Reslotted From Phones to Cleanup, Same Seat Rate Either Way
BPO agents are getting moved onto AI-monitoring queues where the work is fixing chatbot mistakes, at the pay of the phone role the bot replaced.
Your queue changed but your payslip didn't. One week you were taking calls. The next, you're staring at chatbot transcripts, catching the moment the bot promised a refund it can't issue or told a customer their order shipped when it never left the warehouse. Same headset, same floor, same seat rate.
The industry calls it a reslot. Management calls it upskilling. What it means in practice: the AI handles the easy tickets, and you handle everything the AI broke, at the wage set for the phone job the AI supposedly retired.
The bot gets the credit, you get the mess
Here's how the math works for the account. Deflection rates go up because the chatbot fields the first contact. On paper, that looks like fewer agents needed. In the room, it means the humans left behind absorb a harder mix. The routine questions are gone. What lands on your screen is the angry customer, the edge case, the transaction the bot mangled and can't unmangle.
The pay grade never moved to match. You're doing recovery work, resolving escalations, correcting a system that made confident promises with no authority behind them. That's more skill, more judgment, more emotional labor per ticket. The rate card reads like nothing happened.
Handle time still rules the floor
The old metrics didn't retire either. Average handle time, first-contact resolution, quality scores. Now they get applied to tickets the bot already fumbled, where half the call is you untangling what the customer was told before they reached a person.
Try explaining that a refund the chatbot approved was never actually possible. That conversation takes longer. It scores worse. The dashboard doesn't care that you inherited a problem someone else's software created.
Some accounts add a new layer on top: reviewing the AI's output before it goes out, flagging bad responses, feeding corrections back so the model improves. That's training the thing that shrank the headcount, folded into a shift with no extra line on the payslip.
The bargain that broke
The BPO pitch to a generation of Filipino workers was simple. English fluency plus a stable seat equals a real wage, night differential, an HMO, a path up. The tradeoff was the graveyard shift and the metrics grind. Most people took it because the numbers cleared rent.
Reslotting quietly rewrites the deal. The seat survives, but the work underneath it got harder while the rate held flat. Attrition does the rest of the cost-cutting, because people burn out and leave, and the account calls that headcount reduction instead of admitting the job got worse.
Ask about the pay adjustment for the new queue and the answer is usually that the role code didn't change. Same tier, same band. The chatbot took the volume. You took the wreckage. And the seat rate is exactly what it was the day before the bot logged in.