Snowflake’s AI Efforts Accelerate, but Can Margins Withstand the Pressure?
Snowflake just posted the kind of AI growth numbers that send stocks soaring, but a quiet revision buried in management's guidance reveals a cost problem that could haunt the bull case for quarters to come.
This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.
Snowflake (NYSE:SNOW | SNOW Price Prediction) just delivered a quarter that reframes the debate around enterprise AI economics. Q2 FY27 revenue hit $1.55 billion, up 35.09% year over year, with product revenue of $1.49 billion, up 37% and marking a third consecutive quarter of product revenue growth acceleration. Non-GAAP EPS came in at $0.62 versus the $0.447 estimate, a 38.7% surprise. Shares fell more than 4% ahead of yesterday’s earnings, but were soaring 22.8% ahead of the opening bell this morning, or over $68 per share, to around $374.
Why This Quarter Matters Now
CEO Sridhar Ramaswamy framed the quarter around a widening AI flywheel: “AI continues to compound our advantages, creating a flywheel effect across the business. CoWork and CoCo are driving transformational outcomes for our customers, while fueling rapid adoption, user growth, new workloads, and overall platform consumption.”
Adoption metrics back him up. Cortex AI surpassed 9,100 accounts, CoCo added over 2,000 accounts to reach 9,100, and CoWork expanded to 5,800 accounts. Sayari is using CoCo to migrate 12 billion records at half the cost, and 1Password and Indeed selected Snowflake for AI transformation. Net revenue retention held at 126%, and RPO climbed to $9.00 billion, up 30% YoY.
Margin Question Gets Sharper
Management raised the year: FY27 product revenue guidance moved to $6,070 million (36% growth) from $5,840 million (31%), and non-GAAP operating margin guidance was raised to 14.5% from 13.5%. Non-GAAP operating margin in Q2 expanded to 15.3% from 11% a year ago.
The tension is buried in a single line: FY27 non-GAAP product gross margin guidance was trimmed to 74.0%, down from the 75.0% target management defended just one quarter earlier. On the Q1 call, CFO Brian Robbins was blunt about the mechanics:
“You’re absolutely right. Our AI products have a lower gross margin than our core platform.”
Ramaswamy has pointed to usage controls as the offset: “We are creating the controls that one needs in order to keep cost manageable as things continue expanding.” Bandwidth savings from a new $6 billion five-year AWS contract and an expanding $200 million OpenAI partnership are also expected to cushion the mix shift.
What Investors Should Watch Next
The bull case rests on core-platform consumption accelerating faster than AI mix dilutes gross margin. Q3 guidance calls for $1,588 to $1,593 million in product revenue (37-38% growth) and 15.5% non-GAAP operating margin. The bear case is visible on the GAAP line: a $262.97 million operating loss, $456 million in Q2 stock-based compensation, and a product gross margin guide moving in the wrong direction. Morningstar raised its price target to $284 from $255 but now views shares as overvalued amid intensifying AI competition. With SNOW up 39.42% year to date, the AI narrative is now doing the heavy lifting.
Contact [email protected] for any questions or corrections.








