Chatbots Which chatbot to choose? Guide to different types of bots are increasingly present in our lives: they can order coffee in the morning for us, automate sending emails at work and play our favorite playlist as soon as possible. let’s go home.
It’s not science fiction: the examples I gave above are examples of chatbots that really exist and used by millions of people in the world. Starbucks, for example, created a bot to order their coffee quickly and intuitively.
See also : How to create a chatbot in python?
Starbucks bot for ordering coffee. Incredible, right? The world of virtual assistants is a young but rapidly growing world, and it is still unknown to most people (luckily we created a quick introduction to the topic).
Related topic : How to choose an agency to create a chatbot?
The youth of the market often urges us to think that it is enough to install a bot on the website to be ready to welcome our customers. Nothing more wrong: let’s remember that the technology is just one way to offer a certain service. We must therefore think first of all about the service we want to offer: customer support? Appointment booking? Legal advice?
We define how we will help our customers, and then we choose the type of chatbot that best suits each project. But what are these guys? Is there really a better type of chatbot than anyone else? Let’s look at the most used ones together and answer the questions step by step ✅
Plan de l'article
- Rules Chatbot vs Free Language Chatbot
- Personal Chatbots Vs. Team-Bots
- GDPR I authorize the processing of personal data E-mail Sign up for the newsletter Open-domain Vs Closed-domain chatbots
- Super-bot Vs. Specific-domain bots
- Voicebots Vs. Textbots
- And now?
Rules Chatbot vs Free Language Chatbot
The first distinction we can make is that between default rules and free input .
With a rules chatbot you can define a series of steps and guide the user with buttons or visual elements. A free language chatbot, on the other hand, allows the user to type what he wants and provides the most relevant answer: The most advanced solutions use artificial intelligence to analyze what the user writes, understand the meaning of the question, and give the correct answer.
Example of free text (left) and rule (right) chatbot. Which of the two types do you want to use? Obviously it depends on the purpose of your chatbot and the context in which it will operate, we cannot define it a priori. In Awhy we try to combine the two concepts: your customers can speak freely with the chatbot, and depending on what they ask for, our artificial intelligence answers with free text or guides them through a series of steps.
Benefits of a Policies Chatbot
- They don’t need language comprehension algorithms
- They limit the complexity of the conversation and reduce the occasions when the chatbot can’t respond
- They can drive the step-by-step user depending on a previously defined stream
Benefits of a Free Text Chatbot
- Dialogue is more natural, resulting in a better end-user experience
- The domain of conversation widens: if on the one hand it increases the risk of unmanaged questions, on the other hand, it allows you to understand what users are asking during interactions with your brand and improve with time
Try our chatbot!
Personal Chatbots Vs. Team-Bots
Another important distinction, this time more at the design level, is that between personal chatbots and team chatbots.
Personal chatbots have a focus on the individual user, converse with them and try to be a small, all-round assistant: the example of the chatbot to order the Starbucks coffee made above is a good example of a personal chatbot.
Instead, team chatbots are intended to facilitate a process or activity at the enterprise level. Remaining on the subject of food and drink, let’s imagine a bot that collates the orders for lunch of the entire team, perhaps on a slack channel: you’ll have to foresee more interactions from different users. Usually this type of bot involves a longer analysis phase, as conversations become more complex as the number of interlocutors increases.
An open-domain chatbot (literally open domain ) aims to entertain the user in a general conversation. This type of chatbot is very difficult to accomplish (almost impossible), because:
- It needs lots of training data before being put online
- Even with a lot of data available to us, it is impossible to predict where each conversation will lead.
- It’s very difficult to control how the chatbot will answer individual questions
A closed-domain chatbot (closed domain) instead operates in a well-defined context, as it could be that of the frequently asked questions you receive from your customers. In this case, the chatbot can be trained precisely enough to respond correctly to your users’ questions and improve the perception of your brand.
Super-bot Vs. Specific-domain bots
A super-bot is a chatbot that manages to expose other services within it. An example? Google and Amazon voice assistants. When we ask Google Assistant ‘Play My Favorite Playlist’, the assistant manages to play it using one of the music apps installed on our device (Spotify, for example). Spotify is therefore the service displayed within Google Assistant.
A specific-domain bot, on the other hand, is a chatbot that exposes exclusively a service: for example, a chatbot for the customer support of a given e-commerceOk Google, play the playlist for my workout. . This type of bot allows you to customize the content and intents that we will recognize, for a more vertical interaction on a certain type of service and brand.
Voicebots Vs. Textbots
The distinction here is very simple: voicebots integrate speech recognition technology, textbots work only with text.
In recent months, also thanks to the pandemic we are experiencing, speech recognition is an increasingly demanded feature: hybrid forms are being outlined, in which a textbot can be interfaced with voice systems to also allow this type of interaction.
Mixed systems are those that currently manage to fill the shortcomings of both types, in fact:
- Voice systems don’t respond well in noisy environments, and are not suitable for long, complex operations where the end user prefers to have everything in the eye;
- Text systems may be impossible to use in certain situations, such as driving or in the kitchen.
We have seen together the main types of chatbots available on the market: Keep them well in mind when you’re going to deploy your virtual assistant. To answer the initial question, there is no type of chatbot better than another. Each project has different needs: it will be your job to identify the type of chatbot that best suits them.
Finally, remember that there are endless facets in this world, facets impossible to describe in an article. Are you approaching a type of bot that I haven’t treated? Report it in the comments or send me an email, I’ll be happy to discuss it together!