The question hanging over Snowflake’s September 2 print isn’t whether Cortex Code is working — Q1 fiscal 2027 product revenue of $1.33 billion, up 34% year-over-year, already answered that. It’s whether the market has priced in the answer twice. Shares have run roughly 178% off an April 10 low of $118.30, per TIKR, and now trade above TIKR’s own $307.37 average analyst target even as Wells Fargo’s July 29 call for $500 anchors the bull case. That’s not a consensus; it’s a bet still being placed in real time.
The mechanism worth watching isn’t the coding agent itself but what it does to Snowflake’s unit economics. CFO Brian Robins told investors that the company’s top 25 customers now average $34 million a year in spend, up from $22 million two years ago, while the median Fortune 2000 account still spends just $2.4 million — a 14x expansion runway that management is funding without adding people (net headcount up 17 in a recent quarter, per TIKR’s reporting). That’s the operating-leverage story a 155x-forward-earnings, still-GAAP-unprofitable stock needs to hold, not just a good AI headline.
A consumption business trading at 98x EV/EBITDA doesn’t need a good quarter on September 2 — it needs a quarter that resets the range between Wells Fargo’s $500 and the Street’s lagging $307 average, or the easy re-rating is over.
Two dated data points suggest which side has momentum: TD Cowen lifted its target to $370 from $300 on August 20, following similar raises from Deutsche Bank, Truist, BofA and UBS, and Point72 reportedly grew its stake roughly 2,449% to 2.81 million shares worth $716.1 million entering Q2. Neither confirms the bull case — hedge fund filings and sell-side targets move on momentum as much as fundamentals — but they show institutional money positioning for confirmation, not surprise. The real test on September 2 is whether product revenue growth accelerates again past 34%, or merely holds it; a merely-good number, against this setup, could read as disappointment.
For four months in early 2026, the fear was that AI agents would gut consumption software, and the stock fell 56.30% from its high to an April 10 low of $118.30. Wells Fargo analyst Ryan MacWilliams argued the opposite after surveying Snowflake customers: AI is driving them to upload more data and run more workloads, not fewer.
