Ask The Expert: How Should I Monetize Ai Agents?

Hello, GeekWire readers. We launched our new business advice column earlier this month, featuring different experts from the Pacific Northwest tech community, answering questions about an array of topics. It’s inspired by common questions from tech workers and entrepreneurs.

We’re back with another question, this time focusing on how to monetize AI agents.

Our expert this week is Manny Medina, the co-founder and former longtime CEO at Outreach, a Seattle-based sales automation software company that has raised nearly $500 million and reached a $4.4 billion valuation in 2021. Medina, who previously was a director at Microsoft, stepped down as Outreach CEO in September and remains executive chairman.

Do you have a good question, or are you an expert that wants to contribute to a future column? Email us at tips@geekwire.com. Our next column will highlight the debate over remote and in-person work. — Taylor Soper, GeekWire editor.

Q: I’m a chief revenue officer at an enterprise software company. We are trying to figure out how to price our new AI products and how to differentiate from our traditional SaaS business model. Do you have suggestions?

MANNY MEDINA: AI agents are growing rapidly, but monetizing them remains a challenge. After speaking with several founders and CEOs, I’ve identified key approaches companies are taking, each with their own benefits and tradeoffs.

The simplest and most common way to monetize AI agents is to bundle them into existing seat-based pricing. It’s easy to implement, straightforward for customers to understand, and works well as an enterprise add-on or upsell. However, this approach exposes you to variable LLM costs, which can eat into your margins.

Another popular method is charging based on consumption, like credits or usage metrics. It’s quick to set up and helps you pass LLM costs directly to customers with a margin added. But consumption pricing has a downside: customers may struggle to connect credits to real value. This lack of clarity can lead to churn, and it leaves you vulnerable to competitors offering cheaper alternatives.

Outcome-based pricing is gaining traction as a way to align agent performance with customer goals. Salesforce, Sierra, Intercom, and recently Zendesk have announced outcome-based pricing. By tying pricing to measurable results — like tasks completed or tickets resolved — you make it easier for customers to see your value. This approach strengthens loyalty and reduces switching, but it’s harder to implement and often requires significant investment in back-end systems.

For even greater alignment, success-based pricing ties cost to higher-value achievements, like meetings set or customer resolutions with high satisfaction. This model differentiates you in the market and makes pricing part of your strategy, but it’s complex to build. Measuring success varies by customer and by agent, and as agents take on more sophisticated work, tracking value becomes increasingly nuanced.

The reality is that AI agent monetization is a journey. Most companies start with simpler models like bundling or consumption-based pricing, then evolve toward outcomes and success as their product and customer understanding matures. Often, the best approach is a hybrid — combining elements of consumption, outcomes, and success to align pricing with the value your agents deliver.

Ultimately, monetization is about connecting what your AI agents do with what your customers care about. Start simple, iterate, and aim for pricing that reflects real value.

Previously: What are my career options at a slow-growing software-as-a-service company?