February 19, 2019
Conversational user experience (UX) is a user experience that mimics an actual human conversion as closely as possible. Essentially, the user chats to a machine that learns what you want based on what you’re saying, adapting its tone until it finally delivers what you want. In other words, conversational UX isn’t just about deploying a chatbot that addresses a customer by their first name or plucking out a recurring customer’s order history; it goes beyond that to add a layer of human connection between your brand and your customer. The process involves combining chat, voice, or any other natural language-based technology.
Creating a great conversational UX may sound like a mammoth task, but it doesn’t have to be. Here are some ideas on how to get started with conversational UX:
Control the conversational flow
The first thing you need to do is craft the narrative before even thinking about building the machine. Why? Because the best conversational UX creates the illusion of a free-flowing and natural conversation. In reality, the dialogue is meticulously scripted in a manner that controls the flow of the narrative.
Planning the conversational narrative involves the following steps:
- Identifying spots in the script where specific questions are likely to be asked;
- Deciding how these questions will be answered by the machine; and
- Strategically placing human errors within the script such as typos (believe it or not, typos can help create a strong conversational UX!)
The next step involves creating a flowchart of the conversational narrative. Marketing automation software can help you design the flowchart as you go; a solution like Autopilot will even allow you to create powerful and interactive mockups of your conversational narrative and share them with your team for feedback and revision.
It’s also important to determine where the responses to the above questions will reside and how that knowledge will be organized. Will the machine use an API or will the answers will hard-coded? If someone asks a difficult question, will they be transferred to a live person or will there be a trigger for an email to be sent?
Add machine learning/AI
Conversational UX doesn’t necessarily have to include machine learning and AI, but it would certainly help. After all, AI is all about narrowing the funnel of interactions that need human intervention. When AI is machine learning-based, it can expand its capabilities over time; as a result, it can learn how to “understand” what a customer is saying and provide a more natural, conversational and satisfying experience. Implementing AI into your conversational UX involves configuring your machine to pick up your brand’s particular nuances.
Don’t set and forget
Unfortunately, conversational UX is not static; it requires constant monitoring. When executing a conversational UX strategy, it’s important to analyze its performance over time. Most importantly, the program must adapt to external factors that are like to change constantly, such as the target audience it’s designed to serve or the behaviors of your customer segment. Autopilot can help you capture all that data, making it easy for you to adjust your tactics accordingly.