Trading cheat sheet (SNOW)

• When: Wednesday, May 27th, 2026, after the close.
• What matters most: consumption steadiness (less explanation, more “it just is”).
• AI check: Cortex usage that looks repeatable and budgeted, not one-off demos.
• Listen for: anything that sounds like customers are still optimizing hard.
• Also listen for: margin commentary that hints AI workloads are either helping… or getting expensive.

Snowflake doesn’t really sell “a data warehouse.” Not in the way people picture it. It sells a meter. The bill grows as workloads pile in, and the workloads pile in when the org finally decides the data platform is where work happens, not where data sits.

If that sounds a little fuzzy… it is. Consumption always feels fuzzy until it doesn’t. You don’t see the decision get made, then you see it all at once.

Here’s the thing: this quarter probably won’t be won by the flashiest Cortex soundbite. It’ll be won by whether the base business feels calm. Calm is underrated.

One line I keep coming back to: last quarter’s net revenue retention was 125%. That’s still real expansion. But I’m not treating it like a trophy. I’m treating it like a thermometer. Is it stable? Is it drifting? And do they talk like they know exactly why?

Now Cortex. What’s interesting is that Snowflake has made parts of Cortex monetization pretty “countable” through usage-based credits (tokens and service units). That’s good because it can scale quickly when customers move from experimenting to shipping. It’s also unforgiving: if costs feel spiky, customers can slam the brakes. Fast.

Slight tangent, but it matters: the buyers who greenlight AI inside big companies aren’t always the ones who own the budget. Finance shows up later. The platforms that win are the ones where finance doesn’t get surprised on month-end close.

So into May 27, I’m basically listening for two tones. Tone #1: “consumption is steady and broad-based.” Tone #2: “Cortex is becoming normal.” If they have to keep qualifying either one, that’s information too.

Worth a look. Not because it’s a sure thing. Because the language on this call tends to tell you where customer behavior is headed next quarter, not where it was last quarter.