The outsourcing industry spent thirty years selling one product: people in chairs. You bought agents by the seat, paid by the hour, and hoped the volume added up to good service. That logic held when every contact needed a human. It holds a lot less well now that software can finish a password reset before an agent would have read the ticket.
Most providers know this. Fewer have changed their contracts. What usually happens instead is an AI line item bolted onto the same seat-based deal, which leaves you paying for the automation and the headcount it was supposed to shrink. If you are weighing a partner right now, these are the things that give away a provider still working from the old playbook.
The proposal opens with a headcount
Watch the first number in the pitch. When it is the size of the team, the provider is anchoring on an input. The thing you are paying for is resolution: contacts closed, customers kept, calls that do not come back a week later. A partner built around results will lead with resolution rate and cost per resolved contact, and let team size be a consequence of those rather than the headline.
Your bill only moves when volume moves
Seat pricing has a quiet flaw. It charges you for capacity whether or not that capacity does anything. There is one question that exposes it: if your automation deflects sixty percent of contacts next quarter, does the invoice shrink? When the answer is no, you are renting chairs.
The AI is sitting on top of the old process
Plenty of legacy providers switched on a chatbot and called it a transformation. You can usually spot the retrofit, because the bot collects a few details and then hands the customer to an agent who starts over from the top. AI-first support works the other way around. It resolves what it can end to end, and when it does pull in a person, the full history comes along so nobody restarts the conversation.
You are asked to trust the automation blindly
If a provider cannot show you which contacts the AI resolved, which it escalated, and why, then you are taking deflection on faith. Ask for a live view of it. A partner confident in its numbers hands one over without much fuss. Vague dashboards and a quarterly PDF usually point the other way.
The quarterly review is about staffing
Notice what fills the business review. If the hour runs on shift coverage, attrition, and requests to add seats, the relationship is pinned to labor. If it runs on retention, first-contact resolution, and what last month's contacts revealed about your product, you have a partner reading the same scoreboard you are.
None of this makes seat-based work wrong everywhere. Complex, low-volume support that resists automation can still be worth staffing by the hour. The mismatch appears when that model gets stretched over thousands of repetitive contacts a machine could have closed, and you are the one covering the gap. If you want to move off it, measure your own cost per resolved contact first, get a rough read on how much of your volume is automatable, and take both into the next vendor conversation. Ask to be priced against them. The providers worth your time will say yes.
Frequently asked questions
What is outcome-based BPO pricing?
It charges for results such as resolved contacts, retained customers, or completed tasks, rather than for agents or hours. The idea is to line the provider's incentive up with your results instead of with the size of the team.
Is seat-based pricing always a bad deal?
No. For complex, low-volume work that automation cannot touch, paying skilled people by the hour is reasonable. Trouble starts when the same model gets applied to high-volume repetitive contacts, where you end up funding work software could handle.
How do I move off a seat-based contract?
Start by measuring your current cost per resolved contact and estimating how much of your volume could be automated. Then ask prospective partners to quote against those figures. A provider that believes in its automation will take the terms.