Could a conversation with a computer change how we engage?March 14, 2017
Conversational User Interfaces (CUIs), more commonly known as voice recognition tools or chatbots, have big implications for engagement professionals, and we should be excited – very excited.
The question on my lips is not whether the rise of conversational interfaces can help us to engage, but how we can best use them to engage in meaningful ways.
Who doesn’t love the idea of being able to speak to a computer or robot in ways that you would a human? Having an actual conversation that is understood and responded to in human terms.
It’s the common language of general conversation that make this form of artificial intelligence so appealing.
Unlike Graphic User Interfaces (GUI), which require humans to learn the computer’s language by clicking on buttons and navigating around pages, conversation interfaces take very little effort from a human’s point of view, you just need to type or speak.
When engaging with people online via a survey or interactive tool we are seeking responses to set questions or feedback on particular topics. People respond in straightforward ways by clicking, commenting and sharing. This can involve personal choice which leads to a set outcome, like with Avatar Surveys, but there is no conversation involved.
With CUIs this idea of set responses can be flipped on its head. The user can steer the direction and provide input in ways we’d not imagined. The style of language used can be more engaging and conversational, making people feel they are being spoken to one-on-one.
Think of a Choose Your Own Adventure Book, which has several predetermined courses, but the reader chooses which course to take and ultimately their fate. Similarly with CUIs, the user steers the conversation, and we learn a whole lot more about that user because of their choices.
I’m thinking of an interactive survey, which is designed like a conversation. We might ask for example what a user’s favourite colour is. The user says pink, so we then ask if they’d ever choose to buy a pink car. The user says no, but we are keen to hear why if their favourite colour is pink, why wouldn’t they want a pink car? The conversation eventually concludes that while the user’s favourite colour is pink, this doesn’t necessarily influence their choices.
The difficulty in this type of engagement is designing the tool that collates data that has purpose, but is also conversational in nature. A challenge is also being able to collectively analyse the responses so they make statistical sense and to gather data on a larger scale.
These challenges will become more prevalent in the future, especially as CUIs become easier to design and implement.