The chatbot world is big, dynamic and innovative. To keep you up to date, we have put together a chatbot wiki that explains all the important terms briefly and concisely. Get ready to become a chatbot expert after reading this article!

1 What is a chatbot?

The word chatbot is a combination of the words chat and robot. Chatbots are computer programs that simulate a human conversation.

They use written text or spoken language. Chatbots provide answers to repetitive questions or can perform repetitive tasks. The word bot is often used as an abbreviation for chatbot.

2 What is a chatbot avatar?

The chatbot avatar is the face of the chatbot, so to speak. He represents the bot. Since a chatbot is still a computer or software and not a human being, the face of the chatbot should not be a real human image.

Chatbot avatars signal to the user that the chatbot is a bot and not a human being.

3. what is a good chatbot greeting?

Basically, the same rules apply when communicating with a chatbot as with a live chat. This also applies to the greeting. The success of the chatbot begins with the right greeting. It must therefore motivate the user to continue chatting with the bot.

Ideally, the greeting should arouse a certain curiosity in the user. The greeting should also be used to introduce the chatbot and create the right expectations for the user.

A very interesting study on the use of different chatbot greetings can be found here.

4 What is a good chatbot farewell?

There is no general rule for the farewell phrase of a chatbot. The same rules apply here as in a human conversation.

The chatbot should say goodbye accordingly. Depending on the chatbot’s goals, the bot should end with a call-to-action or collect feedback from the user.

5 How does a chatbot comply with the GDPR?

In principle, the same data protection guidelines apply to chatbots as to websites, newsletters and other media. First and foremost, it must be checked whether and which personal data is collected.

It must then be ensured that the users have actively given their consent to the processing of this data. Depending on the type and amount of data collected in the chatbot, the classic cookie banner may already be sufficient. In very complex cases, however, consent directly in the chat window is mandatory.

In the podcast episode by Sophie Hundertmark and Marcel Griesinger, you can even listen to the topic of chatbots and data protection.

6 Which chatbot tools are available?

There are countless chatbot tools. Most of them differ in price, functions, data protection requirements and compatibility with other software programs that the company already uses (interfaces).

Companies should select the right chatbot tool depending on their requirements in terms of data protection, functions and interfaces.

If possible, the chatbot concept is created first (see the chatbot canvas), and then a suitable chatbot tool is selected.

7. can I combine chatbots and live chat?

Yes, in this case we are talking about hybrid bots. A hybrid bot combines the advantages of chatbots with those of a live chat and at the same time reduces the disadvantages of both.

Technically speaking, a hybrid bot is a chatbot that breaks off at a certain point in the dialog and switches to a live chat, i.e. a chat with a real person.

For the user, switching between chatbot and human should take place without any major hurdles. The user remains in the same chat window. There is therefore no channel change.

The human agent receives the information from the chatbot conversation in the background so that it does not have to ask for it again. More about hybrid chats here.

8 What is a Facebook Messenger chatbot?

Chatbots are basically automated instant messaging software that automatically responds to user queries.

However, when we talk about Facebook Messenger bots, we are referring exclusively to chatbots that are integrated into Facebook Messenger. Some companies that are represented on Facebook have an automated start message in Messenger.

With a Messenger bot, all communication with the target group can now be automated. Find out more about the possible uses and benefits of Facebook Messenger bots here.

9 What is a WhatsApp chatbot?

WhatsApp is one of the best-known messenger platforms in the world. As a rule, it is people who exchange information or small talk with each other here. However, it is also possible to create chatbot profiles so that a person chats with a fictitious person, i.e. a chatbot. In this case, we are talking about WhatsApp chatbots, or WhatsApp bots for short.

There are a few special features to bear in mind with WhatsApp bots. WhatsApp does not yet allow buttons to be integrated into the conversation. Users can therefore only reply using text messages or emojis.

Furthermore, companies do not have the opportunity to contact their target group without being asked. In WhatsApp conversations, the user must always ask the chatbot a question before it can respond.

