These are the times that try investors’ souls (at least the few of us out there still with one). The latest bull charged right off a cliff a few weeks ago, driven by a multitude of factors – a softening U.S. economy, uncertainty over interest rates, the start of World War III in the Middle East, and disappointing earnings from Big Tech after spending big bucks on AI. Savvy investors don’t pay attention to this noise since they’re dollar cost averaging into quality names at reasonable simple valuation ratios. One such quality name is big data platform vendor Snowflake (SNOW).
A lot has happened since we last checked in with our favorite cloud data storage company. For starters, Snowflake got itself a new CEO in February 2024 when Sridhar Ramaswamy took the reins from Frank Slootman. A former top Google executive who grew the internet giant’s advertising business from $1.5 billion to more than $100 billion, Ramaswamy joined Snowflake just last year after it acquired Neeva, which he had co-founded. Originally created to develop an alternative search engine that prioritized users’ privacy, Neeva pivoted into generative AI for enterprise applications when it turned out nobody cared Google tracks their porn-watching behaviors. Snowflake took an interest in Neeva to enhance the AI and search capabilities within its data cloud platform. Now, with Ramaswamy at the helm, the company is going all in on AI. More on that later.
Data breach no big deal
The new CEO immediately finds himself navigating some pretty choppy waters, including slowing revenue growth and a sliding net retention rate. (More on that later as well.) It certainly did not help that a group of cybercriminals stole data from more than 165 Snowflake customers starting in mid-April. However, news of the data breach did not emerge until the end of May after the company had reported its Q1-2025 earnings. We wanted to wait until the Q2-2025 numbers had been released this week to see what effect the incident had on the company’s bottom line.
The short answer: Not much. Product revenue reached $829 million, up 30% year-over-year, beating previous guidance. Snowflake CFO Mike Scarpelli said the cybersecurity incident has zero impact on the company’s consumption revenues. Meanwhile, Ramaswamy emphasized that after “multiple investigations by internal and external cybersecurity experts, we found no evidence that our platform was breached or compromised.” Of course, that did not stop at least one class-action lawsuit being filed against Snowflake but it sounds like the onus is on customers like Ticketmaster and AT&T for not enabling better security protocols like multi-factor authentication.
Snowflake Expands into AI
You might recall in our last article on Snowflake that we caught Ramaswamy’s predecessor, Slootman, having somewhat of a Blockbuster moment when he said this about generative AI during the company’s Q4-2023 earnings call:
“One of the challenges with these new technologies [is] that people come up with a lot of interesting questions, but without a solid business model, that’s not going to take off. So we take a very pragmatic view. We do anticipate that Snowflake data will be a very, very big driver of a large language model in conjunction with many, many other data sources. So we think that the gravity around data will drive a lot of this action activity to our platform.” CEO Frank Slootman
Under Ramaswamy, Snowflake has now appended the letters “AI” to its previous label of “data cloud company.” He has been instrumental in launching Snowflake Cortex, a fully managed AI service built into the Snowflake platform. It allows businesses to leverage machine learning and other AI goodies directly on their data with minimal specialized knowledge. It also provides access to large language models (LLMs) for working with unstructured text data. More recently, the company released its own LLM, Arctic, which it claims outperformed leading open models.
And it’s not stopping there. Snowflake plans to spend about $50 million on AI chips from Nvidia this year, which may have knock-on effects to the company’s gross margins due to increased hardware costs. That seems like pocket change compared to the billions and billions that companies like Meta and Alphabet are investing. Yet the same skittishness that investors have shown over perceived excessive spending on AI without immediate payoffs may temporarily hit Snowflake stock as well.
However, Snowflake says more than 2,500 accounts were now using AI on its platform on a weekly basis. As of Q1-2025, about 40% of customers processed unstructured data on Snowflake, with 1,000 customers signing up for this functionality over just a six-month period. Another 1,600 customers are using a new feature called Notebooks that allows data scientists to work collaboratively within the Snowflake environment, enhancing their ability to analyze data and build models. Just remember, the only measure of AI product adoption is revenue growth, which brings up a good question.
What’s the Story with Snowflake’s NRR?
So, why does the net retention rate (NRR) still look like this if customers are adding new bells and whistles under a consumption-based business model?
There’s likely a confluence of factors, starting with the fact that revenue growth has decelerated significantly over the past few years, from 120% in fiscal 2021 to 70% in fiscal 2023 to 34% in fiscal 2024. While Snowflake has now increased its full-year revenue guidance twice on the back of a strong first half, year-over-year growth is expected to be “just” 26%. Scarpelli has noted that over time, the net revenue retention will converge with the company’s revenue growth rate, so presumably we’re close to equilibrium. And, in fact, NRR inched down to just 127% in Q2-2025.
The addition of new, smaller customers may be diluting the overall NRR as well. For instance, Snowflake added 1,735 new customers between Q1-2024 and Q1-2025. Only about 6% of those represent customers spending $1 million or more on the platform, and larger enterprise customers with deeper pockets typically have higher expansion rates.
Management has declined to provide future guidance on net retention rates. In fact, the C suite seems to be focusing more on overall revenue growth and remaining performance obligations than emphasizing net retention rate as a key metric.
Should We Be Worried About Snowflake Stock?
So, let’s talk about that revenue growth, which is decelerating but certainly not collapsing. Again, we probably have a confluence of factors at play here. For one thing, Snowflake has already captured a significant piece of the data-cloud-computing pie, with online estimates ranging from about 20% to 35%. Another factor is competition in this space is fierce, with Databricks being one of the main rivals. In fact, a lot of the AI strategy rolling out of the Snowflake R&D department appears to be an attempt to catch up with Databricks, which has developed a cloud data platform purpose-built for machine learning and data science. You can find a pretty good technical comparison between the two platforms here but the consensus seems to be that while strong in data warehousing and analytics, Snowflake is still catching up to Databricks in terms of native AI capabilities.
On one hand, we’re encouraged that Snowflake is aggressively pursuing its AI strategy under Ramaswamy. On the other hand, this pace of innovation may strain resources and lead to quality issues or integration challenges. In Q2-2025 alone, Snowflake announced nine new products and made more than 15 capabilities available to customers. Not all of these are necessarily AI. Overall, however, there seems to be a throw-everything-against-the-wall-and-see-what-sticks strategy at work. We’re also disappointed that Snowflake gives us pretty much the same slide on geographic mix each quarter:
In other words, there is very little movement to expand beyond North America, which means about 80% of the business is concentrated in the United States. That suggests that there is still market to capture, particularly if this graphic on the total addressable market can be half believed:
Still, all of this seems like a lot of quibbling over a company that is still experiencing above-average revenue growth for a software-as-a-service (SaaS) company. In fact, of 100 public SaaS companies in the United States with revenues above $100 million that McKinsey analyzed in 2021, the median revenue growth rate was just 22%. It simply gets harder to maintain stratospheric growth as a company matures.
Conclusion
Shares did slip more than 10% after the Q2-2025 earnings report in after-hours trading, dropping the simple valuation ratio ($39.95 billion market cap/$3.3 billion annualized revenue) to just 12. Investors interested in adding shares of a quality company certainly could not be faulted for jumping at Snowflake stock at that price point. However, jump with your eyes wide open: Snowflake is now an AI stock, so expect even more volatility than ever before.
