following table shows the advantages and disadvantages of the NLP approach based on a SWOT analysis:
Click chatbot: guided dialogs and buttons
The second approach to send a user the right answer to their request is through guided dialogs. Buttons or so-called quick replies to messages guide the user to the destination. The buttons can be suggested topics (e.g. reset password) or, in the case of a quick reply, a “yes” or “no”. Based on what the user selects, he receives an answer from the chatbot.
The user's freedom is limited, but by entering them via buttons, the chatbot can interpret the concerns more precisely. This results in a more meaningful answer. With good conversational design, the user can be quickly led to the right destination and made aware of further content. However, the specification of buttons limits the user's freedom and sends fewer or no text messages to the chatbot. This has the disadvantage that no new intents can be recorded. These are the strengths, weaknesses, opportunities, and risks of a click chatbot:
Checklist for the advantages and disadvantages of both types of chatbots
The following table shows the advantages and disadvantages of the two types of chatbots:
Chatbot types: Which should you choose?
It is impossible to say in general whether the rule-based click bot or the AI-based chatbot, which enables and understands free text input, is better. Because it always depends on which goal is to be achieved with the bot and what the company wants from the chatbot. A click bot is therefore ideal for simple and seasonal campaigns or marketing campaigns. Because the clear advantage of click bots is that they do not need to be trained and are much easier to set up than AI chatbots.
However, if companies want to use a permanent solution, e.g. in customer service or marketing, it could make more sense to spend a little more time and resources and implement an AI chatbot. Because thanks to many years of experience, we can definitely confirm one thing: The time and effort invested in setting up an AI chatbot initially pays off after a few weeks or months. Once the AI chatbot has been trained and the answer texts have been formulated, the chatbot develops independently thanks to AI and machine learning.
Nevertheless, it should be mentioned that an AI chatbot that allows free text input always carries the risk that user requests may not be understood in some cases. But there is also a strength in this weakness: Every question that is not understood is valuable for the chatbot, because it learns from it. You can also protect yourself by using so-called quick replies, i.e. buttons. So that even if the chatbot has not understood a request, it still offers the user further options, e.g. forwarding it directly to a human employee.
