SNOW Stock - Snowflake Inc.

Under the leadership of CEO Sridhar Ramaswamy, Snowflake has undergone a fundamental strategic pivot, repositioning itself from a cloud data warehouse to an AI Data Cloud. As of February 2026, the company is transitioning into Fiscal Year 2027 (FY27), following a year (FY26) defined by the rapid monetization of generative AI (GenAI) capabilities via Snowflake Cortex.

Revenue Growth Attribution: GenAI vs. Core Warehousing

For the upcoming fiscal year, the attribution of product revenue growth remains weighted toward core data activities, though GenAI has emerged as the primary growth catalyst and a significant direct contributor.

• Direct GenAI Contribution: In late FY26, Snowflake achieved a $100M AI revenue run rate, reaching this milestone one quarter earlier than internal projections. While this represents approximately 2-3% of total product revenue, it is the fastest-growing segment within the portfolio.
• Core Data Warehousing & Engineering: The "core" business—comprising data warehousing, data engineering (Snowpark), and data sharing—continues to drive the vast majority of the projected 27-28% year-over-year product revenue growth. Management has indicated that core workloads remain the "bread and butter" of consumption, providing the foundational data layer required for AI applications.
• The "AI Influence" Multiplier: A critical metric for the upcoming year is the "influence" factor. Management reported that AI capabilities influenced nearly 50% of new bookings in the second half of FY26. This suggests that while GenAI-specific consumption (Cortex credits) is a smaller direct line item, it is the primary driver for new customer acquisition and core storage/compute expansion.

Strategic Shift Under Sridhar Ramaswamy

Ramaswamy’s tenure has been marked by an "innovation overdrive," focusing on reducing the friction between data storage and AI execution.

• Snowflake Cortex & Intelligence: The launch of Cortex AI and Snowflake Intelligence has shifted the narrative from passive data storage to active "agentic" AI. Over 1,200 customers are already using agentic AI capabilities in production, driving a "second wave" of consumption credits.
• Product Velocity: Snowflake released approximately 250 new capabilities in the first half of FY26 alone. This velocity is designed to capture "AI-ready" workloads that would otherwise migrate to hyperscalers (AWS, Azure, GCP) or specialized AI platforms.
• Open Data Strategy: The aggressive push into Apache Iceberg and the acquisition of Crunchy Data (Snowflake Postgres) reflect a shift toward an open, interoperable ecosystem, aiming to prevent vendor lock-in while keeping Snowflake at the center of the compute layer.

Financial Implications & Risks

The transition to an AI-first model introduces new dynamics to Snowflake’s financial profile:

• The "AI Tax" on Margins: GenAI workloads are significantly more compute-intensive and rely on expensive GPU infrastructure. This has led to a "margin paradox" where high revenue growth is met with pressure on non-GAAP operating margins, currently guided in the 8-9% range for the full year.
• Consumption Volatility: Unlike traditional SaaS subscriptions, Snowflake’s model is purely consumption-based. While AI drives higher credit usage, it also makes revenue more sensitive to customer "optimization" efforts, where enterprises tune models to be more efficient, potentially slowing revenue growth in the short term.
• Net Revenue Retention (NRR): NRR has stabilized around 125%. For the upcoming year, the company’s ability to maintain or grow this figure depends on whether GenAI consumption can offset the natural maturation and price-performance improvements in core SQL warehousing.

Summary of Projected FY27 Outlook

The enterprise adoption of open table formats, specifically Apache Iceberg, represents a fundamental shift in the data architecture landscape. For Snowflake, this transition is a "double-edged sword" that simultaneously threatens its traditional storage-based "data gravity" while opening new avenues for compute-driven growth.

📊 Impact on Storage Revenue: The "Headwind" vs. "Offense"

Traditionally, Snowflake’s business model relied on customers ingesting data into its proprietary Internal Stage (FDN) format. Open formats like Iceberg allow enterprises to keep data in their own cloud storage (e.g., AWS S3, Azure Blob) while using Snowflake only for compute.

• Direct Revenue Headwinds: Snowflake management has explicitly identified Iceberg adoption as a headwind for storage revenue. In recent fiscal cycles, storage has consistently accounted for approximately 11% of Snowflake’s total consumption revenue. As more customers migrate to "Iceberg Tables," this high-margin storage revenue is expected to decline or stagnate.
• The "Offense" Strategy: CEO Sridhar Ramaswamy has noted that Iceberg allows Snowflake to "play offense" by querying data that was previously "dark" or trapped in external data lakes. In Q3 FY2025, Snowflake reported that approximately 500 accounts had already adopted Iceberg, often bringing in new workloads that were never previously on the platform.
• Compute Preservation: While storage revenue may face pressure, the more lucrative compute revenue (which drives over 90% of the bill for most customers) remains the primary target. By supporting Iceberg natively, Snowflake ensures it remains the preferred "query engine" even if it no longer "owns" the physical bits.

🔐 Evolution of Customer Lock-in: From Storage to Governance

The "lock-in" mechanism is evolving from Data Gravity (the difficulty of moving petabytes of proprietary data) to Governance Gravity.

• Polaris & Horizon: Snowflake’s launch of the Polaris Catalog (an open-source Iceberg REST catalog) and Snowflake Horizon (a governance layer) is a strategic move to maintain control. Even if data is in an open format, customers often remain "locked in" to the security, access control, and metadata management policies defined within Snowflake's ecosystem.
• Interoperability as a Feature: By open-sourcing Polaris, Snowflake is attempting to prevent a "Betamax vs. VHS" scenario where a competitor’s catalog becomes the industry standard. This allows Snowflake to claim "zero vendor lock-in" at the storage layer while entrenching itself as the indispensable Management Plane.

