What does a conversation designer do

At a high level, a conversation designer defines and creates compelling and delightful conversational AI experiences, such as text-based chatbots and voice assistants.

For those who are familiar with the role of a UX designer, there are a number of similarities, although some aspects are specific to conversation design.

The conversation designer's involvement in the software design and development process is as follows:

  • Use case definition. Reviewer / advisor. The high level use case is defined by the product management team and is informed by user insights, business objectives and company direction.

  • User needs. Executor / advisor. Conducting user research and desk research to gain insights into the target users, their wants and needs to use as input to the conversation structure and prompt wording. In larger organizations with dedicated UX researchers, this research can be lead by UX researchers in collaboration with the conversation designer.

  • Assistant personality. Executor. Defining the assistant personality based on existing brand guidelines and brand values is an important part of conversation design. Larger organizations may have a content team that can act as an advisor.

  • Conversation structure. Executor. Defining and building the flow of the conversation.

  • Wording and sample dialogs. Executor / advisor. Prompt wording and overall . In larger organizations, a team of content professionals can take on this task. The final prompt wording generally needs review from the legal department, similar to content on the website or other digital channels.

  • Natural language understanding (NLU). Executor / advisor. Defining intents, gathering sample utterances and training the intents to build the NLU model for natural language in put. In larger organizations, a data scientist/ML engineer/AI trainer may be available to define intents and entities and train them. Given that the conversation designer uses the intents in the conversation, a close collaboration is recommended.

  • Usability testing using prototypes and beyond. Executor. Gathering user feedback iteratively to then improve the current design.

  • Advocate for conversation design and best practices to create engaging conversational experiences. Executor. Spreading awareness around conversational AI and building conversational AI products, since conversational AI and conversation design are still new to many.

Soft skills are important. Ideally, a conversation designer is a:

  • great collaborator: working with the product team, technical team and possibly third parties

  • great communicator: communicate ideas clearly, evangelize and explain what you do as a conversation designer. Exercise patience and listen to others

  • organized individual: needs to be able to manage and schedule their workload within the larger effort

  • understands the big picture: while the conversation designer's prime directive is to create compelling experiences for the user, the conversation designer must at all times consider the entire multi-channel customer experience and the strategic direction of the company

A successful conversation designer:

  • understands and has experience defining and designing digital products using a comprehensive approach including user-centered design and attention to company goals and growth

  • understands AI and how building AI products is different from building other digital experiences such as web or mobile applications

  • has a background in one or more of the following: UX, (computational) linguistics, psychology/cognitive science, writing

  • has passion for the field of conversational AI

Last updated