Snowflake: AI Growth Is Real, but the Multiple Has Already Priced It

Snowflake cleared the bar that actually matters for its bulls on Wednesday, and the reaction says the market is listening. Management reported product revenue of $1.49 billion, up 37% from a year ago, on top of total revenue of $1.55 billion that grew 35% and beat the Street's $1.48 billion estimate. The stock jumped about 20% in extended trading.

The reason this quarter matters is that it is a second consecutive confirmation, not a one-off. Product growth had spent months decelerating through 2025 — the story skeptics doubted was whether customers would actually write bigger checks for AI-driven data workloads. Instead, growth accelerated from 34% a quarter ago to 37% now, net revenue retention held at 126%, and management raised full-year fiscal 2027 product revenue guidance to $6.07 billion from $5.84 billion. The number of customers generating more than $1 million of trailing product revenue rose 27% year over year to 828.

What is driving the acceleration

The mechanism is now visible rather than a pitch. Snowflake's AI coding agent, CoCo, added more than 2,000 accounts in the quarter, bringing its base to roughly 9,100. Under the company's model, AI tools sit on top of the same consumption-based data platform, so every workload an agent or app generates is metered consumption. That is the flywheel the bull case promised: AI is not displacing Snowflake's core spending but adding to it. The company also closed a five-year, $6 billion cloud commitment with Amazon Web Services, capitalizing on the momentum.

Management reinforced the acceleration with guidance. It raised full-year product revenue again and projected third-quarter product revenue near $1.59 billion, and it lifted the adjusted operating margin outlook to 14.5%.

The cost of buying that growth

Here is where the good-company/good-stock distinction does its work. The fastest-growing parts of the business are AI products that carry a lower gross margin than the core platform. Snowflake is defending its roughly 75% non-GAAP product gross margin guide for the year, but the way it holds that line is telling: it is offsetting the AI margin drag with cheaper bandwidth under the new AWS contract rather than seeing the margin expand naturally. In the prior "best-growth-quarter-ever" print, gross margin and adjusted free cash flow conversion actually slipped as AI consumption accelerated.

That matters because of the price. The stock had already roughly doubled over the trailing three months before this report, trading around 16 times forward sales — and before Wednesday's pop, against roughly 90 times trailing operating cash flow versus about 16 for the S&P 500. The AI turnaround is no longer a contrarian bet on a fallen name; the market has re-rated it as a growth compounder, and the multiple now prices in continued 30%+ product growth.

The honest read

The evidence supports the growth story — this is real paid demand, corroborated in revenue, retention, and guided commitments, not management language. That is a meaningful upgrade in what the stock's thesis rests on. But a good company is not automatically a good stock at this multiple once the acceleration is already out and largely discounted.

For a holder, the quarter validates the position and the raised guidance extends the runway. For someone who has not bought the pop, the risk is not that the story is fake — it is that you pay ~16 times forward sales for growth whose final quality has not yet been proven at scale. The falsifiable tests over the next two quarters are specific: does product gross margin hold near 75% as AI grows as a mix of revenue, does net revenue retention stay near 126%, and — most importantly — does free cash flow conversion stop deteriorating as the fastest growth rolls in. If those hold, the premium is defensible; if the AI mix keeps pulling margins and cash flow down, the multiple has run ahead of the proof. That is a reason to watch and let the next two quarters decide, rather than a reason to stop caring.