Salesforce: Proving That AI Needs A Software Partner

Salesforce delivered a strong Q2 beat, with revenue up 11% y/y to $11.35B and pro forma EPS up 102% y/y to $5.90, notably exceeding expectations. Current RPO rose 14% y/y, signaling accelerating demand, while AI momentum was reinforced by Claudeforce’s launch with Anthropic. The partnership and new AI-driven revenue stream prospects should support CRM sentiment and likely move the stock in the near term.

Analysis

This is less about one beat and more about a regime shift in how the market should underwrite enterprise software: AI is no longer automatically a margin headwind or a science project. If a large installed-base vendor can show usage translating into durable demand, the read-through is higher confidence on pricing power and lower fear of seat erosion across the category, which should help the software multiple set broadly rather than just this name.

The second-order winner is the broader AI stack, not just the application layer. Real enterprise adoption implies more inference, more orchestration, and more workflow integration, which is constructive for hyperscalers and infrastructure beneficiaries; the losers are smaller SaaS vendors that rely on "AI roadmap" messaging without proof of monetization. If buyers conclude that AI spend is being captured inside incumbent platforms, point solutions with weaker distribution may face slower net retention and more aggressive pricing pressure over the next 1-3 quarters.

The key risk is that this remains a narrative until management proves it is incremental revenue, not bundled retention. The falsifiers are simple: cRPO growth slowing back toward revenue growth, no full-year guide raise, or evidence that AI attach is lifting costs faster than gross margin. Over 6-18 months, the market will decide whether this is a durable multiple re-rating or a one-quarter relief rally; if AI features do not show up in disclosed monetization metrics, enthusiasm can fade quickly.

Contrarian take: consensus may still be underestimating how negative the setup had become for enterprise software, so even modest proof points can force multiple expansion. The cleaner expression is to own the platform with the clearest demand signal and avoid overpaying for "AI optionality" elsewhere until the data show actual attach rates. If the next print confirms the trend, this can become a leadership name rather than just a sentiment trade.

Key Decisions for Investors

• Long CRM on pullbacks over the next 1-2 weeks; treat this as a 1-3 month multiple-re-rating trade. Falsifier: next-quarter cRPO decelerates materially or management refuses to lift FY guidance.
• Buy a 3-6 month CRM call spread rather than stock if implied volatility remains elevated post-print; structure for defined risk while capturing follow-through from estimate revisions.
• Relative value: long CRM / short a basket of weaker SaaS names with less proven AI monetization (for example HUBS and TWLO) over the next quarter. Thesis: capital rotates toward vendors with embedded workflow depth and away from higher-burn software stories.
• Use MSFT, AMZN, GOOGL, and ANET as second-order AI infrastructure beneficiaries to watch; if CRM AI adoption is real, inference and orchestration spend should follow within 1-2 quarters.
• If CRM fails to hold the post-earnings breakout and the next disclosure shows no improvement in AI-related attach or margin, de-risk aggressively; the move would then be largely sentiment-driven rather than fundamental.