Salesforce AI agents handle 40% of support traffic with high resolution rates

Salesforce has integrated Fin, an AI-native customer support platform acquired from Intercom, into its Service Cloud business, which generates nearly $11 billion in annual operating value. Fin utilizes outcome-based pricing, charging for resolutions rather than seats, a model that aligns with broader industry trends toward consumption-based metrics. In an internal pilot, Salesforce deployed Fin in 14 days, slightly faster than the average 16 days for external customers. The system now handles over 40% of traffic on help.salesforce.com and achieves a 65% resolution rate without human intervention. This shift allows Salesforce to capture a larger share of the labor budget by replacing manual support tasks. The company’s internal deployment demonstrates the practical application of AI in reducing operational costs while maintaining service levels. By automating a significant portion of support interactions, Salesforce is testing the viability of AI-driven resolution as a core component of its enterprise offerings.