NLU vs. ITR Chatbots... Which one should I use?


Even though there is no doubt about the importance of chatbots when it comes to tasks like acting as a customer support agent, HR tool or as a sales representative, there is still one question that causes some confusion: What type of chatbot should I use?

The answer to this question depends on a few factors, but most of the time it comes down to just two:

  • If you are interested in managing and creating a smooth user experience (CX)
  • The number of topics you want your chatbot to cover (scope)

NLU- Chatbot


An NLU (Natural Language Understanding) chatbot will facilitate a more open, easy-going, and natural conversation with the user. Additionally, it does not have the limitation of presenting a list of topics to the customer, meaning that the bot has an opportunity to grow in scope almost in an unlimited way and be trained to understand more topics, questions and actions.

On the other hand, an ITR (Interactive Text Response) chatbot is a good option when its functionality is limited to a few topics with one or two submenus. The user will browse through an option tree and their search will be limited to a set of predefined options that branch out of that specific option. Even though ITR chatbots are an effective option (there are no interpretation errors), one must keep in mind that this type of bots are also limited in scope and the browsing experience may seem unnatural which can lead to a poor customer experience.

So far, the answer to the question asked at the beginning seems obvious. Everyone would opt for an NLU chatbot that provides a variety of capabilities and allows growth in its functionality. Why isn’t this the obvious answer then? A little thing called “chatbot training”.




Chatbots that use NLU need to be trained so that the Artificial Intelligence understands the customer’s query in an open format, just like a human would. This process can be done with data that has been previously extracted from interactions with different users (live chats, social media, etc.) or with custom-made synthetic data based on the chatbot user’s profile.

The capability to training a virtual assistant or chatbot in a couple of days is something relatively new, but it has already been successfully implemented by some leading companies around the world. As a result, they have been saving millions of dollars in customer support while also increasing customer satisfaction.

The “future” is here to change the interactions you have with your customers. Ready to give it a try?

Contact Us


Subscribe Here!