Agentforce Crosses $1B, But Salesforce's Moat Is Still Being Tested

Salesforce stock is down 33% this year. The market has essentially been pricing in a slow death by AI. Then Wednesday's numbers arrived.

The company posted Q1 revenue of $11.13 billion against a consensus of $11.05 billion, and adjusted EPS of $3.88 versus an expected $3.12—a beat of nearly 25% on the bottom line. Net income climbed from $1.54 billion to $2.11 billion year over year. By most measures, a solid quarter.

The stock barely moved after hours.

The Number That Actually Matters

Bury the lede at your peril: Agentforce annualized revenue hit $1.2 billion, up 205% year over year, crossing the $1 billion threshold for the first time. That's the headline Salesforce CEO Marc Benioff has been building toward since the company launched its AI agent platform in late 2024.

Agentforce lets businesses deploy autonomous AI agents across sales, customer service, marketing, and commerce workflows—essentially automating tasks that used to require human labor or clunky rule-based bots. The $1.2 billion figure is still small relative to the company's $111 billion total quarterly revenue run rate, but the trajectory is the argument. Salesforce is betting that its decades of customer data and workflow integration give it an edge no AI startup can replicate overnight.

The quarter also included the acquisition of commerce startup Cimulate and sales startup Momentum (both undisclosed terms), plus a notable win: the U.S. Veterans Health Administration agreed to adopt an AI agent system built in Slack. Public sector reference deals matter for enterprise sales cycles.

Sponsored

Advertise with Us

[email protected]

Sponsored

Where the Story Gets Complicated

For all the Agentforce momentum, two numbers gave investors pause.

First, guidance. For the current quarter, Salesforce projected revenue of $11.27 billion to $11.35 billion—the midpoint lands just below the $11.36 billion analysts expected. Full-year guidance was raised to $45.9 billion to $46.2 billion, but the midpoint of $46.05 billion again trails the $46.12 billion consensus. These are small misses, but in a stock already under pressure, small misses sting.

Second, and more structurally important: remaining performance obligation (RPO) came in at $67.9 billion, short of the $68.61 billion consensus. RPO is the pipeline—contracted revenue not yet recognized. It's the best forward-looking indicator of whether enterprise customers are actually committing to Salesforce's platform at the rate the company needs. A shortfall here doesn't just disappoint for one quarter; it raises questions about whether the AI narrative is converting into signed contracts at scale.

The Existential Question Hasn't Gone Away

While Salesforce beat on earnings, the broader anxiety driving its 33% decline this year hasn't been resolved. The thesis against the company—and against enterprise SaaS broadly—goes like this: as general-purpose AI models (GPT, Claude, Gemini) become capable enough to handle sales automation, customer service, and data analysis, companies will stop paying premium SaaS license fees and build their own solutions on top of foundation models.

This pressure isn't unique to Salesforce. ServiceNow, Workday, and SAP face the same structural challenge. The incumbents' answer is roughly the same: we have the data, the integrations, and the trust relationships that AI startups don't. Salesforce's specific version of that answer is Agentforce—layer AI agents on top of 20+ years of CRM data and call it a moat.

The 205% growth rate suggests the pitch is landing with some customers. The RPO miss suggests the urgency isn't universal.