This is the year that AI agents will pick up even more momentum in business!—That’s what it seems like when you read the tech headlines. Every major vendor has announced or (by now) even delivered the first Agentic AI assistants and platforms. Working just on Generative AI seems so ‘2023.’
From a software vendor’s perspective, Agentic AI is both an opportunity and a threat: You create a compelling vision and product, and you tie your customers even closer to your ecosystem. You don’t, or you overlook a new and more nimble player, and you’ll find yourself in uncharted territory, unable to compete on the same terms.
So how do you turn this Agentic AI hype into business outcomes (aka revenue), then?
Software Pricing has Always Been Tricky
The core principles of how software works are evolving. Historically, software developers have explicitly programmed systems (e.g., “if this, then that, else do…”). With the emergence of machine learning about 10 years ago, systems have been able to detect patterns in data (e.g., “people who’ve bought this hammer have also bought this toolbox”) as opposed to relying on explicit instructions for certain, narrow tasks. Agentic AI takes this even another step further as users only need to define a goal that the software breaks down into smaller subgoals and executes (e.g., “gather information from these sources and draft a marketing brief “). But that’s not the only evolution.
Evolution from Programming to Agentic AI
Over the years, software deployment paradigms have evolved as well—from installing and maintaining software in your data center on your hardware (on-premise) to renting software installed on hardware in a provider’s data center (Cloud).
Along with this evolution, the monetization models have further evolved, too—from perpetual licenses to user-based (e.g., for applications) or transaction-based (e.g. for platforms) pricing models. Software vendors have evolved these commercial models into value-based pricing metrics with Generative AI. The latter has already been difficult since it takes the cost plus a margin as a delta to the business value that it creates—and that value can fluctuate between companies of different sizes, for example.
Thank you for reading The AI MEMO. Feel free to share it with your network.
NEW ONLINE COURSE: Monetizing Agentic AI: From Technology to Revenue Growth
Agentic AI is the hottest topic of 2025. The right commercial model makes the difference between making cash or burning through it! Define the commercial model for AI agent-based solutions that will create new revenue streams and enthusiastic customers, based on 20+ years of experience in enterprise software. (Use code AGENTS50 and get 50% off.)
Disrupting Established Players via New Monetization Models
Incumbents typically have a larger installed base than startups and scaleups. They can more easily upsell new products, but they also often have a more complex, historically grown product portfolio to begin with. This portfolio means that incumbents have challenges to compete with more nimble players on new metrics other than “user-based", “transaction-based,” or “consumption-based” pricing.
New market entrants do not have these same constraints. They are in a position to disrupt the market by introducing an entirely new monetization model. If product & GTM leaders set this course early, they can gain the lead position while incumbents debate the best answer.
Salesforce created the Cloud software category and paradigm more than 25 years ago. It is a good example of creating such a new category.
Agentic AI presents another big shift in pricing and monetization models. But to do it right and to make it easy to understand for their customers, software vendors need to put in a lot of work upfront. As AI agents take off, the number of users (and “per-user”-subscriptions is expected to decrease).
What makes Agentic AI more complex to price than previous generations of software:
Mix of resource-, transaction-based platform components
Unknown number of iterations between agents
Variable prompt (token) length per instruction
Setting a unit of measure and price point is critical to avoid leaving money on the table or risking cost overruns on case of higher consumption/ usage by customers.
Usage-based scaling to mitigate cost overruns
On top of it all, Agentic AI enables software vendors to provide true “outcome”-based monetization models in which customers only pay by business-related KPI (e.g. case resolved. However, this model requires an in-depth understanding of the market and of its customers.
Summary
Software is constantly evolving—from traditional programming to recognizing patterns in data and defining a goal.
Startups and scale-ups have a good chance of disrupting industries or creating entirely new ones. Agentic AI presents an opportunity to move toward value- or outcome-based monetization models. But it’s not an easy task!
Join my bi-weekly live stream and podcast for leaders and hands-on practitioners. Each episode features a different guest who shares their AI journey and actionable insights. Learn from your peers how you can lead artificial intelligence, agentic AI, generative AI & automation in business with confidence.
