How Does Snowflake Make its Money?
Snowflake operates a cloud-based data platform that allows organizations to store, process, and analyze large volumes of data. Unlike traditional software that charges flat subscription fees, Snowflake uses a consumption-based pricing model — customers pay only for the compute and storage resources they actually use. This creates revenue that scales directly with how much data customers process.
Snowflake’s platform runs on top of AWS, Azure, and Google Cloud (all three major public clouds), giving customers flexibility to run workloads wherever their data lives.
Revenue Breakdown
| Revenue Stream | FY2025 (Jan) | FY2024 (Jan) | YoY Growth |
|---|---|---|---|
| Product Revenue | $3.43B | $2.67B | +28.5% |
| Professional Services & Other | $0.22B | $0.19B | +15.8% |
| Total Revenue | $3.65B | $2.86B | +27.6% |
Product Revenue — 94% of Revenue
Product revenue is the consumption-based core of Snowflake’s business. Customers use Snowflake credits to pay for:
- Compute: Running SQL queries, data transformations, and ML workloads. Priced per “virtual warehouse” (compute cluster) per second of usage.
- Storage: Storing structured and semi-structured data in Snowflake’s managed storage layer. Priced per terabyte per month.
- Data Transfer: Moving data between cloud regions or providers.
- Snowpark (data engineering): Running Python, Java, and Scala code natively within Snowflake.
- Snowflake Cortex (AI/ML): Running AI models, vector search, and LLM functions directly on Snowflake data.
- Snowflake Marketplace: A data marketplace where third-party providers list datasets and services.
The consumption model means Snowflake’s revenue is driven by customer data growth and usage intensity rather than seat count. When customers run more queries, build more dashboards, or train more models, Snowflake earns more.
Professional Services — 6% of Revenue
Implementation support, training, and optimization consulting. This segment is intentionally small and low-margin — Snowflake prefers partners (Deloitte, Accenture, etc.) to handle implementation work.
Income Statement Overview
| Metric | FY2025 | FY2024 |
|---|---|---|
| Total Revenue | $3.65B | $2.86B |
| Cost of Revenue | $1.32B | $1.09B |
| Gross Profit | $2.33B | $1.77B |
| Operating Expenses | $3.20B | $2.74B |
| Operating Income | -$0.87B | -$0.97B |
| Net Income | -$0.83B | -$0.84B |
Key Financial Metrics
- Product Gross Margin: 73.5% — Strong for a cloud-native platform. The main cost is cloud infrastructure (AWS/Azure/GCP) where Snowflake is a tenant. Margins have been improving as Snowflake negotiates better cloud pricing at scale.
- Operating Margin: -23.8% — Still unprofitable on a GAAP basis, primarily due to massive stock-based compensation (~$1.5B). On a non-GAAP basis (excluding SBC), Snowflake achieved a 7% operating margin.
- Revenue Growth: +27.6% — Solid but decelerating from 36% the prior year. Growth deceleration is Snowflake’s most debated metric among investors.
- Net Revenue Retention Rate: 127% — Existing customers increased their spending by 27% year-over-year. This high retention rate means Snowflake can grow significantly even without acquiring new customers.
- Remaining Performance Obligations (RPO): $6.9B — Committed future spend from existing contracts, providing strong revenue visibility.
Where Does Snowflake Spend its Money?
- Cloud Infrastructure (~$1.1B): Paying AWS, Azure, and GCP for the compute and storage that powers the Snowflake platform. This is the largest cost and directly tied to customer consumption.
- R&D (~$1.83B): The largest operating expense. Snowflake employs some of the highest-paid engineers in enterprise software, building Cortex AI, Snowpark, Iceberg Tables, and performance optimizations.
- Sales & Marketing (~$1.48B): Enterprise sales is expensive — large company deals involve long sales cycles, solution architects, and proof-of-concept engagements. Snowflake’s sales team targets Fortune 500 and Global 2000 companies.
- Stock-Based Compensation (~$1.5B): The elephant in the room. SBC is ~41% of revenue, among the highest ratios in enterprise software. This is the primary reason GAAP profits remain deeply negative.
- G&A (~$0.40B): Legal, finance, and administrative functions.
What to Watch
- Consumption growth reacceleration — Snowflake needs existing customers to increase usage meaningfully. AI workloads (Cortex, vector search) are expected to drive the next wave of consumption growth.
- AI and Cortex adoption — Snowflake Cortex puts AI capabilities directly in the data platform. If customers adopt Cortex for inference and fine-tuning instead of going to separate AI platforms, it could drive significant incremental consumption.
- SBC trajectory — Stock-based compensation at 41% of revenue is extraordinarily high. Investors want to see this ratio declining toward 20-25% as the company scales.
- Competition — Databricks (private) is Snowflake’s most direct competitor, offering a similar data lakehouse platform. Google BigQuery, Amazon Redshift, and Microsoft Fabric also compete. The data platform market is intensely competitive.
- Iceberg Tables and open formats — Snowflake is embracing Apache Iceberg, an open table format. This reduces data lock-in but could lower switching costs. Snowflake bets it can win on performance and ease of use even without proprietary formats.