Snowflake (SNOW)

Cloud data warehouse/analytics platform (Snowflake) — query massive data for BI/AI; runs on hyperscalers

AI exposure 23% — derived from reported segment revenue, not asserted.

Where it sits

Software Stack → Data & Infrastructure Software · figures rebased to Q2 FY2027, reported 2026-09-02

Snowflake Data Cloud — warehouse/lakehouse storage and compute, data sharing, Cortex AI, consumption-priced — leads 2 businesses in this sub-tier that together are 100% of revenue, so Data & Infrastructure Software is effectively the whole company rather than one leg of it.

Reported segments

• Product (consumption) — 96.4% of revenue
• Professional Services — 3.6% of revenue

Snowflake ($1,546.8M Q2 FY2027 total, +35% YoY; product $1,491.9M, +37%). Product consumption 96.4% = primary B5 Data & Infrastructure Software — the data cloud, with Cortex AI and the CoCo/CoWork agent products driving a third consecutive quarter of acceleration; Professional Services 3.6% ($54.9M, +0.8%, effectively flat). Consumption-priced, so AI workloads show up directly in revenue rather than as seat expansion. Remaining performance obligations $9.00B grew 30%, against 38% a quarter earlier — the forward book decelerating while recognised revenue accelerates.

How the 23% is built

• Product (consumption) — 96.4% of revenue × 24% AI-driven (estimate). Raised from 20%. Product revenue accelerated to 37% from 32% a year earlier with AI named as the step-up, CoCo passed 9,100 accounts and CoWork 5,800, and Cortex Sense and the Cortex AI Gateway extended the surface — on a base that is still mostly conventional data warehousing
• Professional Services — 3.6% of revenue × 5% AI-driven (estimate). Implementation on the same platform, and the line grew 0.8% while product grew 37%

Investment read

Moat. Workloads, governance and sharing relationships create real switching costs — net revenue retention is 126%, 828 customers now spend above $1M of trailing product revenue, and 692 net new customers landed in the quarter. But the data-gravity component is weaker than it was: Apache Iceberg's near-universal adoption makes the storage layer portable, so lock-in now has to come from compute, governance and the sharing network rather than from custody of the data itself.

Bottleneck / pricing power. Consumption pricing with the AI attach behind it — non-GAAP product gross margin held at 74.7% while product revenue accelerated to 37% and full-year operating margin guidance rose from 13.5% to 14.5%, which is what pricing power looks like in a usage model. The limit is that Databricks contests the same workloads directly and hyperscalers bundle analytics at near-zero incremental cost.

Role in the AI stack. The cloud-data-warehouse + AI-analytics layer.

Connected companies

• Amazon (AMZN) — SNOW is supplied by: AWS compute/storage underpins Snowflake's data cloud
• Datadog (DDOG) — SNOW competes with: Adjacent cloud-data platforms; Snowflake's Observe acquisition creates emerging observability overlap with Datadog, but largely complementary
• MongoDB (MDB) — SNOW competes with: Cloud data platforms — operational DB (MongoDB) vs analytical (Snowflake)
• Alphabet (GOOGL) — SNOW competes with: Cloud data warehouse — Snowflake vs Google BigQuery (serverless analytics)
• Microsoft (MSFT) — SNOW competes with: Cloud data warehouse — Snowflake vs Azure Synapse / Microsoft Fabric
• Oracle (ORCL) — SNOW competes with: Cloud data platform — Snowflake vs Oracle Autonomous Data Warehouse

Others in Data & Infrastructure Software

• IBM (IBM)
• MongoDB (MDB)
• Cloudflare (NET)
• Datadog (DDOG)
• S&P Global (SPGI)
• Moody's (MCO)

All 174 companies · What moved this quarter · How the map is built