Snowflake Re-Accelerated on Its AI Assistant — the Backlog and Cash Have to Catch Up

Snowflake lifted its full-year product revenue forecast in Wednesday's fiscal second-quarter report and credited its AI assistant with a re-accelerating top line, sending the stock higher in extended trading. This was a 90-point quarter against a hurdle raised by three months of AI hype: product revenue of $1.49 billion grew 37% year over year, the third consecutive quarter of accelerating growth, and it cleared the top of the company's own guidance.

The market already owned most of this story. SNOW entered the report up roughly 39% this year and about 73% over the prior four months, at a $106 billion market cap and around 21x trailing sales. The report has to answer a different question: whether an acceleration this visible is durable enough to justify a multiple that leaves no room for the growth to fade.

The assistant is pulling workloads in

The driver is the AI/agentic layer, not the aging data warehouse. CoCo, an AI coding agent that lives inside the platform, passed 9,100 customer accounts and added more than 2,000 in the quarter; CoWork, the assistant aimed at knowledge workers, expanded to 5,800 accounts. The CFO put the mechanism plainly, pointing to "strength in both our core data platform and a meaningful step-up in AI revenue."

That is the move the market has trained investors to reward: AI adoption showing up in the revenue line rather than in slideware. Third-quarter product revenue is guided to about $1.59 billion, up 37% to 38%, and management raised full-year product revenue to $6.07 billion, or 36% growth, from a prior $5.84 billion (31%). For a company growing this fast, that is genuine forward visibility.

The two measures that lead revenue did not accelerate with it

Net revenue retention held flat at 126%, and remaining performance obligations — the backlog of contracted future work — grew 30% year over year to $9.00 billion, roughly flat to slightly below Q1's $9.21 billion. RPO is lumpy, and the prior quarter included a large AWS commitment, so one soft quarter is not a trend. But it is the number to watch: when near-term consumption accelerates while the booked order book grows more slowly, the burst can outrun the pipeline.

Then there is cash. Free cash flow in the quarter was $83.8 million, about a 5.4% margin and 6.0% on an adjusted basis, against a 23% full-year target. On a trailing-twelve-month basis it looks healthier — around $1.17 billion, up about 59% year over year — and much of the quarterly dip is seasonality, since collections concentrate early in the fiscal year. Still, a re-rating built on a 36% guide and a 21x sales multiple requires that cash conversion to actually land.

The old doubt was answered. The new question is leverage.

The old worry about Snowflake — that AI tools would commoditize the data warehouse and slow its growth — was answered in the opposite direction: the assistant is drawing workloads in and re-accelerating revenue. The new question is whether that acceleration feeds the backlog and cash, or runs ahead of both.

Next quarter's $1.59 billion product-revenue guide is the hurdle. If net revenue retention holds above 126% while backlog growth re-accelerates toward the 36% revenue pace, the AI flywheel is real. The signal that breaks the narrative is the reverse: product revenue decelerating off this guide while the agentic buildout keeps free cash flow far below the 23% target. Extended trading is sentiment; those conversion numbers are the proof.