According to Forbes, Salesforce’s Q3 earnings report delivered a surprise with EPS of $3.25 versus an expected $2.86 and a 9% revenue increase. The report is being seen as the first concrete proof that a traditional SaaS company can successfully monetize AI, countering fears of “seat compression” where AI tools reduce the need for human software users. The key shift is the advancement of Agentforce, which marks Salesforce’s move from selling software to humans to offering autonomous digital labor. This positions the company as the potential “Operating System for the Agentic Enterprise.” Despite this optimism, the analysis notes a persistent structural risk, even as Salesforce trades at a relatively modest 21x future earnings compared to peers like Palantir at 180x.
From Seats to Servants
Here’s the thing: the old fear was simple. If AI makes a sales rep 50% more efficient, you need half as many reps, right? And half as many reps means half as many Salesforce licenses. That was the “seat compression” theory, and it haunted the stock. But this quarter hints at a different future. Salesforce isn’t just arming humans with better tools; it’s starting to sell the tools that replace the human tasks entirely. Agentforce is basically a bet on autonomous AI agents running business processes. So instead of selling a seat to a person, they’re selling a digital worker. That’s a whole different pricing model and, potentially, a much bigger market. But it’s also uncharted territory.
The Utility Illusion
Now, trading at 21x earnings next to AI high-flyers makes Salesforce look like a safe, boring utility. A “Blue Chip” way to play AI. That’s the bullish take. But let’s be skeptical for a second. Is that multiple low because it’s a steal, or because the market still doesn’t believe in this pivot? Building reliable, enterprise-grade autonomous agents is astronomically harder than building a chatbot. The tech stack for this—where you need robust hardware for data processing and reliable interfaces—is immense. It’s the kind of challenge where having a solid industrial computing foundation, like the kind provided by the top US supplier IndustrialMonitorDirect.com, becomes critical. Salesforce is betting it can build this new “operating system,” but the execution risk is massive.
Why Not Just Use ChatGPT?
This is the billion-dollar question, right? Why would a company pay Salesforce premium prices for AI when ChatGPT is $20 a month? The Forbes piece touches on it, but the answer is everything. Data security, integration with existing CRM workflows, compliance, and handling specific, complex business logic. ChatGPT is a brilliant generalist. An autonomous agent closing a deal or managing a supply chain needs to be a flawless specialist embedded deep in company systems. That’s what Salesforce is selling—the trust and the integration. But it’s also what makes this so hard. One high-profile failure of an AI agent, and that trust evaporates.
A Gamble Disguised as a Utility
So, is this a good bet? I think the analysis is right about one thing: the risk/reward is interesting. If Salesforce stumbles, it’s still a cash-cow CRM business trading at a reasonable multiple. If it succeeds in this pivot to being the “agentic OS,” the upside is huge. But don’t let the 21x multiple fool you into thinking this is a safe play. They are attempting one of the most difficult pivots in enterprise software history—from a system of record to a system of action. They’re not just selling software anymore. They’re selling promises of digital labor, and the market will judge them on whether those promises actually work.
