Outcome-based pricing: a commercial pricing model in which a BPO vendor’s compensation is tied, in whole or in part, to a pre-defined business result, such as resolved tickets, collected payments, or retained customers, rather than to hours worked or seats filled. In practice, pure outcome pricing is rare; most real contracts use a hybrid structure that combines a base rate with a performance bonus anchored to that result.
What Does Outcome-Based Pricing Actually Look Like in a BPO Contract?
Most live contracts are not pure outcome deals. A vendor paid only when outcomes are achieved carries real cash-flow risk, which creates its own problems: underpaying agents, cutting corners on quality, or walking away from a difficult process mid-engagement. What actually gets signed is usually a base-plus-bonus structure, where the vendor earns a floor rate for hours or transactions, and an additional bonus when a defined metric clears a defined threshold.
A collections team, for example, might earn $10 per hour base plus 3% of dollars recovered above a baseline recovery rate. A customer retention team might earn a per-seat rate plus $15 per customer retained above the prior-quarter average. The “outcome” component is real, but the base keeps operations stable on both sides.
The version closest to pure outcome pricing I see work consistently is per-resolution or per-transaction pricing: the vendor bills only for completed, verified units of work. Zendesk and similar platforms have made this more tractable by giving buyers per-ticket resolution data that can be audited. Outcome-based pricing in AI-assisted support is attracting real attention right now for exactly that reason: when a bot resolves a ticket, the resolution event is logged, timestamped, and attributable without ambiguity. That auditability is the missing ingredient in most outcome contracts.
Why Attribution and Baseline Rules Make or Break This Model
The most common point of failure is not the pricing formula. It is agreeing on what counts as the outcome and who gets credit for it.
If your BPO vendor handles tier-1 customer support and churn drops 8%, how much of that drop is attributable to the vendor? Your product team shipped two updates that quarter. Your marketing team ran a loyalty campaign. Your vendor handled calls faster. Attribution is genuinely hard, and if you do not solve it contractually before signing, you will fight about it on every invoice.
The same problem appears in outcome-based pricing examples across industries, from pharma rebates tied to patient outcomes to SaaS deals tied to platform adoption. The underlying challenge is the same: downstream results are multi-causal, and a vendor can only control their slice of the chain.
What I would require before signing any outcome-based contract:
- A pre-contract baseline diagnostic. You need at least 90 days of historical data on the metric you are tying pay to. Vendors who push to skip this step are usually betting that the baseline will be set low. If the metric is not already instrumented in your systems, the model is not ready to deploy.
- Explicit attribution rules. Define which actions, channels, and time windows count toward the vendor’s credit. Write these into the statement of work, not the MSA boilerplate.
- Exclusion clauses for external shocks. A product outage, a pricing change, or a seasonal spike should not count against or for the vendor. Define these events in advance.
- An audit right. You or a neutral third party should be able to verify the outcome data. Vendors who resist this are a signal worth heeding.
Outcome-Based vs. Other BPO Pricing Models
Outcome-based pricing sits at one end of a risk-sharing spectrum. Here is how it compares to the models buyers most commonly consider:
| Model | How You Pay | Best Fit | Risk to Buyer |
|---|---|---|---|
| Per hour | Hourly rate per agent | Pilots, variable volume, complex work | Paying for effort, not output |
| Per seat / dedicated FTE | Monthly flat per agent | Stable volume, process ownership | Underutilized capacity |
| Per transaction | Fixed fee per completed unit | Repeatable back-office tasks | Output without quality control |
| Outcome-based (hybrid) | Base rate plus performance bonus | Measurable results with clean attribution | Baseline disputes, bonus gaming |
| Pure outcome / gainshare | Percentage of value created | Mature partnerships, fully instrumented data | Vendor financial instability, attribution fights |
Per-transaction pricing is often the practical middle ground. It links payment to completed work without requiring the vendor to absorb the full risk of a multi-causal business result. For documented, repeatable processes, it is often the better starting point than a pure outcome deal.
When Outcome-Based Pricing Is Worth Pursuing, and When It Is Not
Outcome-based pricing earns its complexity when the outcome is clean, the data is already running, and the vendor has meaningful control over the result. Collections, lead qualification with a defined conversion event, and AI-resolved support tickets are the cases where I have seen it work without constant contract disputes.
I would be careful with outcome pricing when the process is new or not yet documented, when the metric requires data engineering work to track, or when the vendor’s contribution is one of several factors driving the result. If your operations are not yet clean, you will not get a good outcome deal. You will get a vendor who prices in the measurement risk, and you will pay more than a straightforward per-hour or per-transaction structure would have cost.
The broader lesson: do not outsource to an outcome-based model to avoid defining what success looks like internally. Define it first. Then structure the contract around it. The model rewards buyers who have already done the measurement work, not those hoping the vendor will do it for them.
If you are evaluating vendors and want to compare pricing structures across qualified BPO providers, get outsourcing quotes from the Global BPO Index directory.