Salesforce Just Had Its Worst Day in Years — and a $25 Billion Buyback Made It Worse

ServiceNow's CEO admitted the company got hit by delays tied to the Middle East conflict. But he said the impact was mostly about timing — not a drop in actual demand. Still, the conservative guidance was enough to trigger the stock's worst single-session decline on record.

The contagion hit Salesforce harder than a generic sector rotation would explain. Both companies sell to the same types of big customers: large multinationals, government-linked organizations, and companies with complex global buying teams. When geopolitical disruption freezes purchasing cycles, it tends to hit these buyers first because their approval chains are longest and most sensitive to headline risk.

That overlap matters. Salesforce earns a lot of revenue from multinational and public-sector accounts. These are exactly the buyers most likely to delay or rework software deals when geopolitical uncertainty spikes. The market didn't just reprice CRM on generic fear. It repriced CRM on the specific bet that Salesforce's pipeline contains the same frozen deals ServiceNow just disclosed.

The debt doesn't go away. Salesforce added long-term obligations to its balance sheet to buy shares at higher prices. If the stock stays down — or falls further — shareholders are left with more debt and diluted returns on the capital that was spent.

This matters because the buyback was supposed to signal strength. Instead, it locked in a cost basis that the market has already moved below. The bonds maturing in 2066 mean Salesforce will be servicing this debt for forty years regardless of where the stock trades next quarter.

Management may ultimately be vindicated if the stock recovers over the next 12–18 months. But in the near term, the optics are brutal: billions in new debt, shares bought at higher prices, and a stock sitting near its trailing 52-week low.

Salesforce's Agentforce platform — its flagship AI product — is built on a pricing model that assumes AI augments human workers rather than replaces them. More AI agents means more software seats, more licenses, more revenue per customer. The unit economics — how much money you make (or lose) on a single sale depend on headcount staying stable or growing.

But large enterprises are signaling the opposite. Major tech companies have announced significant workforce reductions citing AI-driven efficiency gains. If other companies follow that template, the assumption baked into Agentforce's pricing breaks. Fewer employees means fewer seats. And AI starts hurting per-seat revenue instead of helping it.

ServiceNow's conservative guidance showed that geopolitical disruption can freeze purchasing cycles regardless of how compelling the AI pitch is. Salesforce, with a workforce of approximately 83,000 as of its fiscal year-end in January 2026, is betting that its customers will use AI to do more — not to employ fewer people. If the market starts to believe the opposite, the valuation math changes dramatically. The timeline for when AI investment turns into SaaS revenue growth could get pushed out by quarters, not weeks.

Salesforce issued bonds with maturities ranging from 2028 to 2066 just six weeks before the sell-off. Those bonds were priced when the stock was trading significantly higher and the outlook was rosier. The bonds almost certainly contain change-of-control provisions — standard clauses that give bondholders the right to demand early repayment if a new party gains effective control of the company.

That creates a structural conflict. Say an activist investor builds a big enough stake and pushes for dramatic changes — a breakup, a forced sale of a division, or new board members. Bondholders could argue those moves trigger change-of-control protections. The activist's equity play and the bondholders' contractual protections can collide.

Salesforce's cash-flow generation is strong enough that a credit crisis is unlikely. But the mix of new debt and potential activist pressure creates a limit. No outside investor can push too hard for big structural changes without running into it. And that limit didn't exist before March.

The stock trades at a historically compressed valuation relative to its expected earnings. It has fallen sharply from its 52-week high as of the time of writing. And the company has deeply diversified revenue — no single customer accounts for more than 10% of sales, according to its most recent annual filings.

Activists prize exactly this profile: recurring revenue, depressed valuation, no single customer over 10% of sales. Funds that have run campaigns during enterprise-software sell-offs before — names like Elliott Management, Starboard Value, and ValueAct — have targeted companies with this exact setup again and again. A full acquisition is unlikely at this market capitalization. But an activist campaign pushing for cost cuts, strategic divestitures, or a change in capital allocation is very much on the table.

The irony is that the $25 billion buyback was supposed to preempt exactly this kind of pressure. Instead, going underwater may have made the activist case stronger. Management spent $25 billion buying stock at a price the market has already rejected. And the new bond covenants from Domino 4 may cap how far any activist can push before triggering bondholder protections. That tension between equity activists and debt holders is the real story the market hasn't priced.