Dreamforce 2025, Salesforce’s flagship event set in San Francisco, featured a flurry of product updates, customer testimonials, and a beating drum that the company is the indispensable orchestrator of AI agents.

Salesforce needs a fresh engine for the next leg of growth, and it is betting big that Agentforce is it. The company set a long-term revenue target of more than $60 billion by fiscal year 2030, which implies roughly ten percent organic compounded growth from FY26 to FY30. It also rolled out a profitable growth framework that aims to have the sum of constant currency subscription and support growth plus non-GAAP operating margin add up to fifty by the end of FY30. Those are ambitious markers. To hit them, Salesforce is pointing almost everything at agentic AI.

Agentforce is the centerpiece of that plan. Salesforce says more than twelve thousand customers across sizes and industries are already on board. The pitch is actually pretty simple. Expand agentic AI across the business and there is a path to three to four times more ARR over time. That kind of promise is exactly what a company needs when the core CRM market is mature, and expansion depends on new categories that create new budget lines.

Recent results offer limited proof. Data and AI revenue was about $1.2 billion dollars in Q2, an increase of one hundred twenty percent year over year. Whether that pace holds and translates into broad Agentforce adoption remains uncertain, and the $60 billion target clearly depends on sustained execution rather than a single quarter.

A range of Salesforce customers were highlighted at the event, including Williams-Sonoma, Dell, Pepsico, and FedEx, though each at varying stages of Agentforce implementations. America’s largest virtual accounting firm, 1-800Accountant, implemented Agentforce in January 2025 to help manage surging demand from more than a million small business users. CTO Ryan Teeples says his company now autonomously handles up to 70% of chat inquiries during tax season, providing real-time, accurate answers to tax questions without hold times or phone calls. Teeples calls Agentforce “a game-changer,” noting how the digital labor platform has freed accounting professionals to focus on complex advisory work while maintaining fast, personalized service.

Finding the Right AI Pricing Model for CX Success
Consistent Agentforce revenue growth will largely depend on clear, scalable AI pricing models. Salesforce is signaling caution on outcome-based pricing. As David Schmaier, president and chief strategy officer, put it at a session with industry analysts, “We don’t know want to get into an argument with a customer over the definition of an outcome.” That stance fits a world where disputes over what counts as a resolved ticket or a deflected call can swamp the value of the deal. It also echoes recent pushback when vendors tried to flip existing contracts to outcome terms without enough safeguards.

Meanwhile, buyers of conversational technology are testing outcome-based models because they align spend with value delivered. Contact center buyers seem to want options like pay per successful containment, pay per qualified lead, or pay per fully handled workflow. These models reward precision and continuous tuning but require shared definitions, reliable attribution, and transparent measurement.

The practical read is simple. Salesforce will anchor on seats, usage, and platform consumption. Outcome-based pricing will remain a niche tactic (at best) for specific customers who really push for it, but definitely not a Salesforce norm.

Command Center: Turning Conversations into Insights
Command Center for Service looks like Salesforce’s bridge into conversation intelligence. The feature set lands very close to what dedicated vendors offer, with orchestration across channels, visibility into customer satisfaction in real time, and the ability to spot knowledge gaps and create articles. Add the Customer Signals Intelligence (cutely known as CSI) component and the direction is clear. Salesforce wants service leaders to extract value from every conversation, whether a human call, an automated exchange, or—most importantly for the future—a blended session where a person and an AI agent work together.

This framing clearly fits the vendor’s broader agentic story. Conversation intelligence is not just scorecards and call summaries. It becomes a control surface that turns raw interactions into guidance, playbooks, and workflow changes. Supervisors, managers, and team leads get tools that recommend actions, monitor fleets of human and virtual agents, and surface patterns that matter for coaching and quality. CSI, and the other tools in Command Center, should deepen that by unifying signals from voice, chat, and digital channels into a single model of intent, sentiment, and outcome. Complementing these innovations is a new AI-powered Service Rep Assistant, a “co-pilot” that helps agents draft responses, generate plans, and provide guided resolutions in real time.

If Salesforce delivers on what its promising, service leaders could run operations with a tighter loop between insight and action. Salesforce would be not only selling analysis of conversations; it would sell a system that tunes behavior in the moment and measures the impact after the fact.

Derek Top, Principal Analyst with Opus Research, contributed to this post.

Categories: Conversational Intelligence, Intelligent Assistants, Intelligent Authentication, Articles