Chat can support customer service, teamwork, or automated assistance. A good chat experience communicates who or what is responding, whether a reply is expected, and what happens when nobody is available.
A chat window is like a conversation at a help desk: people need to know who they are speaking with and whether the conversation is still active.
Set expectations about response
Tell people whether they are messaging a person, a team, or an automated system. If replies are not immediate, show realistic availability or response-time information. Do not use a live typing indicator or online status unless it reflects the system’s actual state.
Keep conversation context
Messages should remain in a readable order, show useful sender and time cues, and preserve context when someone returns. For long conversations, provide ways to find recent or important messages. Avoid making a user repeat details that are already present and safe to reuse.
Design for interruptions
People may switch tabs, lose connectivity, or receive a delayed response. Save drafts where appropriate, indicate sending and failure states, and avoid duplicate messages after a retry. Provide a route to another support channel when chat cannot resolve the issue.
Protect people and information
Make it clear what information is stored and who can read it. Avoid asking for secrets such as passwords in a chat. Provide reporting or blocking tools where users interact with one another, and ensure notifications do not expose sensitive message content on a shared device.
Frequently asked questions
Key takeaways
- Identify the responder and set honest timing expectations.
- Preserve message order, context, and recoverable drafts.
- Show network and delivery states to prevent lost or duplicate messages.
- Protect personal information and provide suitable escalation paths.