Finvest
SNOW Cloud software · AI · Data cloud · Consumption model · Thesis updated June 11, 2026

AI growth is real, but priced for perfection

01 Running thesis

AI has restarted the growth story

Snowflake is back in a stronger growth phase. In Q1 FY27, product revenue grew 34% year over year, and Net Revenue Retention was 126%. Net Revenue Retention means how much the same customer group spends this year compared with last year. A number above 100% means customers are expanding.

The bull case is simple: companies need clean, governed data before they can use AI well. Snowflake already holds that data for many large customers. As those customers add AI work, they use more Snowflake capacity. New products such as Cortex Code and Snowflake Intelligence give Snowflake a way to charge for that demand directly.

The bear case is also real. Growth at this scale gets harder each year. AI products can carry lower gross margin than the core data platform, and the market already expects a lot. That fits Finn's weak valuation score.

The newest risk is not about growth at all. The Q1 FY27 10-Q says Snowflake has been made aware of additional cyberattacks on customer accounts since May 2024. The company says these attacks used similar methods tied to weak customer safeguards, such as missing MFA. Even so, the brand risk lands on Snowflake too.

May 2026The Q1 FY27 10-Q confirmed 34% product revenue growth and 126% Net Revenue Retention. It also made clear that cyberattacks on customer accounts are ongoing, which keeps legal and reputation risk high.
May 2026Q1 FY27 showed a clear growth re-acceleration. Management raised FY27 product revenue growth guidance to about 31% after seeing stronger core platform demand and a meaningful lift from AI products.
Mar 2026The FY26 10-K confirmed 29% revenue growth and a wider product set, including Snowflake Postgres and Observe. It also repeated that customer account attacks were not limited to one past event.
Feb 2026Q4 FY26 strengthened future visibility, with RPO up 42% year over year and a largest-ever deal of more than $400 million. Net Revenue Retention stayed healthy at 125%.
Dec 2025The Q3 FY26 10-Q showed RPO of about $7.9 billion and Net Revenue Retention holding at 125%. The filing also kept the customer account security issue in focus.
Dec 2025Q3 FY26 gave the first clear proof of AI monetization, with a $100 million AI revenue run rate reached earlier than expected. Product revenue grew 29% year over year.
Sep 2025The Q2 FY26 10-Q confirmed that Net Revenue Retention stayed at 125% for a second straight quarter. That challenged the old bear case that customer expansion was structurally fading.
Aug 2025Q2 FY26 improved the thesis as product revenue growth accelerated to 32% and Net Revenue Retention rose to 125%. AI and Iceberg adoption added support for new growth drivers.
02 Business model

Pay for what customers use

Snowflake mainly makes money through a consumption model. Customers buy capacity, then spend it as they run queries, move data, build apps, or use AI tools. This model can grow fast when customers move more work onto the platform.

The main revenue line is product revenue. In Q1 FY27, product revenue was $1.334 billion out of total revenue of $1.391 billion. Professional services and other revenue was much smaller at $56.6 million.

This model cuts both ways. If customers slow projects, optimize workloads, or delay AI rollouts, usage can weaken quickly. Snowflake also pays cloud providers to run the platform. In Q1 FY27, third-party cloud infrastructure costs were about 73% of cost of product revenue.

Management is trying to hold a 75% product gross margin target while AI products grow. That is an important test. If AI drives revenue but pulls margins down, the business can look less attractive even with strong top-line growth.

03 Product portfolio

Data core, AI edge

Cash cow

Core Data Platform

This is the base platform for data warehousing, data lakes, data engineering, analytics, and data sharing. It is still the main engine behind product revenue.

Growth engine

Cortex AI and Snowflake Intelligence

These AI and machine learning tools help customers build AI features on top of their own data. Management said adoption has been the fastest for any new products in company history.

Growth engine

Cortex Code

Cortex Code, also called CoCo, launched in Q1 FY27 and became a meaningful revenue contributor. More than 7,100 accounts were using it as of Q1 FY27.

