Oracle Shifts to ‘Action Units’ for Agentic AI Pricing, Forrester Analyst Highlights Developer Burden
Oracle is transitioning its pricing model for agentic applications within its Fusion suite, moving away from traditional per-user, per-month subscriptions to a consumption-based system called “Action Units.” This shift comes as AI-powered agentic applications become more prevalent, and developers grapple with the long-term maintenance of these systems, according to industry analysts.
Understanding Action Units
According to Oracle’s Rachelson, Action Units are a single consumption metric, each valued at approximately one cent. These units are consumed at varying rates depending on the complexity of the AI task performed, ranging from simple queries to more intensive operations like document or video creation. All AI processing across Oracle’s Fusion pillars draws from a shared pool of Action Units.
Forrester Insights on Agentic Application Maintenance
Kate Leggett, Principal Analyst at Forrester, emphasizes that the ongoing maintenance of custom agents and agentic applications represents a significant, often underestimated, burden for developers and CIOs. Kate Leggett is a leading expert on customer relationship management (CRM) and customer service strategies. Leggett has recently published insights on the evolving landscape of AI in customer service, including the agentic shift and the implications for SaaS pricing models.
The Rise of Agentic Applications and AI in Customer Service
The move towards agentic applications is fueled by the increasing capabilities of AI. As Leggett notes, customer service operations are evolving from reactive, cost-heavy, manual functions to proactive, revenue-driving engines powered by AI. This transformation is attracting significant investment from C-level executives and board members eager to capitalize on the potential ROI of AI in customer service technology.
Recent Forrester Analysis
Recent blog posts by Kate Leggett at Forrester cover key trends in the industry:
- The Salesforce Consulting Landscape: Leggett observed that Salesforce, with $35.7 billion in FY 2025 revenue and over 150,000 customers (including 90% of the Fortune 500), continues to dominate the CRM vendor landscape.
- The “SaaS-pocalypse” and AI-Native Alternatives: Leggett explored the potential impact of a recent software stock sell-off, suggesting a possible investor shift towards AI-native solutions, whereas cautioning that the enterprise core is not disappearing but rather transforming.
- Google’s Entry into the Contact Center Market: Leggett highlighted Google Cloud’s Customer Experience Agent Studio, announced at the NRF event, and its potential impact on the CX and contact center landscape.
- AI Realities in Customer Service (2026 Predictions): Leggett predicted that while AI won’t fundamentally transform customer service in 2026, foundational operate will be crucial.
The shift to Action Units and the broader adoption of agentic AI applications signal a significant evolution in how software is priced and consumed, with a growing emphasis on value-based, consumption-driven models.
Related reading