Snowflake Inc. (SNOW)

News Summary

• Snowflake's two agentic AI products—CoCo (AI coding agent, 9,100 accounts, +2,000 in the quarter) and CoWork (5,800 accounts)—together accounted for approximately half of the acceleration to 37% YoY product revenue growth, confirming AI workloads are now a material, not marginal, revenue driver for enterprise data platforms.
• Snowflake disclosed that AI workloads carry lower contribution margins, prompting a downward revision to non-GAAP product gross margin guidance to 74%; simultaneously, management highlighted model-neutrality via its Cortex AI Gateway—dynamically routing to optimal models including open-weight inference—as a competitive differentiator, with customers expressing regret over single-provider frontier model commitments, a dynamic directly relevant to inference infrastructure positioning.
• A large network equipment manufacturer completed a full Teradata migration in under three quarters versus a prior 2–3 year typical cycle, illustrating how AI-accelerated data migration is compressing enterprise deployment timelines and pulling workloads onto cloud data platforms faster—a demand-pull signal for underlying compute and storage infrastructure.
• Q2 FY2027 revenue of $1.55B (+35% YoY) and product revenue of $1.49B (+37% YoY) beat consensus by ~$70M; non-GAAP EPS of $0.62 crushed the $0.45 estimate; full-year product revenue guidance raised to $6.07B (36% growth, up from 31%), Q3 product revenue guided to $1.59B vs. $1.50B consensus, and full-year adjusted operating margin target lifted to 14.5% from 13.5%; RPO reached $9B (+30% YoY).
• SNOW surged ~22–24% in after-hours and premarket trading to a record high above the prior 52-week high of $341.95, with Cantor Fitzgerald raising its target to $405, Citigroup to $395, KeyBanc to $375, and TD Cowen to $370; CNBC framed the move as the fourth-largest single-session jump since Snowflake's 2020 IPO, with analyst sentiment shifting from skepticism to validation of durable AI-driven acceleration.

Financial Highlights

Earnings Call Highlights

Snowflake's Q2 FY2027 call was defined by a confident, almost triumphant management posture as the company reported its third consecutive quarter of product revenue growth acceleration, reaching 37% year-over-year and raising full-year guidance to 36% growth. The dominant theme was the 'agentic enterprise' narrative, with CEO Sridhar Ramaswamy positioning Snowflake as the governed control plane for enterprise AI, a claim backed by concrete adoption metrics rather than just aspiration. Analysts were broadly constructive, probing the quality and durability of the acceleration rather than questioning its existence, signaling a sentiment shift from skepticism to validation. The call revealed that AI products—primarily Cortex Code and CoWork—contributed approximately half of the growth acceleration, with the other half coming from a reinvigorated core platform, a more balanced and durable mix than prior quarters. Management's willingness to discuss internal AI productivity gains, outcome-based selling via the Frontier Engineering program, and gross margin compression from AI workload mix all added credibility and nuance to the bull case.

• Snowflake's AI products (Cortex Code and CoWork) contributed approximately half of the growth acceleration to 37% year-over-year product revenue, with Sridhar Ramaswamy stating 'they contributed approximately half of the acceleration that we are seeing'—a more balanced split with the core than prior quarters and a signal that AI revenue is becoming material, not marginal.
• CoWork surpassed 9,100 accounts, adding over 2,000 net new accounts in a single quarter, while Cortex Code expanded to 5,800 accounts; critically, management disclosed that accounts using CoWork consume more of the core platform, confirming a structural consumption multiplier rather than a standalone product dynamic.
• Non-GAAP product gross margin guidance was revised down to 74% from a higher prior expectation, explicitly because fast-growing AI workloads carry lower contribution margins today—a direct and honest disclosure that AI infrastructure costs are compressing margins even as operating leverage improves, a dynamic highly relevant to AI chip and compute economics.
• Model neutrality is emerging as a competitive differentiator: Christian Kleinerman confirmed customers who made large commitments to a single frontier model provider are expressing regret, and Snowflake's Cortex AI Gateway dynamically routes tasks to the optimal model—including open-weight models where Snowflake runs inference itself, offering meaningfully different and improving unit economics.
• Management disclosed that a major Teradata migration by a large network equipment manufacturer is completing in under three quarters, a process that 'would have taken probably 2 to 3 years in any previous time,' illustrating how AI-accelerated migration is pulling workloads onto the platform faster and compressing the sales and deployment cycle.

Sources

• Snowflake spikes 22% on healthy results and AI coding momentum
• Snowflake - Financials - Quarterly Results
• Snowflake Shares Rise After Q2 Revenue and Earnings Exceed ...