Why Salesforce Plunged Over 40% in the First Half of 2026

Salesforce shares fell 40.9% in H1 2026 amid investor fears that agentic AI could disrupt traditional enterprise software, despite the company beating revenue/earnings expectations and raising the low end of FY2027 guidance. Management countered with AI expansion (Agentforce reaching a $3.4B annualized run rate) and multiple tuck-in acquisitions, including the $3.6B purchase of Fin. Confidence was reinforced by a $25B accelerated share repurchase (part of a $50B authorization) that reduced shares outstanding by ~10% over days, though it increased debt.

Analysis

The market is treating agentic AI as a structural margin compression event for horizontal SaaS, but the more likely near-term effect is a mix shift: license revenue may slow while usage, outcomes, and integration layers take more of the wallet. That favors platforms with deep workflow embedding and proprietary customer data, which is why CRM is better positioned than the blanket selloff implies. The buyback helps cushion EPS, but it is not a substitute for proof that AI monetization can offset slower seat growth; the added debt also reduces flexibility if growth wobbles.

Over the next 1-3 quarters, the key catalyst is not generic AI messaging but whether management can show Agentforce translating into measurable net retention, cRPO stability, and a visible path to consumption pricing. If that evidence is absent, the stock can stay trapped in a de-rating regime even if reported results remain “good enough,” because investors will look through current earnings to future monetization risk. The second-order beneficiaries are less obvious than the losers: metering/billing infrastructure and outcome-based software models should gain pricing power as buyers demand pay-for-performance contracts.

Contrarian view: the selloff may be overdone because AI is more likely to unbundle low-complexity features than replace the systems of record where compliance, identity, and integrations matter. CRM’s real moat is not model quality; it is distribution, switching costs, and the ability to bundle AI into existing enterprise spend. What would falsify the bullish case is a quarter or two of decelerating RPO/cRPO plus no evidence that AI attach rates are expanding enough to offset seat pressure; at that point, the low multiple is just a value trap with leverage attached.

Key Decisions for Investors

• Initiate a starter long in CRM on further weakness over the next 2-6 weeks, targeting a 3-6 month hold; risk/reward is attractive only if the stock stays near a low-teens earnings multiple and the next print shows cRPO or RPO stabilization.
• Pair trade: long CRM / short a higher-beta software name with thinner workflow moats such as ZM over the next 1-3 quarters; thesis is that AI compresses standalone communication layers faster than integrated systems-of-record platforms.
• Do not add to CRM ahead of the next earnings cycle unless management quantifies Agentforce attach, usage revenue, or consumption-pricing adoption; absent that proof, treat the buyback as a floor, not a catalyst.
• Set a thesis-falsifier alert on CRM: if cRPO/RPO growth slips for two consecutive quarters or net leverage rises without offsetting FCF growth, reduce exposure by 50% and reassess.
• Watch the broader SaaS complex for a relative-value rotation into metering/billing and outcome-based software; if consumption pricing becomes the dominant narrative, the best long expression may be infrastructure names rather than the most obvious AI software beneficiaries.