There is a recurring debate in monetization design that usually goes the same way: someone declares that per-seat pricing is dead and usage-based pricing is the future. Then someone else points out that usage-based pricing creates revenue volatility that makes forecasting impossible. Then someone suggests outcome-based pricing as the neutral solution, and the conversation collapses into a discussion of how to define and measure outcomes.

What gets lost in this debate is that each monetization model is a different answer to a different question. They are not interchangeable — they have different internal logic, different tradeoffs, and different organizational requirements. Choosing one means accepting its costs as well as its benefits.

This post maps the actual tradeoffs across the four primary models.

Usage-Based Pricing

What it is: A variable charge based on consumption — API calls, records processed, emails sent, minutes used, actions taken. Revenue scales with usage.

What it buys: Alignment. Revenue and value delivery move together. When the customer gets more value, they pay more. When they get less, they pay less. The expansion mechanism is built into the model: customers grow into higher usage tiers as they get more value.

What it costs: Predictability. Revenue becomes a function of customer behavior rather than a function of contracts. If your customer base has seasonal usage patterns, your revenue will too. Forecasting requires understanding not just how many customers you have but how they behave — and that behavior is harder to model than contract renewals. Usage-based pricing also creates "sticker shock" risk: customers who don't set budgets can get surprised by bills, which creates churn at the moment of highest value delivery.

When it works: Usage-based pricing works best for products where value is proportional to consumption and where consumption is something the customer can control and budget for. API-heavy infrastructure products, communication platforms where message volume varies, data processing tools where records processed is the relevant unit.

When it breaks: It breaks when usage is not proportional to value, or when usage is driven by things outside the customer's control. If your product creates dependencies that spike usage unpredictably, customers will feel punished for having adopted you. It also breaks in enterprise contexts where procurement teams need to budget for software in advance — pure usage billing is often incompatible with how enterprise budgets work.

Outcome-Based Pricing

What it is: Pricing tied to a business outcome rather than a product metric — revenue sharing, savings capture, or a fee tied to a measurable result. The company takes on risk that the product will deliver, and gets compensated when it does.

What it buys: Maximum alignment. If the product delivers the outcome, pricing is justified. If it doesn't, the customer pays less or nothing. Outcome pricing removes the customer's risk from the purchase decision and can enable deals that would otherwise be blocked by procurement risk aversion.

What it costs: Measurement and trust. Outcome-based pricing requires agreeing on what the outcome is, how it's measured, and who controls the measurement. These are organizational and contractual problems, not just pricing problems. Outcome-based pricing also requires the vendor to trust their own product enough to bear the outcome risk — which requires evidence that the product actually causes the outcome, not just correlates with it.

When it works: Outcome-based pricing works in highly specific, measurable contexts where both parties have high confidence in the causal relationship between the product and the outcome. Some performance marketing tools, certain HR and recruiting products where output is measurable, some infrastructure products where the outcome is well-defined.

When it breaks: It breaks in most contexts because outcomes are rarely attributable to a single product. If a customer's revenue increases after adopting your product, did your product cause the increase, or would it have happened anyway? Outcome pricing that can't answer this question cleanly creates disputes at renewal that poison the relationship.