The popularity of chat is growing at a rapid rate, and businesses are finding it hard to ignore. According to a study by Betaworks, messaging apps have now surpassed social networks in popularity and usage, and the trend is not slowing.
The potential to reach customers through chat is massive and companies like Facebook and Kik are investing heavily in not only chat platforms but AI frameworks that allow them to respond in more human ways. With more than a billion combined users, applications built on the combined Facebook and Kik platforms have incredible reach.
In addition to AI, investments in Cognitive Services such as Microsoft’s Language Understanding platform LUIS, developers now have access to powerful entities known as “Domains”. These domains contain language and knowledge specific to industries that allow chat bots instant access to terminology thus providing valuable business context to every interaction.
There is no doubt that with chat comes tremendous business value but how do you design something that has no UI? How do we align our design process to this new method of interaction to ensure we are getting the most out of our chat applications? The answer is Conversational UX.
Conversational UX studies how users interact with chat systems and looks for ways to introduce enhanced functionality and new features that address business objectives while dealing with the unique challenges of the chat interface. We are able to achieve this by integrating new tools with some of the techniques uses in a typical User Experience Design engagement. In many cases the classic deliverables are still extremely powerful but for very different reasons.
At the center of good UX design is the persona. The guiding principle of User Centered design is that we must understand our users goals and design outward. Conversational UX is no different. By starting with well crafted profiles of our target audiences we are able to create powerful design strategies that are tailored for the mostly text based interactions of the chat platform.
With personas in hand we can get to the work of defining our Conversation Flows. The Conversation Flow is a visual map of queries and responses between the users and our chat bot and allow us to better understand how visitors will reach a goal, such as “order confirmation” for example. However, the conversation flow is about much more than just linear conversation mapping
In an environment where we cannot count on flashy UI to attract the attention of users we must find other ways to give them feature hints or guide them to new services. Conversation Flow diagrams are used to plot types of interactions and arrange them in priority based on business importance, frequency of use and complexity.
Quite often we are able to use a common interaction to gradually guide users to other services they may not have considered. For example, a common usage of chat is often to check the status of an order. While most customers would think of this immediately they may not have considered the idea of chat as a “Personal Shopper”. As designers, our goal is to find ways to introduce new ideas during common interactions allowing us to surface features that surprise and delight customers.
Once we have started mapping our conversation flows we can begin to build and test through simulations. These simulated interactions allow us to quickly test our conversation structures as well as the “Microcopy” we have chosen for each response. Is our system utilitarian or conversational in tone? Does the way we are using language reflect the style we want, make the user interaction feel natural and encourage comfortable and engaging interactions?
Besides style and tone there are also some very practical considerations to building an AI driven chat bot.
- How do we handle gender-specific pronouns?
- How do we gracefully handle unsupported responses and topics?
- When do we inject suggestions and help?
- How many versions of responses to we create to enhance variety?
- When do we resort to buttons versus text?
- How will be validate user input?
All of these questions are vitally important and are tested through conversation flow and documented for copywriters.
It’s exploding popularity makes a strong business case for chat and the advances in AI, Natural Language Processing(NLP) and Cognitive Services allow for quick build times. Even so, the design considerations are every bit as important. Before building, businesses must clearly define the objectives and focus development efforts on delivery of great core features with an eye on expansion over time. Conversational UX is the key to exceptional chat experiences.