10 What is Conversational AI?

Conversational AI is the term used to describe AI-supported, automated dialog systems. This is a form of artificial intelligence (AI) that enables users to engage in automated, natural-language dialog via chatbots or voice assistants.

11 What is Natural Language Processing?

Natural Language Processing (NLP) is the process of language processing. This includes, for example, classifying texts into categories, language recognition and automatic translation as well as the automated summarization and generation of texts.

Actions are executed based on the processed text. NLP therefore plays a particularly important role in AI-based chatbots. NLP enables free text input from users to be understood and answered correctly by the chatbot.

12 What is conversational marketing?

Conversational marketing means direct and individual communication between brand and customer in real time. This is usually done via a chat interface. Companies are in direct dialog with their target group and thus transmit their marketing messages.

Thanks to chatbots, companies can automate their conversational marketing. Conversational marketing only became popular thanks to chatbots.

13 How much do chatbots cost?

In order to answer the question of the costs of a chatbot, it is first important to know and understand the individual cost drivers. A chatbot project basically comprises the following budget items:

• Concept creation

Depending on the scope and use case of the chatbot, the chatbot concept can take a lot of time. In some cases, target groups are even tested here. You can read more about the chatbot concept and prices here.

• With or without AI – rule-based vs. ai-based

A chatbot with AI (Artificial Intelligence) generally incurs significantly higher costs than a simple rule-based chatbot. An AI-based chatbot has to be “trained”, so to speak. Intents must be created.

This means: The intentions that a user may have must be defined, and then example sentences and answers must be written for each intention. With a rule-based chatbot, only a type of dialog tree is created.

• Interfaces

If a chatbot requires interfaces, i.e. if connections to third-party systems are required, the entire chatbot project can quickly become complex. Depending on the interface, programming can quickly take a week.

• Multilingualism

Chatbots that are to exist in several languages usually take longer to implement than if the chatbot only works in one language. It is often advisable to start with just one language and add other languages later.

• Data security

Where should the chatbot be hosted? Depending on the hosting requirements, additional costs may also be incurred here.

• SaaS tool or in-house development

There are now many good SaaS (Software as a Service) providers that offer platforms with which companies can create a chatbot without much development effort.

If this is not an option for the chatbot project, you will have to resort to in-house development. As a rule, in-house development is more expensive than using a SaaS platform.

More about chatbot costs here.

14 What is the chatbot engine?

The chatbot engine only comes into play with intelligent chatbots. In so-called AI-based chatbots, the chatbot engine is the central control unit and performs the following important tasks in particular.

• Send and receive messages.

The engine receives messages via the chatbot’s input field and sends messages to the user at the end. It forms the interface between the user and the chatbot.

• Process messages

In the engine, you decide what to do with the message received from the user and how to reply to the message.

• Manage the meetings.

The engine ensures that the right users receive the right answers or that the desired actions are carried out. It is also possible that the engine must first identify the user in order to react correctly or possibly not allow certain actions at all.

• Remember context.

Good AI-based chatbots are able to respond differently depending on the context. The engine remembers the user’s inputs. If possible, she refers to this in the further conversation or uses the previous information for a specific answer.

• Mediate live chat.

In some cases, the chatbot is directly connected to a live chat. In this case, it is up to the engine to decide when the conversation should be handed over to the human agent. The engine must also ensure a smooth transfer so that the user does not notice any media discontinuity.

15 Where can I find chatbot templates?

A good chatbot must be precisely tailored to the company and the target group. Chatbot templates can only help to a limited extent here.

By using ready-made chatbot templates, the individual bots lose their individuality and originality. However, these two factors are crucial for the success of the bot.

It is therefore advisable to avoid fixed templates and instead use templates such as this chatbot canvas. The chatbot canvas helps you go through all the important steps from the idea to the finished chatbot.

However, if you still want to be inspired by templates, you can read this article.