Review|SNOW FY27Q2: Beyond a Single Product, a Data Flywheel That Merits a Longer Horizon

This was another exciting quarter for Snowflake. Our channel checks on CoCo and CoWork were very positive, but we were somewhat cautious on FY27Q2 due to tough comps and more constructive on FY27Q3. The results suggest that the data flywheel driven by CoCo and CoWork is stronger than we expected and has not been materially constrained by tough comparisons.

A Review of Our Snowflake View Over the Past Several Quarters

We first turned positive on Snowflake's inflection in FY25Q3. We viewed Iceberg as incremental rather than dilutive and called for Snowflake to slow hiring and refocus on operating margin. SNOW has, in fact, slowed hiring meaningfully over the following quarters.

In our FY26Q3 Preview, we expected Gen2 optimization and Snowpark Connect to weigh on Snowflake. Frontier labs were also optimizing usage on Snowflake, creating revenue risk.

In our FY26Q4 Preview, we argued that Gen2 optimization was complete, Snowpark Connect had shifted from a headwind to a tailwind, and frontier labs had finished optimizing and were beginning to reaccelerate consumption.

In our FY27Q1 Preview, we spoke with six experts about Cortex Code. We made the bold call that Cortex Code was the fastest-growing product in Snowflake's history, growing far faster than Snowpark. We also argued that Cortex Code is not just a coding agent for data teams. It also addresses non-data use cases, allowing it to grow much faster than products such as Text to SQL.