The Promise of an 'Agentic Enterprise'
For the past year, Salesforce has been aggressively pushing a new vision for business software centered on 'agents'—AI tools designed to not just assist human workers but to take on tasks and achieve goals autonomously. This concept, dubbed the 'Agentic
Enterprise,' is powered by Salesforce's Einstein AI platform, which has been rebranded and bundled into a top-tier offering called Agentforce. Recent earnings reports show impressive momentum, with Agentforce generating over $1.5 billion in annual recurring revenue. The company's strategy involves integrating AI deeply into its core products like Sales Cloud, Service Cloud, and especially Slack, which it positions as the conversational hub for this new way of working. The goal is to move AI from a novel feature to the fundamental operating system of a business.
The Monetization Puzzle
The central challenge for Salesforce is turning this vision into a clear and profitable business model. The company has moved away from selling AI as a collection of separate, confusing add-ons. Instead, it's pushing customers toward premium bundles like the Agentforce 1 Edition, which packages AI agents with its CRM, data tools, and Slack for a single, high-end price. However, this shift raises new questions. As AI agents begin to perform tasks traditionally done by human users, the classic per-user, per-month software license becomes less relevant. This is forcing Salesforce and its customers to explore consumption-based pricing models, where costs are tied to usage, such as the number of AI actions or API calls. Getting this transition right is critical, as analysts note that clear signs of customers upgrading to these premium tiers are necessary to justify the company's AI investments.
The Specter of 'Chaos'
The 'chaos' Salesforce must avoid comes in several forms. First is pricing complexity. While bundling simplifies options, the introduction of consumption-based credits can create budget uncertainty for customers, who may struggle to predict their AI usage and costs. Second is the data problem. AI agents are only as good as the data they are fed, and many companies struggle with siloed, inconsistent, and unstructured information. Salesforce's Data Cloud is meant to solve this by creating a unified view of customer data, but it remains a complex product to implement. If the data foundation isn't solid, AI tools can produce unreliable results, eroding trust. Finally, there's integration chaos. By opening its platform via 'Headless 360' and partnering with other AI leaders like Anthropic, Salesforce is allowing external AI assistants to access and use its data, which raises questions of governance, security, and control.
What to Watch at Dreamforce
Dreamforce 2026 will be a litmus test for how well Salesforce is managing this transition. Customers and investors will be looking for more than just impressive demos; they want proof. Key announcements will likely center on showing tangible business value and customer success stories with hard numbers—metrics that have been largely absent from pre-event messaging. Expect a heavy focus on simplifying the AI story, clarifying pricing models, and demonstrating how Data Cloud, Slack, and Agentforce work together seamlessly in the real world. The success of the event won't be measured by the splashiness of new features, but by whether Salesforce can convince thousands of business leaders that its AI strategy is not just powerful, but also practical, manageable, and, most importantly, profitable for everyone involved.
