Zero-Touch Resolution: Definition, Measurement and Examples
The word doing the work is any. A conversation the bot appeared to finish, which quietly created a ticket for someone in operations to action on Monday, is not zero-touch. Neither is one an agent silently monitored. Neither is one where a supervisor approved a refund the bot promised. Most automation dashboards count all three as automated, which is how a 60% "automated resolution" figure survives contact with a headcount that never went down.
Measure it against work completed, not against tickets avoided. A password reset that finished in the conversation is zero-touch. A customer redirected to a help-centre article is deflection, and says nothing about whether the password got reset.
Published enterprise figures sit lower than containment numbers by a wide margin, which is the point. They are counting a stricter event. Start with three or four high-volume intents where the system has write access to the
underlying record: order status, EMI due date, KYC document upload, service appointment booking. Read-only integrations cap zero-touch resolution permanently, because the bot can describe the problem but not end it.
Then watch the reopen rate on those intents specifically. Zero-touch resolution that comes back in 48 hours is just false resolution with a better name.
Often confused with: Containment counts sessions that stayed inside the bot. Zero-touch resolution counts problems that ended there. The gap between the two numbers is the honest measure of how much work automation is actually doing.