Snowflake Stock Nearly Doubled On A Product Its Own Forecast Could Not See

The company’s guidance method excluded a product it had only just launched, and management said so two months before the run began.

Snowflake (SNOW) stock has gained about 98% over the past three months, from $142.56 to $282.90, while the S&P 500 returned 2.5%. Its largest cloud peers went the other way over the same window: MSFT and AMZN fell 8.8% and 12.7%, and GOOGL slipped 3.7%. Start with what the run was actually about.

The Outlook Went From 27% To 31% In One Report

At the fiscal Q1 2027 report in late May, inside the run, product revenue came in at $1.334 billion, with growth accelerating to 34% year over year, from 30% in fiscal Q4 2026 and 26% a year earlier. Net revenue retention rose to 126%, and non-GAAP operating margin expanded more than 300 basis points year over year to 12%. The bigger item was the outlook: management lifted the fiscal 2027 product revenue guide to $5.84 billion, from 27% to 31% growth, and the full-year non-GAAP operating margin guide from 12.5% to 13.5%. By the company’s own explanation, the largest single driver of that raise was Cortex Code, its coding agent.

The February Guide Management Said Could Not Include Cortex Code

Now go back three months. At the fiscal Q4 2026 report in late February, the company guided fiscal 2027 product revenue to roughly $5.66 billion, or 27% growth, after posting $1.23 billion of product revenue and 30% growth in the quarter itself. By management’s own account, that guide could not contain Cortex Code: the forecast is built from observed consumption behavior, the coding agent had reached general availability only weeks earlier, and no model could yet describe a product already in use at more than 4,400 customers. Management said there was a lot of upside sitting outside the number it had just published. Three months later it published a bigger one.

The AI Adoption Curve Was Public A Quarter At A Time

The broader AI adoption around Cortex Code was not hidden either. Snowflake Intelligence, the business-user side of the platform, was still in public preview as of the fiscal Q2 2026 report; by fiscal Q4 2026 the company counted more than 2,500 accounts, almost doubling quarter over quarter. At the fiscal Q3 2026 report in between, the AI line crossed a $100 million revenue run rate a quarter earlier than the company had expected, and AI influenced 50% of the bookings signed in that quarter. By fiscal Q4 2026 remaining performance obligations were growing 42% year over year, accelerating for a second straight quarter and ahead of the 30% product revenue growth in the same quarter. Snowflake Intelligence and Cortex Code appeared together in a company announcement a week before the run began. One further pre-surge indicator is the one most often misread.

What The Pre-Surge Options Reading Did Not Say

Implied volatility sat in the 88th percentile of its own trailing one-year range in mid-March and the 95th percentile by mid-April, at a reading of 61.2. Options were priced for a large move in either direction, up or down. Nothing in that told you which way.

Legible As An Understated Guide, Not As A Doubling

So were the signs real? Yes, and unusually explicit, though not in the trailing total revenue line. Heading into the run, trailing twelve-month revenue was $4.68 billion, up 29%, roughly steady with the prior few quarters, and on a reported basis, not the non-GAAP one, trailing twelve-month operating margin of -31% was already a materially narrower loss than its three-year average of -38%. The acceleration was showing up in bookings and product adoption first. The tell was management saying out loud that its own guide excluded the product it had just launched. What it did not give you was size or timing, and the pre-surge headlines also included law-firm class-action reminders over a class period that ended in February 2024. Turned forward, the same idea is a screen for names where a raised outlook shows up before the price does. The stock now sits at $282.9, the top of its trailing 52-week range, a very different entry than $142.56 was.

Catching The Signal Was The Easy Half

Sizing was the hard half. Someone who read the February signal correctly still had to decide how much of their money belonged behind a company whose reported operating loss was narrowing but nowhere near closed. That reading is weaker today, with the guide raised and the stock at its high. Reading one setup well is not a system. Owning a rules-based portfolio that survives being wrong about any single name is, and that is the idea behind the Trefis High Quality Portfolio. The Trefis High Quality (HQ) Portfolio has a track record of outpacing a benchmark that combines the three major indices – the S&P 500, S&P Mid-cap, and Russell 2000.