Conversational UI (CUI)
Synonyms: Chat UI, dialogue Interface, conversational interface, bot UI, natural language interface, CUI, VUI (voice user interface)
Definition
Use cases
- Handling ambiguity: When a bot misunderstands a request, users need a clear recovery path. Repeating “I didn’t understand” over and over usually causes abandonment.
- Context retention: Users expect the system to remember earlier inputs. If a travel bot forgets the destination halfway through booking, the experience breaks immediately.
How it's used in practice
- Define the voice early: Decide how the assistant should sound across onboarding, support, and error states so the experience feels consistent.
- Design recovery flows: Create fallback responses that help users move forward instead of blaming them for incorrect input.
- Prototype conversations before coding: Use Wizard of Oz testing to simulate conversations with a human operator before building the actual system.
- Maintain conversational memory: Preserve relevant user context across turns so people do not need to repeat information.
Challenges & limitations
- Open-ended input: Users can phrase the same request hundreds of different ways, making misunderstandings unavoidable.
- Low discoverability: Unlike graphical interfaces, conversational systems rarely show users what actions are possible, which can leave people unsure what to ask.
Free resources
- IBM Bot Design Patterns — Enterprise-grade conversational UX principles that scale beyond simple FAQ bots.
- AI UX Design Guide – Conversational UI Patterns — Real design patterns from ChatGPT, Slack, and Siri broken down for practical use.
- Department of Product – UX of Conversational Features — 20+ real-world conversational UX teardowns from Google, Adobe, Pinterest, and Figma.

