Salesforce pushes outcome-based contact center AI

88

TL;DR:

Salesforce’s new Agentforce Help Agent charges only for resolved customer issues, raising the stakes for outcome-based AI pricing in contact centers.

Article:

Salesforce has launched Agentforce Help Agent, an autonomous AI customer service agent that charges businesses only when it resolves a customer issue from start to finish. The product, announced on June 25, is built on the Agentforce 360 Platform and is set for general availability in July 2026, alongside pay-per-resolution pricing.

The move matters because it shifts contact center AI from usage-based billing to outcome-based pricing. Salesforce says customers will not be charged if a user escalates to a human agent or gives negative feedback. That makes the vendor share more of the risk, but also puts pressure on clients to define what a “resolution” means before procurement gets dazzled by the invoice model.

Help Agent can be deployed across voice, web, portal and messaging channels, with prepackaged workflows for case management, appointment scheduling, order updates and account management. Salesforce says Agentforce has already handled 4.3 million inquiries on its own help site and resolved 70% of them.

The broader market is moving in the same direction. Gartner predicts agentic AI will autonomously resolve 80% of common customer service issues by 2029, cutting operational costs by 30%. “Agentic AI has emerged as a game-changer for customer service,” said Daniel O’Sullivan, Senior Director Analyst at Gartner.

For enterprises, the upside is clear: faster service, lower support costs and fewer repetitive tickets for human teams. The risk is equally plain. Bad knowledge bases, weak escalation rules or vague resolution metrics could turn pay-per-resolution AI into a new form of service opacity. The next test is not whether AI can answer. It is whether customers agree the problem was actually solved.