I, Chatbot. The emergence of AI as customer service

If you’ve ever used a chat feature to get support from a website or application it is possible you were actually talking to a robot.  Chatbots as they’re called, are the latest trend in customer service.  They combine artificial intelligence, programmed responses and rules sets to allow the user to interact, via chat, with a service chatbot to help them do simple tasks such as answer frequently asked questions, order a product, or make a reservation.

The last time you called your bank, insurance company or internet service provider you likely went through a phone tree.  A human voice answers asking you to, “press 1” to continue in English or to respond verbally with the nature of your call for example, “auto claims.”

Chatbots are almost identical to a phone tree; however, when using a chat feature on a computer or application the experience is different.  It is clear when in a phone tree that you are not speaking to a human, and it can be frustrating.  A chatbot can give the impression of intelligence, making the user’s experience more enjoyable.

Here is an example of what the start of a chatbot transcript might look like:


@MikeR:  Hi, I am having trouble uploading a photo.  Can you help me?

Chatbot:  Good afternoon, Mike.  I would be happy to help you.  So you are having trouble uploading a photo, is that correct?

@MikeR: Yes.  It keeps almost completing the upload and then stopping.

Chatbot:  I am sorry to hear that, Mike. I need some information from you before we get started solving your issue.  Can you provide me with the type of browser you are using?


The Chatbot will gather the appropriate information and suggest solutions for the user.  At the same time it is collecting data about user behaviors and logging any potential issues that need to be addressed.

Well written pre-programmed responses are what make it seem as if you are interacting with a real person.  Answering frequently asked questions this way enables the user to feel more engaged and chatbots can operate around the clock.

What is important if you are a global company using this technology is to ensure that the responses in all languages are not robotic. You may have perfectly written interaction text in your native language but that’s no guarantee it will read as smoothly after translation.

Transcreation (editing source content for readability in another language) and localization (using terminology and sentence structure that read as if written by a native speaker) are key to ensuring that the user experience is positive.

Interested in getting your own chatbot?  It’s not as hard as it may seem.  A lot of the work has already been done.  There are apps that you can install and developers ready to get you started.  Here are a few resources that may be useful:

Once you’re ready to go global, don’t forget us!  Transfluent can help by localizing your chatbot’s interactions in any language.  The best chatbots have responses that sound human and we are human-powered translation at its best.

Image: Robot by Robin Zebrowski, licensed under a CC by 2.0