⚔️ Competitive Dynamics: Databricks and Microsoft Fabric

The competition has intensified as rivals also pivot toward open standards to erode Snowflake’s market share.

• Databricks & Tabular: Databricks’ acquisition of Tabular (founded by Iceberg’s creators) for a rumored $1B - $2B was a defensive move to bridge the gap between its native Delta Lake and the surging Iceberg format. This acquisition aims to unify the two formats, potentially making Databricks a more versatile choice for enterprises seeking a "Lakehouse" architecture.
• Microsoft Fabric (OneLake): Fabric poses a significant threat through its OneLake "OneDrive for data" concept. Its Direct Lake mode allows Power BI to query data in Delta Parquet format without any data movement, offering a level of integration that Snowflake struggles to match for Microsoft-centric organizations.
• Market Positioning: Snowflake is countering by positioning itself as the "Gold Standard" for SQL performance and ease of use, recently reporting Q3 FY2026 product revenue of $1.16B, a 29% YoY increase, despite the competitive noise.

⚠️ Risks and Long-term Outlook

The primary risk for Snowflake is the commoditization of the query engine. If Apache Iceberg makes data truly portable, the "switching cost" for an enterprise to move from Snowflake to a cheaper or more integrated engine (like Databricks or a hyperscaler's native tool) drops significantly.

• Margin Compression: To compete with "open" alternatives, Snowflake may be forced to offer more aggressive pricing or discounts on long-term contracts, which was cited as a factor in its -8% stock decline following its December 2025 earnings report.
• AI as the New Moat: Snowflake is betting that its Cortex AI and Snowflake Intelligence (agentic AI) will provide a new layer of value that open-source engines cannot easily replicate, shifting the competitive moat from "where the data sits" to "how intelligently the data is used."

Following the stabilization of cloud consumption patterns, Snowflake (SNOW) has demonstrated a pivot toward large-scale enterprise commitments, particularly within the Global 2000 (G2k) segment. As of the latest fiscal reporting and strategic announcements in early 2026, several leading indicators suggest a meaningful acceleration in Remaining Performance Obligations (RPO) over the next two quarters (Q4 FY26 and Q1 FY27).

📈 RPO Acceleration & G2k Deal Dynamics

The primary leading indicator for RPO acceleration is the recent shift in "mega-deal" velocity. In Q3 FY26, Snowflake reported that RPO growth accelerated to 37% YoY, reaching $7.88B. This outpaced product revenue growth of 29%, signaling that forward-looking contract value is building faster than current recognition.

• Nine-Figure Commitments: Management disclosed closing 4 "nine-figure" deals (contracts exceeding $100M) in the most recent quarter alone. These large-scale G2k renewals typically carry multi-year durations (often 3–5 years), providing a massive "step-function" increase to the RPO backlog.
• G2k Concentration: The G2k customer count reached 776, with an average trailing 12-month (TTM) spend of $2.3M per account. The "land-and-expand" motion is now moving into a "standardization" phase where G2k entities are consolidating disparate data silos onto Snowflake to prepare for AI workloads.
• Consumption-to-Booking Conversion: A critical leading indicator is the behavior of large customers who had previously been "purchasing as they consume" to manage budgets. Management noted a transition where these accounts are now returning to large multi-year Enterprise Discount Programs (EDPs) to lock in better pricing for anticipated AI-driven compute needs.

🤖 The AI Catalyst: From Experimentation to Production

The transition of AI from "pilot" to "production" is the most significant driver of RPO in the G2k segment. Snowflake achieved a $100M AI revenue run rate one quarter earlier than anticipated, suggesting that the pipeline for AI-related RPO is hardening.

• OpenAI Strategic Partnership: In February 2026, Snowflake signed a landmark $200M multi-year partnership with OpenAI. This deal natively integrates frontier models (including GPT-5.2) into Snowflake Cortex AI and Snowflake Intelligence. For G2k customers, this removes the "friction of movement," allowing them to commit to larger RPOs because they can now run high-reasoning agents directly on their governed data.
• Agentic AI Adoption: Over 1,200 customers have already adopted Snowflake Intelligence for agentic workflows. Because these agents require high-concurrency compute, G2k customers are signing larger RPO contracts to ensure they have the "credits" necessary to scale these agents across their global workforces.

🛠 Strategic Expansion & Vertical Maturity

Snowflake’s recent M&A and partnership activity acts as a leading indicator for RPO by expanding the "surface area" of what a G2k customer can commit to.

• Observe Acquisition: The ~$1B acquisition of Observe (January 2026) allows Snowflake to capture observability and AIOps workloads. This enables G2k CIOs to consolidate their monitoring budgets into their Snowflake RPO, effectively increasing the "wallet share" per contract.
• SAP & Vertical Integration: The landmark partnership with SAP to unite mission-critical business data with the AI Data Cloud is driving RPO in the Financial Services and Manufacturing verticals. G2k companies in these sectors are signing larger commitments to facilitate "Zero-ETL" access to SAP data for real-time analytics.
• FedRAMP High Authorization: Achieving FedRAMP High status has unlocked a pipeline of large-scale public sector and highly regulated G2k contracts that were previously restricted, providing a new tailwind for RPO in the first half of 2026.

⚠️ Risks & Analytical Limitations

While RPO is accelerating, it is not a deterministic forecast of revenue. Snowflake’s usage-based model means that even if a G2k customer signs a $100M RPO contract, revenue is only recognized as they consume credits.

• Consumption Lag: There is often a 6–12 month lag between a large RPO booking and the peak consumption ramp.
• Macro Sensitivity: While G2k customers are stable, any significant macro downturn could lead to "optimization" cycles where customers slow down their consumption of the RPO they have already committed to.