The shares had been written off by a chorus of bears who argued that artificial intelligence would hollow out the software industry, that customers would stop paying for seats when agents could do the work. On Tuesday night, Salesforce pushed back. The company reported revenue and profit growth for its fiscal second quarter, raised its full-year outlook, and the stock jumped about 13 percent in extended trading, the kind of move that resets a narrative.
The numbers were solid without being spectacular. Revenue rose in the mid-teens from a year earlier, profit margins expanded, and the company lifted its full-year guidance by a modest margin. The details mattered less than the direction: Salesforce grew, its subscription base held, and its customers kept paying. For an industry that had spent months arguing that AI would end the SaaS model, that was the story.
Salesforce paired the earnings with a product announcement designed to answer the bears directly. The company is deepening its partnership with Anthropic and launching Claudeforce, an offering that embeds Claude, Anthropic’s flagship model, directly into Salesforce’s customer relationship management software, where it can execute sales tasks, handle customer service inquiries, and act on customer data without leaving the platform. The two companies framed it as a marriage of the No. 1 CRM vendor and one of the top-rated AI models, a combination each described as a way to turn AI from a feature into a workforce.
The strategy is a throwback to the playbook that built the cloud era. Salesforce rose by becoming the platform on which its customers’ businesses ran, then extended that position by letting partners build on top of it. The Claudeforce deal applies the same logic to AI: rather than selling access to a model, Salesforce sells the integration, the workflow, and the data, with the model as the engine underneath. Anthropic gets distribution into Salesforce’s customer base, and Salesforce gets a model it can call its own without building one.
The timing is deliberate. The “SaaSpocalypse” thesis, the idea that AI agents would replace software subscriptions and destroy SaaS revenue, has been the industry’s dominant fear for more than a year. Salesforce’s stock had been punished along with the rest of the sector, and its earnings were seen as a referendum on whether the thesis is real. By pairing growth with a concrete AI product, the company is arguing that the opposite is true: that AI makes its software more valuable, not less, because agents need the data and workflow infrastructure that CRM systems provide.
The partnership also reflects a shift in how AI companies are being integrated into the software economy. Earlier deals treated models as a layer beneath the application, invisible to the customer. Claudeforce puts Claude at the center of the product, named in the offering, marketed by both companies, and tied to specific outcomes like faster sales follow-up and shorter service resolution times. Anthropic, which has been signing large compute and distribution deals across the industry, gains a flagship enterprise partner in a market where OpenAI had appeared dominant.
Investors will watch whether the growth holds. Salesforce’s guidance assumes the AI push converts into renewals and new seats, and the company’s history is full of product announcements that took longer to monetize than the stock wanted. The after-hours reaction suggested the market is willing to give the company credit for showing up with a credible answer to the SaaSpocalypse question, even if the revenue from Claudeforce will not show up in numbers for quarters.
The quarter also signals how the AI industry is consolidating around a few pairings. Microsoft has its own copilots embedded across Dynamics and Office. Google has infused Gemini into its enterprise suite. Salesforce’s decision to anchor its AI story to an outside model, rather than building one or licensing quietly, is a bet that the model itself is becoming a commodity and the integration is where the value lives. That is the same logic that drove the cloud platforms to open their marketplaces, and it suggests the next phase of enterprise software will be fought less over who has the best model than over who has the best grip on the customer’s data and workflows.
None of this is guaranteed. Salesforce’s history includes acquisitions and initiatives that promised to transform the company and delivered modest returns, and its stock remains below the levels it reached during the pandemic’s software euphoria. The 13 percent jump, large as it was, brought the shares only back to where they traded before the SaaSpocalypse fears took hold. What the quarter does is buy time and credibility: time for Claudeforce to generate real revenue, and credibility with investors who had stopped listening to the company’s growth story. Whether that translates into sustained outperformance depends on the product’s execution, which will be tested in the fiscal third quarter, when the first sales cycles tied to the Anthropic partnership close.
The broader takeaway is about the SaaS model itself. For a year, the industry’s skeptics argued that AI would compress software revenue, that seats would shrink, and that the platform layer would be disintermediated. Salesforce’s quarter offers evidence for the opposite view: that the data, the workflows, and the relationships embedded in enterprise software are exactly what AI needs to be useful, and that the companies that own those assets are the ones that will capture the value. The stock’s 13 percent jump was a vote on that argument, and the next few quarters will show whether it was right.
