Salesforce Expands AI Bundles: Will Higher Adoption Boost ARR?

Salesforce's Agentforce ARR exceeded $1.5 billion in fiscal Q2, up more than 240% year over year, while combined Agentforce and Data 360 ARR approached $3.9 billion, up more than 210%. Bookings for Agentforce bundle editions more than doubled sequentially and premium Slack upgrades tripled, supporting the case for broader AI-driven recurring revenue expansion. However, AI momentum has not yet accelerated companywide growth materially: Q2 revenue rose 11% to $11.35 billion and fiscal 2027 revenue growth guidance remains 11%-12%, amid intensifying AI-bundle competition from Microsoft and Oracle.

Analysis

CRM’s opportunity is less the initial AI SKU sale than bundling-driven expansion: attaching collaboration, analytics, security and support converts a discretionary pilot budget into a broader platform contract with lower churn. The key underwriting question is whether AI attach raises net revenue retention and remaining performance obligations faster than discounting offsets it; aggregate AI ARR alone is insufficient because it may largely cannibalize existing clouds or carry elevated inference costs. Over the next 1-3 months, quarterly disclosures on paid deployments, pricing realization and operating-margin progression matter more than usage metrics.

Competition raises the bar materially. MSFT can subsidize Copilot adoption through its installed productivity base and enterprise agreements, while ORCL’s consumption construct may be more naturally aligned with variable agent workloads; both can pressure CRM’s seat-based monetization and force bundle discounts. Conversely, CRM’s cross-cloud data layer and workflow ownership make it a credible beneficiary where customers prioritize customer-service automation over generic productivity AI. This is likely a winner-take-most-by-workflow market, not a uniform enterprise-software uplift.

The valuation discount creates asymmetric upside only if management demonstrates that AI is incremental to consolidated growth rather than a mix shift within the installed base. A credible reacceleration in subscription revenue or RPO, alongside stable-to-expanding margins, could support multiple normalization over 6-12 months; failure to show this by the next two earnings cycles leaves CRM exposed to a value trap, especially given an earnings profile that appears to decelerate after FY27. The contrarian view is that investors are correctly skeptical of reported AI ARR, but may be underpricing the potential reduction in churn and sales-cycle friction if bundles become the default renewal motion.

Key Decisions for Investors

• Initiate a 6-12 month long CRM / short IGV pair at a modest 1:1 beta-adjusted size. Target 15-20% relative upside if CRM proves AI-driven RPO and subscription-growth acceleration; exit if the next two reports show no acceleration in remaining performance obligations or if operating margin falls more than 150 bps year over year.
• Use the next earnings print as the entry catalyst rather than chasing AI-usage headlines. Add only if management quantifies paid Agentforce customers, renewal/expansion contribution, and gross-margin or inference-cost discipline; absent these data, maintain a watch position rather than a directional long.
• For a more defensive expression, own CRM versus ORCL over the next 3 months only if CRM’s enterprise renewal metrics improve. ORCL’s cloud multiple is more vulnerable to capex, capacity-delivery and consumption-volatility disappointments; cover the relative trade if Oracle demonstrates sustained cloud growth without a further deterioration in free-cash-flow conversion.
• Do not position long MSFT solely on Copilot seat growth: monitor paid-seat monetization, incremental ARPU and commercial remaining-performance-obligation growth at the next earnings release. A weaker-than-expected conversion from bundled entitlements to paid usage would challenge the broader AI-software valuation complex, including CRM.