How BPO engagements are priced

Five models cover almost every contract: per-seat (a fixed monthly rate per dedicated agent), per-hour or per-transaction, outcome-based, managed service, and AI-augmented blended pricing. Knowing which model a quote uses, and what it quietly includes or excludes, is most of the work of comparing providers.

Ask explicitly, because it varies between providers: QA, a team lead share, tooling, and reporting sit inside some quotes and are billed as extras on others. Per-seat is predictable and simple to budget, but you carry the idle time - the seat costs the same on a quiet day.

Per-hour or per-transaction billing, because cost follows volume instead of a fixed seat sitting idle. The trade-off is quality: a per-unit price rewards speed, so agree quality thresholds and rework terms before signing.

When the outcome is clearly measurable - qualified leads delivered, collections recovered, bookings made. It is the most aligned model in those cases and the hardest to negotiate when the result is ambiguous. Expect the provider to price in their risk; a hybrid base-plus-bonus is the usual landing zone.

The provider owns the whole function against service levels - staffing, workforce management, tooling, and continuous improvement - so you buy the SLA rather than the people. It suits mature processes you want off your plate entirely, and needs the most careful contract because switching costs are highest.

AI absorbs part of the routine volume, so fewer seats handle the same workload, usually with a platform or tooling fee alongside a smaller human team. When a provider quotes this model, ask which tasks the AI handles, who reviews its output, and how the seat reduction is split between their margin and your invoice.