“Chatbots” provide a new kind of user interface to computer software. Rather than navigating with menus and buttons on a mobile or desktop device, a user has a “conversation” with a bot. Conversational user experience (UX) is the science and craft of designing experiences that are engaging and satisfying for users, and ultimately help them get what they need done as quickly and efficiently as possible. Get it wrong at your peril – your exciting and possibly expensive Chatbot will fail in its mission if it doesn’t engage your audience!

Designing conversations is hard. That said, having conversations with real humans is hard! Humans don’t usually enter into conversations with a script in hand. Even if they have a script in hand, the conversation is unlikely to go far before they are well off-script and making it up as they go along. Real conversation involves offering up information, probing for clarification and new information, making demands, and trying to influence the perspective of the other person. Thousands of books on how to sell stuff, lead people better, get on better with people, and get your point across are a testament to how tough it is to explain yourself and understand others, without any software involved!

Conversational user experience

While it’s unlikely that most chatbots need to accommodate having their outlook on life altered in a conversation, the business of collecting information, teasing out clarity, and responding to demands is fundamental to conversational UX. The bit of magic that makes it easier for humans to have conversations with each other is context. No idea or interaction exists in a vacuum. People enter conversations with a great big bag of knowledge of what has happened in the past, what might happen in the future, and they remember the outcomes of previous conversations. All of this context shapes how they interpret what others are saying to them, and helps them choose the best way to explain what they want to say.

context is king

Chatbots are no different, without context they provide deeply unsatisfying experiences. If you’ve already had a long conversation with a chatbot and return the next day with new information, the last thing you want to do is be greeted as a stranger and be forced to navigate the same series of questions to get back to the point where you can use that information. If you’re already a customer of a business, then you should be able to expect that a chatbot will use the knowledge of which products you have purchased in the past to figure out what you might have a question about.

At Ambit, context underpins the intelligent conversations we deliver for our clients. Our platform has an open, “pluggable”, architecture that makes it easy to integrate with existing business systems, where all that magic context is hiding. As conversations happen on our platform, we gather and interpret everything the user is saying, and use this to personalise current and future interactions. Follow our LinkedIn page here.


By Gareth Cronin, CTO at Ambit


Leave a Reply

Your email address will not be published. Required fields are marked *