Snowflake (SNOW) Q2 FY2027 Earnings Call: Q&A Summary

Q: Sanjit Singh with Morgan Stanley asked about the quality and durability of Snowflake's growth acceleration, questioning why CoCo and the platform are the right tools for supply chain and finance use cases, and whether there is any irrational behavior in consumption.

A: Sridhar Ramaswamy responded that growth is coming from a broad swath of customers rather than being concentrated in AI-native companies, and tools like CoCo make optimization and cost management much easier. Brian Robins added that gross retention has been flat, guidance is based on observed behavior, and Snowflake is selling into more personas like CFOs and CROs.

Q: Kirk Materne with Evercore ISI asked what portion of the acceleration is coming from newer products versus the flywheel effect of those newer products on the core platform.

A: Sridhar Ramaswamy stated that AI products contribute approximately half of the acceleration, while other products like Notebooks and Streamlit applications, alongside faster migrations, drive the rest. He noted an example of a network equipment manufacturer doing a Teradata migration in less than 3 quarters.

Q: Karl Keirstead with UBS asked about behavioral shifts between open class and Sonnet class models, whether model neutrality is a competitive advantage, and the margin impact.

A: Sridhar Ramaswamy and Christian Kleinerman noted that customers are interested in switching models to optimize cost, and model neutrality is a strong competitive advantage because customers want flexibility rather than being locked into one model company. Brian Robins added that product gross margin was 74% and the priority is delivering great outcomes for customers while driving overall operating margin leverage.

Q: Raimo Lenschow with Barclays asked about performance and opportunities in regions outside North America, such as Europe and Asia.

A: Brian Robins answered that all regions are performing well and their outlooks are factored into guidance.

Q: Ryan MacWilliams with Wells Fargo asked what changed to cause this AI moment for Snowflake and drive accelerating revenues.

A: Sridhar Ramaswamy explained that AI has lowered the distance between data and value, and products like CoWork and CoCo have made the sales team and customers AI native. He noted that the time to reach 80% of purchased consumption for new customer cohorts has visibly improved.

Q: Matt Hedberg with RBC Capital Markets asked how deep CoCo is used within Snowflake's knowledge worker base and if CoCo acts as a sandbox before moving to CoWork.

A: Brian Robins and Christian Kleinerman stated that CoCo is used across deal desk, tax, accounting, internal audit, FP&A, and treasury with over 150 Snowflake on Snowflake applications, achieving nearly 100% adoption within the finance organization.

Q: Koji Ikeda with Bank of America asked about the consumption uplift for AI adopters relative to nonadopters and what gives confidence that this reflects higher lifetime consumption rather than pulled-forward projects.

A: Sridhar Ramaswamy responded that cohort behavior shows noticeable uplift as CoCo adoption deepens, and confidence comes from the breadth and depth of complex use cases being deployed that previously would have required specialized applications.

Q: Brent Thill with Jefferies asked if there are differences in adoption trends for CoCo, such as bigger lands or more users right out of the gate.

A: Sridhar Ramaswamy stated that Snowflake uses a sophisticated pipeline and methodology to measure CoCo penetration, supported by in-product guidance and 3-hour hands-on labs, creating positive momentum because CoCo makes every customer task go faster and better.

Q: Daniel Knauff with Deutsche Bank asked about Snowflake's ambitions regarding proprietary models like Arctic within the overall model strategy for CoCo and CoWork.

A: Christian Kleinerman clarified that Snowflake is not building models to compete in the frontier class, but continues developing Arctic family models for constrained, specific tasks like document processing, AI functions, and embeddings to provide higher accuracy and efficiency.

Q: Aleksandr Zukin with Wolfe Research asked about the size of the budget opportunity Snowflake can access and requested a preview of the outcome-based Frontier Engineering program.

A: Sridhar Ramaswamy explained that AI allows Snowflake to access new strategic conversations with executives regarding asset distribution, supply chain, and risk. He highlighted the Frontier Engineering program, where Snowflake works with enterprises like Sanofi to deliver business outcomes, getting paid only when those outcomes are delivered.

Q: Tyler Radke with Citi asked how Snowflake views traditional SaaS companies partnering with LLMs and becoming databases, and how competitive overlap looks over time.

A: Sridhar Ramaswamy and Christian Kleinerman stated that customers do not want to copy data across multiple application platforms, preferring to consolidate on a single central platform like Snowflake. They noted that future investments in Hybrid Tables and Postgres will form the foundation for new agent applications.

Q: Samik Chatterjee with JPMorgan asked how the full year guide increase is split between core growth and AI, and whether core consumption trends or visibility have fundamentally changed.

A: Brian Robins explained that guidance is based on observed behavior, noting that AI products are driving additional business and customers adopting AI are also increasing their consumption within the core platform.