Salesforce Announces Prepackaged Agentforce Help Agent
salesforce.com · 13 October 2026
Salesforce has deployed Agentforce Help Agent with pay-per-resolution pricing (charging only when the agent autonomously resolves an issue end to end). On Salesforce's own help portal, Agentforce processed 4.3 million inquiries and resolved 70 percent of them. Salesforce moved from per-seat or flat-fee pricing to outcome-based pricing tied to successful autonomous task completion. Three named customers (PenFed, Fisher & Paykel, PowerSchool) are described as using Agentforce for agentic self-service. Fisher & Paykel reports doubling self-service resolution rates. The agent is deployable in minutes and works across voice, web, portal and messaging channels.
This is the observable fact the cited source supports, stated without interpretation.
LAW_2 predicts a shift from access-based to outcome-based pricing in autonomous-affected sectors. This observation tests that prediction in a commercially deployed, enterprise-scale product with named customers and internal production evidence of resolution rates.
This is a direct instantiation of LAW_2 in practice. Salesforce has shifted from access-based (per-seat, flat-fee) revenue capture to outcome-based pricing that ties revenue to successful autonomous completion of a defined task (customer service case resolution). The shift represents a move toward outcomes pricing in place of usage or access pricing. The mechanism is being embedded in a commercially deployed product offered to enterprise customers. The 70 percent resolution rate on Salesforce's own operations provides an observable constraint on the scalability and reliability of the autonomous agent in a production environment with continuous demand.
This is a direct instantiation of LAW_2 in practice. Salesforce has shifted from access-based (per-seat, flat-fee) revenue capture to outcome-based pricing that ties revenue to successful autonomous completion of a defined task (customer service case resolution). The shift represents a move toward outcomes pricing in place of usage or access pricing. The mechanism is being embedded in a commercially deployed product offered to enterprise customers. The 70 percent resolution rate on Salesforce's own operations provides an observable constraint on the scalability and reliability of the autonomous agent in a production environment with continuous demand.
This is the framework's reading of the fact above. The source does not endorse it and is not cited as doing so.
The text establishes that outcome-based pricing is now being implemented in a major SaaS platform deployed across enterprise customer bases. However, the publication date (June 2026) is concurrent with product launch. The paper's LAW_2 horizon test requires tracking whether, over 3 to 5 years, outcome-based pricing models persist, expand, or collapse under autonomous deployment in affected sectors. Whether Salesforce sustains this pricing model under actual customer deployment, whether revenue durability holds, and whether the structure reveals commercial misalignment remain longitudinal questions. The constraint of resolution reliability (70 percent) is anticipated in the paper as a likely boundary on deployment speed, not a contradiction of the pricing model itself.
Current evidence is recorded immediately. The paper's longer empirical horizon determines when persistence becomes a stronger framework-level test.
Outcome-based pricing at launch does not prove that value capture is moving toward orchestration, governance and verification (LAW_1 and LAW_3). It may reflect marketing positioning or temporary customer accommodation. If Salesforce later retreats to usage-based or per-seat pricing, or if customers negotiate back to access pricing under volume constraints, the mechanism would weaken. Additionally, whether Salesforce itself captures value through orchestration and governance (directing and constraining the agent's actions) in place of production capability is not established by this text alone.
Recorded for every observation assessed at strength 4 or 5. A record that cannot state the strongest competing reading of its own evidence is not published.
Where several outlets report the same development, they are recorded as additional sources on this one observation and are not counted as separate evidence.