Pricing the AI Agent Economy

Pricing for AI agents introduces new challenges compared to traditional SaaS models.

SaaS pricing typically relies on the number of seats—the more seats, the higher the price. AI agents, however, are not "tools" for humans. They're autonomous "doers" that perform entire human workflows on their own. Measuring value by the number of "seats" no longer captures their true impact. Instead, AI agents should be benchmarked against human labor, allowing us to quantify their impact more accurately.

The Customer Experience Use Case

At Decagon, our AI agents resolve customer service tickets autonomously, and we measure their performance in two clear ways:

• Resolution rate: What % of conversations are resolved by the AI agent without needing to escalate to a human?
• Customer satisfaction (CSAT): How happy are customers with the conversation they had, measured by survey results?

We've successfully deployed AI agents across countless customers and support two flexible pricing options:

• Per-conversation pricing: A fixed rate for every incoming conversation, with flexible pricing for higher volumes.
• Per-resolution pricing: A higher fixed rate for each fully resolved conversation, with no charge for escalations. This higher rate reflects the added value of delivering complete resolutions. Larger resolution commitments lower the rate.

Both scale with the agent's work, but we've seen the majority of our customers gravitate towards per-conversation pricing.Why Customers Choose Per-Conversation Pricing

1. ‍Predictable and Transparent. Per-conversation pricing is simple: costs scale directly with usage. Customers avoid unpredictable invoices and the constant renegotiations often required with outcome-based pricing, where the nature of "successful outcomes" are unclear. If a frustrated customer stops responding, does that count as "resolved"? If the agent provides a trivial answer, should it cost the same as a major technical fix? This is especially important for businesses who want to allocate set budgets.‍
2. Aligned Incentives. By focusing on conversation volume rather than overly parsing the definition of "outcome," the incentives are clean. We're not incentivized to push partial resolutions or sidestep tough cases. Instead, we can pour energy into what really matters: getting your customers the right answers and increasing your resolution rates and customer satisfaction. You'll get into situations where a user is upset, leaves, and that gets counted as a resolution. You never want to be in a situation where you're arguing over what a "resolution" is.
3. A Foundation of Trust. Our customers depend on us as trusted partners. They value that Decagon's agents deliver higher deflection rates, drive superior CSAT, and offer full visibility into their logic and decision-making. We focus our partnership discussions on enabling your growth and satisfaction, rather than negotiating over definitions and billing structures.

While we're happy to support both models, our goal with this post is to share the learnings we've seen from countless successful deployments. The vast majority of our customers choose per-conversation pricing. It's transparent, scalable, and aligns perfectly with Decagon's focus on building long-term partnerships that deliver consistent improvements in deflection rates and customer satisfaction.

The Future of AI Agent Pricing

We're building for a future where AI agents are integral teammates—always on, always learning, and always improving. A pricing model that's simple and scalable cements that relationship.

At Decagon, we're committed to creating long-term partnerships. Whether you choose per-conversation or per-resolution pricing, you'll benefit from fast deployments, unmatched transparency, and continuous performance improvements. Together, we're building a future where AI delivers real value for your business and your customers.