Years ago, I used to wish that Amazon would spin off AWS, or Microsoft would spin off Azure, allowing me to invest in their cloud services without also having to invest in all the other low margin things Microsoft and Amazon do. But last year, something better came along. That something is Snowflake, a leader and pioneer of the “data cloud”. Snowflake hit a big milestone in this quarter of processing more than 1 billion queries in a single day!
I’m going to oversimplify Snowflake’s business for non-technical people into three major areas: storage of data, compute (ability to do stuff to/with the data), and data sharing. Data sharing is capable both within an organization and between different organizations.
We have very little to complain about in Snowflake’s recently posted quarter as Snowflake grew product revenues 110% year over year to $214 million. Margins jumped to 72%, up from 66% in the comparable quarter last year. Management indicated in the long term, this number could trend upward into the mid 70’s. The net revenue retention rate was 168%, meaning that, if a customer spent $100 last year on the Snowflake platform, this year they are spending $168.
Additionally, the company now has 117% more million dollar customers than the comparable quarter last year, clocking in at 104. And there are many companies on the cusp of the $1mm dollar mark who could very well join the cohort in coming quarters. Net new customer adds (of any dollar amount) was 393 in Q1. These customers only accounted for 1% of revenues, which is a reminder that Snowflake’s time to fully ramp on to the platform is many months long (roughly 6-9 months). Of these customer wins, several were very notable, including Datadog and Walgreens Boots Alliance.
The company’s balance sheet is healthy, with approximately $5.1 billion in cash, cash equivalents and short-term and long-term investments.
A key part in the Snowflake investment thesis is not just its ability to store and process data, but also to make data sharing easier than ever before. The company outlined that “At the end of the quarter, 15% of our rapidly expanding customer base had data edges (data networking relationships) in place with external Snowflake accounts compared to 10% a year ago. And the number of edges in this period grew 33% quarter-on-quarter.” The more edges get made and used, the more useful (and sticky) Snowflake’s platform becomes. The ease and sheer amount of data sharing further enables the value of the data to grow exponentially.
Regarding guidance, the company is guiding revenues between $235 million and $240 million, representing year-over-year growth between 88% and 92%. This equates to 12% sequential growth from Q1 2021 to Q2 2021, which is exactly the same sequential growth they guided for Q1 2021, only to come in 8 percent higher at 20%. I think it’s very realistic to expect the same thing to happen again. So even though they are forecasting 92% high end growth, it will probably be more like 105% YoY growth on $260mm of revenue.
Part of Snowflake’s business comes from storing data, and the company implemented a compression algorithm that’s so good, it’s going to reduce the company’s revenues by roughly $13 million over the course of fiscal 2022. This might sound like a bad thing to investors, but there is much more money to be made from Snowflake’s compute business. Making storage cheaper will ultimately drive more money into Snowflake’s compute business in the long run. Snowflake indicated they update their compression algorithms roughly every two years, so we can expect this to be something Snowflake does more than once.
An interesting question that came up on the call from an analyst was what advantages does Snowflake have over the hyperscalers (Amazon, Microsoft, Google) that have their own cloud based data warehouse? CEO Frank Slootman’s answer was, to put it simply, better architecture. Snowflake was not architected to straddle the on premise and public cloud environments – it was only made for the public cloud.
Another key aspect driving customers towards Snowflake is governance, a combination of compliance and cybersecurity requirements. Snowflake highlighted its acquisition of a company called CryptoNumerics and its ability to anonymize personal information. This is hyper critical for sharing data and ensuring that personal information is dealt with properly and safely.
No quarter, not even Snowflake’s, is ever perfect. If we temporarily forget about all the new million dollar customers this quarter and turn our attention to net customer adds or fortune 500 customer adds, these numbers came in at a lower rate than previous quarters from a sequential (QoQ) perspective. But company leadership was quick to assert that there will be lumpiness and seasonality in tracking this statistic. Sales reps are much more likely to close new customers, especially big ones such as those in the fortune 500, towards the end of the year, not in Q1.
Another weak point was sequential growth in Remaining Performance Obligation, which clocked in at 7.4%. This appears to be due to seasonality. Last time QoQ (quarter over quarter) growth was about this slow was in the corresponding quarter in 2020.
In summary, Snowflake posted a business as usual quarter. For a company like Snowflake, business as usual is terrific. Based on current guidance and company fundamentals, I am very bullish on Snowflake and will continue to hold for the long term.