The market’s narrative around software stocks has been dramatically reshaped by the rise of artificial intelligence. For a period, a prevailing sentiment, often termed the “SaaSpocalypse,” suggested that AI agents would automate away the core functions of many software companies, leading to a significant erosion of their market value. This perspective, however, may have fundamentally misunderstood the true dynamics at play.
Recent financial results from industry titans like Salesforce suggest a significant recalibration is in order. The initial fear was that sophisticated AI models, particularly large language models (LLMs), would capture an ever-increasing share of AI-driven value, leading to soaring valuations for frontier AI companies. However, the reality is that AI intelligence itself is rapidly becoming commoditized. Frontier models are advancing at an unprecedented pace, with new iterations leapfrogging each other every few months. This intense competition, fueled by well-funded rivals and increasingly capable open-source initiatives, is driving down the cost of raw intelligence. Some analysts even predict a future of recursive self-improvement (RSI) for AI, which could further accelerate these cycles and commoditize intelligence even more rapidly.
In such an environment, the critical question for investors is: where does the enduring competitive advantage, or “moat,” lie? The answer, increasingly, appears to be in **data**. As noted by industry observers, “Lower cost of intelligence increases value of incumbent data.” This aligns with fundamental economic principles: when a crucial input becomes abundant and inexpensive, value naturally migrates to its scarce complements. For AI agents, the indispensable scarce complement is **trusted, proprietary data**.
Consider the practical application: an AI agent tasked with closing a sale still requires a robust system to research the customer, meticulously log every interaction, securely store contracts, and customize specific terms. Without access to reliable and proprietary data, the utility of these advanced AI agents is significantly diminished, embodying the classic dictum: “Junk in, junk out.”
Salesforce, a colossal repository of enterprise customer data, exemplifies this shift. Far from being a victim of a cyclical downturn, the company appears strategically positioned as a long-term structural winner in the commoditized AI landscape. The continuous influx of data into its platform underscores this advantage. Salesforce’s Data 360 service, for instance, processed an astonishing 104 trillion customer records in the recent quarter, a staggering 355% increase year-over-year. Moreover, AI agents themselves are generating more data, which in turn needs to be housed within secure, trusted environments. Salesforce reported facilitating 3.2 billion units of agentic work in the past quarter, nearly doubling from the previous one.
This creates a powerful flywheel effect: AI agents generate work, which in turn generates more data. This expanding data moat enhances the value of each successive generation of AI agents, solidifying the company’s position. This is powerfully reflected in Salesforce’s combined AI and data annual recurring revenue (ARR), which has surged to $3.9 billion, more than tripling in just one year. Its Agentforce, a specific AI offering, has seen its ARR skyrocket from $100 million to over $1.5 billion within 18 months of launch, representing a 240% year-over-year increase.
Salesforce CEO Marc Benioff has been vocal in dismissing the “SaaSpocalypse” narrative, highlighting strong performance metrics. Following a significant surge in the company’s stock, he pointed to record net new average order value (AOV) growth, the strongest in four years. Contrary to skeptical predictions, customer seat growth in sales, service, and Slack all increased year-over-year, with customer attrition near historic lows. Bookings more than doubled quarter-over-quarter, contract terms improved across all segments, and crucially, “Agents are using more Salesforce than ever.”
The financial data strongly supports this optimistic outlook. Instead of diminished pricing power and compressed margins, Salesforce reported non-GAAP operating margins of 34.1%. Adjusted earnings per share nearly doubled, significantly exceeding consensus estimates. Revenue grew by 11% to $11.35 billion, with bookings outpacing revenue growth. The company also raised its full-year guidance, projecting revenues up to $46.4 billion.
Perhaps the most compelling evidence comes from the very companies at the forefront of AI development. If AI agents could truly operate independently of established CRM platforms, leading AI labs would likely be disintermediating them. Instead, they are becoming significant customers. Nine of the top 10 AI companies now rely on Salesforce and Slack, with their combined spending experiencing a remarkable 435% year-over-year increase. This is further exemplified by strategic partnerships, such as the development of “Claudeforce,” a collaboration between Anthropic, a leading AI company, and Salesforce. Benioff aptly described this as “The No. 1 AI in the world, Anthropic, and the No. 1 CRM, Salesforce, coming together.”
These leading AI model developers have concluded that their groundbreaking technologies are not replacements for, but rather dependent upon, robust CRM infrastructure. This underpins Benioff’s assertion that “Salesforce is first and foremost in the data business,” a conviction backed by the company’s historic $25 billion share buyback program.
This paradigm shift is not confined to Salesforce. The broader software industry will likely see a divergence. Companies possessing valuable proprietary data that is essential for AI agents are well-positioned for sustained growth. Conversely, software providers offering only functional utility without strong customer stickiness or unique data assets may face significant challenges. Financial health will also be a critical differentiator. Companies with robust free cash flow and low leverage will have the agility to invest in the AI transition, while heavily indebted firms might be constrained by debt servicing obligations.
Ultimately, the locus of power in the evolving AI economy is shifting. It is moving away from companies whose primary advantage lies in the creation of intelligence – an increasingly abundant commodity – and towards those that own the scarce, indispensable assets that intelligence requires to function. Increasingly, that indispensable asset is data.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/25283.html
