Should You Buy Salesforce Stock Before Feb. 25?

Salesforce shares dropped roughly 8% intraday (about -27% YTD, -43% over 12 months) amid a sector-wide sell-off and investor concerns that Anthropic's new Claude AI plug-in could compress SaaS 'seat' revenue by enabling agentic workflows. Piper Sandler cut its price target to $280 from $315 citing "seat compression" and "vibe coding," even as Salesforce — a Claude partner for its Agentforce chatbot — reported strong AI traction: Agentforce and Data 360 ARR rose 114% YoY to $1.4 billion, Agentforce accounts were up 70% QoQ, RPO increased 11% to $29.4 billion, fiscal 2026 revenue guidance was raised, and the company secured a $5.6 billion U.S. Army contract; earnings are due Feb. 25 after the close, suggesting significant event-driven downside risk but also potential for a post‑earnings rebound.

Analysis

Market structure: The Anthropic/Claude plug‑in dynamic bifurcates winners (AI compute, inference platforms, infrastructure vendors) and losers (seat‑based enterprise SaaS with limited usage pricing). Salesforce (CRM) faces “seat compression” risk despite strong RPO ($29.4B, +11%) and Agentforce ARR $1.4B (+114% YoY); NVDA and AI infra vendors gain pricing power as compute demand rises. Expect demand migration from per‑seat licenses to usage/agent billing over 6–24 months, pressuring gross retention for legacy SaaS unless they reprice.

Risk assessment: Tail risks include rapid regulatory intervention on model access/data (6–18 months), large federal procurement reversals (Army deal rebaseline), or model hallucinations causing liability suits; each could swing CRM ±20–40% beyond base case. Near term (days–weeks) expect elevated volatility around Feb 25 earnings; medium term (quarters) watch seat metrics and Agentforce monetization; long term (years) outcome depends on CRM’s successful pivot to usage pricing and partner economics. Hidden dependencies: CRM revenue quality tied to partner models (Anthropic) and customers’ willingness to pay for Salesforce-specific integrations.

Trade implications: Tactical event trades (earnings straddle on CRM) and structural rotation into NVDA/AI infra are warranted. Options IV on CRM should spike 30–80% into earnings—buying a straddle captures directionally agnostic moves, while using long-dated call spreads limits downside for a constructive multi‑quarter view. Reduce exposure to pure seat‑based SaaS and reallocate to AI compute and defense/contractors with secured federal revenue.

Contrarian angles: The street conflates short‑term plugin headlines with long‑term platform moat erosion; if CRM converts Agentforce growth (>$1.4B ARR) into blended ARPU preservation via usage tiers, current declines (CRM -27% YTD, -43% 12m) may be overdone by 20–40%. Historical parallels: ERP shifts to cloud (2008–2015) punished incumbents then rewarded platform owners who rearchitected pricing—watch CRM’s pricing announcements over 90 days for a potential asymmetric entry.

Key Decisions for Investors

• Buy a tactical earnings straddle on CRM (buy ATM call + put expiring the week of Feb 26), size 0.5–1.0% of portfolio notional; exit within 2 trading days post‑release to capture the IV move and limit directional risk.
• Establish a phased 2–3% long position in CRM only on one of two triggers: (A) post‑earnings beat with FY guide raise (scale 1% immediately, +1% on confirmation of seat/usage metrics), or (B) price weakness—add 1% at −8% intraday from current, +1% at −15%, horizon 6–12 months; set stop to trim if RPO growth falls <5% YoY.
• Implement a relative‑value pair: long NVDA (1.5–3% portfolio) vs short CRM (1–1.5%) to express AI infra outperformance vs seat‑based SaaS; trim NVDA if it rallies >30% or rebalance after next quarter results.
• Reduce aggregate exposure to pure seat‑based SaaS names by 15–25% over 30 days and redeploy into AI compute (NVDA) and high‑visibility federal contractors (allocate ~50% of proceeds); monitor Agentforce ARR growth and CRM pricing announcements over next 90 days as a re‑entry signal.