Steady

Snowpark and Dynamic Tables

These tools help developers and data teams build pipelines and apps inside Snowflake. They make the platform more useful for complex workloads.

Option

Iceberg Tables

Iceberg support helps Snowflake work with the open Apache Iceberg table format. This matters for customers that want open data standards rather than a closed system.

Option

Transactions and Snowflake Postgres

Snowflake Postgres is a fully managed Postgres offering in the transactions product category. The goal is to bring app data and analytics data closer together.

Option

Observe

Snowflake acquired Observe in February 2026 to add AI-powered observability. The acquired business is still small, contributing less than 1% of product revenue growth.

Option

Natoma

Snowflake agreed to acquire Natoma in May 2026 for about $110 million. Natoma is meant to extend Snowflake's control layer for AI agents into enterprise apps such as email and calendars.

04 Business segments

One segment, two revenue lines

Product revenue96%growing fast
Professional services and other revenue4%modest

Snowflake reports one operating segment. For the three months ended April 30, 2026, the revenue mix below uses the filing's product revenue and professional services and other revenue lines.

05 Risk factors

What could break the story

Customer account attacks keep spreading

High impact · Medium odds

Snowflake says it has been made aware of additional cyberattacks on customer accounts since May 2024. The company says the attacks use similar methods and target customers that fail to use controls such as MFA and network access policies. Even if Snowflake's platform is not the root cause, buyers may still slow deals if the brand becomes linked with data theft.

We watchWatch the next 10-Q risk factors, new customer security lawsuits, and any comment on deal delays tied to security reviews.

AI usage fades after early excitement

High impact · Medium odds

The bull case needs AI products to create durable consumption, not one-time tests. Cortex Code adoption is strong, but the key question is whether customers keep spending after early rollouts. If AI usage proves less sticky than core analytics usage, growth can slow again.

We watchWatch product revenue growth versus the FY27 outlook of about 31%, plus any disclosure on Cortex Code consumption uplift.

Lower-margin AI mix pressures profits

Medium impact · Medium odds

Snowflake says new AI products carry lower gross margin than the core platform. The company is offsetting that with other savings, including cloud cost efficiencies. This balance may get harder once AI becomes a larger revenue share.

We watchWatch product gross margin versus the 75% target and management comments on cloud infrastructure costs.

Premium valuation leaves little room for mistakes

High impact · High odds

Finn's valuation score is very low, so the stock already prices in strong execution. A small miss in growth, margin, or security news could matter more than it would for a cheaper stock. This is the main reason the overall score is more cautious than the growth story alone might suggest.

We watchWatch whether product revenue growth, NRR, and guidance all stay strong in the same quarter.

Heavy stock compensation and GAAP losses

Medium impact · High odds

Snowflake still reports GAAP losses. In Q1 FY27, net loss was $295.6 million, and stock-based compensation was $402.5 million. That does not stop the company from investing, but it matters for shareholder dilution and Finn's weak financial health score.

We watchWatch net loss, stock-based compensation, share count, and buyback activity each quarter.
06 Quick answers

In one breath

How does Snowflake make money?

Snowflake mostly charges customers based on how much they use its cloud data platform. More queries, data movement, apps, and AI workloads usually mean more product revenue.

Why is AI important to Snowflake?

AI needs organized and trusted data. Snowflake already stores and manages that data for many companies, so tools like Cortex Code and Snowflake Intelligence can turn existing data into more usage.

What is the biggest risk for Snowflake stock?

The biggest stock risk is that expectations are very high. The business must keep growing fast while handling security headlines, legal costs, AI margin pressure, and GAAP losses.

Did Snowflake itself get hacked?

Snowflake says the customer account attacks used similar methods tied to customers failing to use proper safeguards such as MFA and network access policies. The issue still matters because customers and headlines may link the incidents to Snowflake's brand.