A BPO price is not just a number. It is an answer to the question: who pays when things turn out differently from the plan? Volume falls, calls get longer, the client’s system misbehaves, a campaign converts better than expected. Whoever the billing model makes responsible for each of those surprises is carrying risk, and risk has a price. That is the right way to compare the three common models, and my view, stated up front, is that FTE billing is the most honest for the first year of a new process, while transaction and outcome models earn their place only when the work is stable and measurable.
One illustrative process, three prices
Take an invented back-office process: a client expects about 30,000 transactions a month, each needing around 8 minutes of handling. Assume an agent has about 154 productive-capable hours a month (22 days, 7 hours) and occupancy of 85 per cent, which gives roughly 131 working hours of handling per agent. The workload is 30,000 times 8 minutes, which is 4,000 hours. Divide by 131 and you need about 30.6 agents, so the BPO staffs 31. Fully loaded cost is ₹35,000 per agent per month, so the monthly cost is ₹10,85,000.
Now price it three ways to the same revenue of about ₹18 lakh:
- FTE-based: ₹58,000 per agent per month, 31 agents, ₹17,98,000. Gross margin is about 39.7 per cent.
- Transaction-based: ₹60 per transaction, 30,000 transactions, ₹18,00,000.
- Outcome-based: a smaller base fee plus payment tied to a result, such as accounts resolved or applications approved. The structure varies so much that I will describe it below.
At plan, all three look the same. Their differences show when reality moves.
When volume drops
Volume falls to 20,000 a month. Under FTE billing the client still pays for 31 agents, ₹17,98,000. The client carries the risk, and may complain about paying for idle capacity, which is why FTE contracts often have clauses allowing ramp-down with notice. Under transaction billing, revenue falls to ₹12,00,000. If the BPO cannot reduce headcount quickly and keeps 31 agents costing ₹10,85,000, margin shrinks to about 9.6 per cent. Release agents fast and the cost drops to about ₹7.1 lakh for 20 agents, but notice periods, training and the cost of re-hiring if volume returns make that harder than it looks on paper. So the BPO carries volume risk under per-transaction pricing, and it will price that risk into the rate or demand a minimum volume commitment.
When handle time rises
Back to 30,000 transactions, but the client adds two new checks and the handling time goes from 8 to 10 minutes. The workload is now 5,000 hours, which needs about 38.2 agents, so 39. Cost becomes ₹13,65,000. Under FTE billing the client pays for 39 agents (₹22,62,000 at ₹58,000 each) and the margin percentage stays unchanged. Under transaction billing revenue stays at ₹18,00,000, and margin falls to about 24.2 per cent. The BPO carries handle-time risk. A well-written transaction contract therefore says that a change in the process, requested by the client, triggers a re-pricing. If it does not, the BPO is quietly funding the client’s process changes.
Outcome-based pricing
Here the BPO is paid for a business result: money collected, sales closed, claims settled within a specified quality. A typical structure, with variations, is a modest fixed component to cover base costs and a variable component linked to the outcome. For collections, the BPO may receive a percentage of the amount recovered. This model shares the upside, and clients like it because they pay for results. But the BPO is now exposed to things outside its control: the quality of the debtor pool, product competitiveness, the client’s own pricing, regulatory changes. Outcomes also need a clear, auditable definition and a clean attribution rule. If a customer pays after an earlier reminder from the client’s own team, who gets the credit? A contract that does not answer this will generate disputes every month.
Which model suits which work
| Model | Works best when | Who carries volume risk | Who carries handle-time risk |
|---|---|---|---|
| FTE-based | Process is new, variable or judgement-heavy | Client | Client |
| Transaction-based | Work is repetitive and units are easy to define | BPO | BPO |
| Outcome-based | Result is measurable and mostly within the BPO’s influence | Shared | BPO |
Hybrids are common in practice: an FTE core with a transaction rate for overflow, or a transaction price with a volume band and re-pricing outside the band. They are sensible, because they split risk where each party can manage it.
Where I take a side
FTE billing has a reputation for rewarding inefficiency, since the BPO earns more by using more people. That criticism is fair only if nobody is measuring output. With productivity targets, service levels and a regular review, FTE pricing keeps both sides focused on the work rather than on defending a price. Transaction pricing sounds modern, but it pushes the BPO to cut handle time, and in customer-facing work that can reduce quality. Outcome pricing is attractive on a slide, and fragile in a dispute.
So for a new process whose true volume and handle time nobody knows yet, start with FTE, collect six months of data, then move to a per-transaction or hybrid model with bands and re-pricing triggers. Pricing is easier to get right once you know the numbers. Before that, a transaction rate is a guess with a legal document around it.
Figures and names in the examples are illustrative.
Leave a Reply