Detractor Hour
Score a 3 and a voice calls within the hour, when calling helps.
The job
When a customer scores us 0-6, predict whether a call can recover them, reach the recoverable ones by voice in their own language within 60 minutes, fix what policy allows, and prove which causes a call actually recovers against a control group, so callback capacity and money go where calls work and the rest goes to a real fix.
The moment
A customer scores a survey 2 with a real grievance about a missed appointment. The hourglass appears with its prediction, call will recover: 64%, and 24 minutes later the customer's phone rings: a voice agent quotes the comment back and offers a new slot. Then the board shows a coverage detractor from the same hour with no call at all, stamped call-proof: 8% recoverable, routed to network planning. The Recovery Ledger lands: Called within the hour 47% recovered (42-52, n=1,930); standard email 18% (13-24); prediction hit rate 81%; $3.1M retained this quarter. The Head of CX says it out loud: We've been calling the wrong detractors.
What it does
- Edit the guardrails the agent obeys
- Call now for a missed recoverable detractor
- Hand to a human
- Stamp a cause call-proof and send tell-when-fixed
- Reopen a failed recovery
- Drill any cell to verbatims
What you see
An hourglass rail of today's detractors draining to 60 minutes, each carrying a live recoverability prediction that is graded on screen, a recovery league by cause beside the control group, and a retained-revenue ledger. Ready for the wall screen with masked IDs.
What it moves
Detractors recovered by cause vs the control group (with n and bands), prediction hit rate, and retained revenue in dollars per month
Built for
- CX leaders & VoC
- Contact-centre operations
- Analysts & data
- Finance
