Repeat-Ticket Auto-Resolution Agent
The Manual Process Today
Customer files a ticket for a known-answer question instead of self-serving.
It lands in the agent queue like any other ticket.
An agent realizes it's the same question they've answered before.
They write essentially the same reply again by hand.
Resolved, and the answer disappears into the thread.
The next customer files the same question tomorrow.
Working Assumptions
| Assumption | Estimated | |
|---|---|---|
| Est. Tickets / day | ≈ 150,000 / year | |
| Repeat-Ticket Share | % | ≈ 37,500 repeat tickets / yr (pool) |
| Already Deflected by KB (#2) | % | ≈ 18,750 still reach an agent |
| Handle Time | min | per repeat ticket |
| Avg. Support Agent Wage | $/hr | per agent hour |
Illustrative estimates, not fixed numbers. Edit any value to recompute the pain-point figures below and the Impact Analysis tab, or tap the ⓘ on a row for its rationale.
Note: any edits are saved in your browser only, so different viewers may see different numbers. Use Reset to defaults to restore the original values.
Pain Point
About ~25% of ~150,000 tickets/yr are repeat / known-answer questions (~37,500); after #2 deflects ~50% pre-ticket, ~18,750 still reach an agent.
At ~10 min each, re-answering those ~18,750 repeats by hand is ~3,100 agent hrs/yr.
At ~$20/hr those ~3,100 hrs are ~$63k/yr of agent capacity spent re-answering solved questions, the pool an auto-resolution agent can attack.
Known-answer volume competes with the complex, novel tickets that actually need a person, so the hardest problems wait behind the easiest.