Blocking Abusive Visitors Without Ending the Conversation for Everyone Else
Chat's visitor bans work by email or visitor ID, enforced when a conversation starts or resumes — a small, necessary feature most comparison content never mentions.
Any tool that lets an anonymous stranger start a conversation with your business needs a way to stop a specific person from doing that again. It's not a glamorous feature, and it rarely shows up in comparison articles between live-chat products — which is itself telling. Buyers assume it exists and move on; it's worth actually explaining once.
How a ban is identified
Chat bans work against one of two identifiers: a guestEmail, if the visitor provided one, or a visitorId, the identifier the widget assigns automatically to anyone who starts a conversation, logged in or not. That second option matters because plenty of abusive visitors never give a real email in the first place — banning by visitor ID means you're not dependent on having collected contact information from someone you're trying to block.
When the ban actually takes effect
The ban is enforced at conversation-creation or conversation-resume time, inside the /widget/start flow. In practice, this means a banned visitor is stopped before a new conversation opens or an existing one resumes — not mid-conversation. If someone is actively chatting when you ban them, the ban takes hold the next time they try to start or reopen a conversation, not by ending the one currently open. That's a real, specific boundary worth knowing rather than assuming the system does something more dramatic in the moment.
Why scoping the ban to the individual, not the widget, matters
A ban targets one identifier, not the widget as a whole — every other visitor keeps chatting normally. There's no global "pause the widget" switch bundled into this, and there doesn't need to be: the mechanism is specifically about removing one bad actor's ability to keep starting conversations, not about your support channel as a whole.
When to reach for this
Use a ban for genuine abuse — harassment, spam, someone deliberately wasting agent time — not for a visitor who's simply upset. A frustrated customer is exactly who a support inbox exists for; a ban is for the narrower case where the conversation itself has stopped being one worth having. If in doubt, it's cheap to check the conversation history before banning: the ban only blocks future conversations, so nothing about the visitor's past history changes if you decide not to use it.