Platform overview
Salesforce Marketing Cloud Personalization, formerly Interaction Studio and before that Evergage, handles real-time 1:1 personalization, recommendations and decisioning across web, email and mobile. It now includes native web experimentation: marketers run experiments against a traffic subset, define custom primary and secondary metrics, and review lift, chance-to-beat-control and chance-to-beat-all. Experiment data is drillable in Data Cloud and exportable to BI tools.
Custom engagement signal metrics, compound metrics and first/last-touch attribution models feed the recommenders and the decisioning engine.
Agentforce works in both directions. Personalization context actions run ahead of Agent topics so agents decide with visitor context, and Agentforce Actions turn live conversation intent into structured signals for the decisioning engine.
Agent access is governed at the Salesforce platform level through Hosted MCP Servers using per-user OAuth 2.0 with PKCE. Those servers cover SObjects, Data 360, Tableau Next and customer-exposed Apex/Flow tools, but there's no Personalization-specific MCP server. Pricing now itemizes Agentforce Personalization Decisioning, AI objective-based recommendations and AI next-best offer.
.svg)
