Resolution-Based Pricing

You see this in AI-powered customer service platforms that charge customers based on resolved interactions rather than on seats, volume, or time. Resolution-based pricing ties cost directly to outcomes: the vendor is compensated when an issue is genuinely resolved, not simply when a conversation takes place. This model is designed to align vendor incentives with customer outcomes by making it in the vendor's interest to actually solve the problem rather than just generate activity.

A retail company evaluates two AI support vendors. The first charges per conversation. The second charges only for resolved interactions, as defined by a mutually agreed-upon standard. The retail company chooses the second model because it shifts financial risk to the vendor and creates shared incentive to improve containment, knowledge coverage, and escalation logic. However, the procurement team spends significant time negotiating what counts as a resolution to ensure the definition is accurate and not gameable.

This shows up as a model alignment question as much as a pricing question. Resolution-based pricing can create strong alignment when the definition of resolution is precise and independently verifiable. When it is vague or easily manipulated, it can create perverse incentives where vendors optimize for appearing to resolve issues rather than genuinely resolving them. For teams evaluating AI vendors, the pricing model reflects assumptions about where accountability sits, making it worth examining carefully before signing.