Based on the functionalities and nature of usage, I divided chatbot into 6 types. Also, I added an example at the end of every type as a reference to consider.
Chatbot: A chatbot is a computer program that interacts with humans, providing intelligent answers. It works on a chain of commands and tries to replicate the human touch during these interactions.
For Example, when you are searching online for help about investing in mutual funds, the chatbot will assist you by delivering answers whilst adopting the tone of a customer support agent.
Talking about the future of chatbots, CEO of Manychat Mikael Yang has something to say: According to him, 80% of B2C communication will take place via bot messenger within the next 3 to 5 years.
Based on the functionalities and nature of usage, I divided chatbots into 6 types. Also, I added an example at the end of every type as a reference to consider.
- Menu/Button based chatbot
- Keyword recognition based chatbot
- Contextual chatbot
- Voice-based chatbot
- Service chatbot
- Social messaging chatbot
1. Menu/Button based chatbot
A menu or button based chatbot is the simplest type of chatbot. There is a pre-built knowledge base available and you just have to select from the predefined buttons.
It is quite possible that your query won’t be listed and the bot, therefore, can’t help as it has no capacity to go beyond the listed button or menu. In other words, you can consider it as a “scripted chatbot”, constrained to certain question-answers.
For Example, Poynt bot provides buttons for users to find solutions to certain queries. Furthermore, when you click a particular button, a pre-built decision tree appears, and after some more clicking, you get the answer.
2. Keyword recognition based chatbot
This type of chatbot is more advanced than the previous one. It uses NLP (Natural Language Processing) — an AI application that helps better serve users.
In terms of working methodology, when users type a question and hit enter, the bot analyses the question with NLP, matching against keywords to deliver the correct answer.
For Example, the Poynt device uses NLP to answer user’s questions.
3. Contextual Chatbot
This chatbot is much more advanced than the button and keyword types. The reason is an exciting combo of AI (Artificial Intelligence) + ML (Machine Learning)
When a user types their query, the bot is trying to learn the intent behind the search instead of simply jumping to a predetermined answer. It will store up unique searches from different types of users.
In future similar instances with users, the bot will retrieve previous interactions and answer questions without the need to probe deeper and deeper into the situation.
This saves users time. Instead of 5 to 6 questions, within 1 question, all of the necessary information is shared by bot.
Explore the below example where “Good Bot” identifies the user’s need and displays the entire user order. They just have to approve it.
4. Voice-based chatbot
As the name suggests, these bots work with the user’s voice as an input, providing a solution after AI-assisted interpretation.
For Example, Robot Vera is a recruitment bot that works on speech recognition and is capable of conducting interviews with thousands of candidates simultaneously.
5. Service chatbot
This type of bot provides necessary information after asking certain questions regarding users’ service requests. It is more prevalent in the service industry, such as airlines.
For Example, KLM Airlines has a messenger bot called BB (Blue Bot) that allows users to book their tickets through its handy interface.
6. Social messaging chatbot
This type of bot can be integrated with social media platforms, enabling users to ask questions and get answers, without even leaving the platform. The main benefits are around time-saving and reducing the effort required from users.
For example, Whole Foods has an active Facebook messenger bot to provide users with information about cooking tips and new recipes.
These are my chatbot types. If you are aware of any other type apart from this, then do share your thoughts with the community in the comment section.