CX Economics · Attrition
Agent attrition is a P&L line item. Almost nobody puts it under that heading.
Take an agent who leaves at month seven.
You paid to find them. You paid a trainer and weeks of floor time while they produced nothing. You paid for their first two months of below-average handling and below-average resolution. Then you paid again to replace them.
None of that shows up as a single number anywhere. It is scattered across recruiting spend, training budget, and a resolution rate that looks like a performance problem instead of what it actually is: the cost of turnover, wearing a different name.
Why the number stays invisible
Each piece of this cost sits in a different owner's budget. Recruiting owns the sourcing cost. L&D owns the training cost. Operations owns the dip in quality during ramp-up, and usually just absorbs it into "normal variance." Nobody owns the sum, so nobody presents the sum, so the board never sees a number large enough to justify fixing the actual problem.
Now multiply that per-departure cost by your real exit rate, not the one from the slide deck. That number is usually big enough to fund the fix that would have prevented it in the first place.
Why I won't hand you a benchmark
Every attrition benchmark I have found is American, published by a vendor with something to sell, and repeated unchanged for years. None of them reflect your market, your wage bands, your ramp curve, or your specific mix of channels. Using someone else's number here does not save you time. It gives you a number you cannot defend in the room where it matters.
Build your own. It is an afternoon of work with your finance lead, not a research project.
What actually goes into the number
Four components, all specific to your operation, none of them exotic:
- Sourcing cost: recruiting time and any agency or platform fees per hire.
- Ramp cost: trainer time plus floor hours during which the new agent produces below a fully ramped agent's output.
- Below-average performance cost: the gap between a new agent's handling and resolution numbers and your steady-state average, for as long as that gap lasts, usually the first one to two months.
- Replacement cost: the same sourcing and ramp cost again, for whoever backfills the seat.
Multiply the total by your actual monthly exit count, not a projected or industry-average one. That is your real, defensible attrition cost, specific to your operation.
Do it before the number gets buried again
The right time to run this calculation is before your next budget cycle, not after. A defensible, specific number changes the conversation from "can we afford to fix retention" to "can we afford not to." Once it is expressed in the same currency as everything else on the P&L, it stops competing with other priorities on vibes and starts competing on numbers, which is a conversation operations usually wins.