Chat bots are the current hype and everyone seems to be searching for use cases to introduce chat bots in their applications. We’ve been working on creating a learning chat bot and the lessons learnt prompted me to write about how do we build better bots (with the current technology).
The reason we see a lot of frustration around using bots is that in a hurry to build them, a lot of times, a bad user experience is created. These outcomes are avoidable with a little ‘design thinking’.
Let’s look at some of the common use cases around chat bots to begin with.
What are chat bots being used for?
1. Customer Support
Gartner predicts that 85% of customer interactions will be driven by Artificial intelligence (primarily chat bots) by 2020. With the advent of mobile devices, where the most important use case is messaging, the natural progression to chat bots will bring forth an immense opportunity for intelligent virtual agents that can act as a first point of interaction with the customer and if needed, seamlessly, pass the baton to a human agent. They can help answer basic questions or help a customer make a buying decisions. Apart from bringing down the costs, chat bots can help improve the customer experience by reducing the wait times and providing the most up-to-date information as and when it appears.
There is a wide gap between the demand and supply of quality teachers in the world. United Nations estimates that there will be a need for 69 million new teachers by 2030. It might not be humanly possible to come up with that many good quality teachers and that is where chat bots come into the picture, they can be that teaching assistant the teachers never had. We all know about Georgia Tech’s Professor Ashok Goel who had use an AI based teaching assistant to scale personalized learning. Chat bots are now helping students learn new languages and helping students review their coursework and help improve outcomes.
3. Sales lead generation
Online ads have been the primary source of online lead generation, now imagine if users can interact with your online ads – That is your lead generation chat bot. It can qualify the lead in a much better way by engaging the customer at the right time and help them take a buying decision by transferring to your customer support virtual agent or a human agent by prioritizing as needed.
4. Data collections
Filling forms is boring, chat bots can help. Conversational forms can freshen things up while filling information in web forms. An intuitive UI apart from textual interaction can be very useful in this case.
5. Personal assistant
Siri, Google Now, Alexa and Cortana come to mind. These are keyword triggered bots that help you manage your daily tasks like setting up a reminder, create a list, call or text someone or summon an Uber. For the most part, they depend on searches to provide an ‘answer’. They are increasingly evolving into niche areas like health and shopping. A chat bot can help you eat healthy and do your grocery for you from the list it managed.
Chat bots can help you at your workplace. The pesky HR policy information can be better served by a bot that can keep itself up to date of the frequent changes. There are slackbots that we already use than can trigger a build or a deployment in the cloud as continuous integration becomes mainstream. A career coach bot is another good example which can provide valuable insights into your career.
New use cases will continue to emerge as we progress in our understanding of the areas where chat bots can be useful. Till then, we must experiment.
In the next part, I’ll discuss about some common pitfalls to avoid when building bots.
[About the author: Rishi Arora works as an Analytics Architect at IBM India Pvt Ltd and is currently working on building a Virtual Tutor.
This article represents my personal views; I do not represent the view of my employer (IBM) in any way. This article was originally published on Linkedin.]