AI Pincer Movement

Programming note: Our next issue lands 13 March, where we will dig into market data to see if the tech industry's fundamentals are keeping pace with the narrative.

Salesforce's Dilemma

As we discussed recently, the software industry has been crushed by a new market narrative dubbed the "SaaSpocalypse". The logic—at least the logic favored by the people hitting the sell button—is that if AI is as magical as its cheerleaders claim, it will soon become smart enough to replace the expensive enterprise software we currently use to run our businesses.

But if you zoom out, the market is actually grappling with two somewhat contradictory narratives about AI at the same time.

• Narrative A ("AI Eats Software"): AI will destroy the margins and moats of traditional software companies by allowing anyone to "vibe code" their own custom applications.
• Narrative B ("The Monetization Supercycle"): AI is a once-in-a-generation platform shift that will allow incumbent software companies to charge massive premiums for new, high-margin, "agentic" capabilities.

Usually, an enterprise software company only has to deal with one of these narratives. But if you want to understand why Salesforce is currently getting absolutely hammered by the market, it is because it has managed to get caught in a pincer movement between both of them simultaneously.

The Multiple Compression

A year ago, Salesforce was trading at a forward P/E multiple of 28x. Today, it is trading at 15x. The stock has shed more than a third of its value over the last twelve months.

This is the mechanical result of Narrative A. As Goldman Sachs strategist Ryan Hammond noted on a recent podcast, the market has abruptly changed how it values the terminal growth rate of software companies.

"These stocks traded at a forward earnings multiple of about 35 times... As of the latest readings, they trade at about 20 times," Hammond said of the broader software sector. "It seems like investors have gone from valuing this group of stocks in the 15 to 20 percent growth range to 5 to 10 percent growth in the span of a couple of days."

For a company like Salesforce, the fear is palpable. If your core product is a massive, complex database for managing customer relationships, the market is suddenly terrified (probably irrationally) that a startup with a clean interface and an Anthropic API key is going to build a better, cheaper version of your product over a weekend.

Salesforce CEO Marc Benioff recently dismissed these fears entirely, telling investors: "If there is a 'SaaS-pocalypse', it may be eaten by the 'SaaS-quatch' because there are a lot of companies using a lot of SaaS, because it just got better with agents."

But to believe in the "SaaS-quatch," you have to believe that Salesforce is successfully executing on Narrative B: the monetization supercycle.

The Monetization Problem

And this is where Salesforce's problems compound. While one half of the market is dumping the stock because of the existential threat of AI disruption, the other half is selling because they are looking at the actual sales numbers.

Investors are increasingly suspicious that enterprise adoption of AI is nowhere near the level required to justify a clear monetization path. According to recent Census Bureau data, only 4.3% of the 1.2 million businesses it surveyed report using AI in any of their business functions. It is very hard to build an AI monetization case when 95% of the economy hasn't even plugged the thing in yet.

Salesforce's answer to this is a product called Agentforce, which it describes as an "enterprise agentic AI solution." The marketing materials promise that companies can "safely deploy agents that work for their customers, suppliers, and employees 24/7."

But if you look at the actual people tasked with implementing this software—the Salesforce administrators and developers who live in the trenches—the reality of Agentforce is a slow-motion disaster. In a recent Reddit discussion among Salesforce professionals, the feedback was uniformly brutal.

The primary complaint is that Salesforce is trying to staple an unpredictable AI model onto a legacy system that demands absolute precision. As one user noted, a CRM system is "a giant pile of if/then/else" logic that is meant to be 100% predictable, whereas LLMs are fundamentally unpredictable. Attempting to force the two together results in a system that is "convoluted to set up" and "glitchy."

This technical glitchiness can become a business-ending risk when you apply it to Salesforce's core segment. Enterprise customers pay for predictability, especially when it comes to their most valuable asset: customer relationships. You don't want to risk an AI agent going rogue on a $5,000 SMB account, let alone a $5 million enterprise account that funds an entire division. This is the heart of the company's monetization problem: Salesforce is selling a technology defined by its probabilistic nature to a customer base that pays a massive premium for deterministic, career-saving certainty.

The Infrastructure Envy

While software companies like Salesforce are desperately trying to convince the market that they have a monetization path, the market has clearly decided that the real money in AI right now is in the physical infrastructure.

The data from the hardware layer is staggering:

• Broadcom recently projected that its AI chip sales will top $100 billion next year (up from $20 billion in 2025).
• Dell saw its stock jump 17% in a single day after forecasting that its AI server revenue will grow 103% to $50 billion in fiscal 2027.
• Hon Hai (Foxconn), the primary assembler of Nvidia's servers, just reported a 22% climb in sales driven by the global AI buildout.

Investors are happy to reward Broadcom, Dell, and Hon Hai because their revenue is tangible, immediate, and secured by the frantic spending of hyperscalers.

Salesforce, however, is struggling to prove its piece of the AI pie actually exists. It is trying to sell a highly complex product to a market that mostly hasn't plugged AI in yet. The result was an earnings report that leaned heavily on Agentforce marketing while offering a lukewarm long-term revenue outlook. To distract from the glaring absence of AI-driven monetization, management announced a massive $50 billion buyback. It is a classic bit of financial engineering: if you can't convince the market that your AI revolution is real, you can always just use your cash to manufacture demand for your own shares.

More on AI Adoption:

• The State of Enterprise AI (Deloitte)
• 2026 AI in Professional Services Report (Thomson Reuters Institute)
• AI Adoption Research (UK Department for Science, Innovation and Technology)

📊 Data > Narrative

We pull key data points to show you the mathematical reality of what's happening in tech.

Growth Parity, Capital Disparity

We compared the fundamental health of the Hyperscalers (Amazon, Meta, Alphabet, Microsoft, Oracle) against the Top 30 Enterprise SaaS companies.

• Hyperscalers (Median):
  • Revenue Growth (YoY): 16.7%
  • Gross Margin: 65.4% (vs. 3Y Median: 65.9%)
  • Free Cash Flow Margin: 18.2% (vs. 3Y Median: 20.8%)
  • CapEx to Revenue: 27.2%
• Top 30 Enterprise SaaS (Median):
  • Revenue Growth (YoY): 16.7%
  • Gross Margin: 76.9% (vs. 3Y Median: 75.6%)
  • Free Cash Flow Margin: 30.0% (vs. 3Y Median: 28.0%)
  • CapEx to Revenue: 2.0%

The Takeaway: Both groups are growing revenue at exactly 16.7%, but the cost of that growth is drastically different. Hyperscalers are compressing their cash margins to fund a massive 27% CapEx burden, while Enterprise SaaS is quietly expanding margins, against the backdrop of SaaSpocalypse.

P.S. we are launching the Tech Sector Diagnostic—a weekly dashboard tracking exactly how capital, margins, and multiples are flowing across tech companies. Click here for beta access.

You received this message because you are subscribed to ARPU newsletter. If a friend forwarded you this message, sign up here to get it in your inbox.