Wolfe Research published a research note on April 25th, initiating coverage of Snowflake (SNOW) with an outperform rating. While I conduct my own research before making an investment decision, their rationalization highlighted a few strengths of the Snowflake thesis that I agree with. These included the ability to generate meaningful FCF over time and new tailwinds from the AWS relationship, which is the source of most of the $1.2B in new hyperscaler partnership bookings over the last year. See my prior posts for more background on the Snowflake investment thesis and their evolving relationship with the hyperscalers.
A significant growth driver that Wolfe highlighted was the potential of the Snowflake Powered By program. Wolfe referenced “multiple VC’s” (they spoke to several) who are funding companies to build native data services on the Snowflake platform. If you aren’t familiar with Powered By, the program allows companies to harness the Snowflake data processing platform to run their data-centric businesses. Snowflake lists a number of current customer examples, including Aladdin by Blackrock and parts of the Adobe Experience Cloud. There is also an observability company Observe and a security analytics company Lacework that participate. The promotional video for the program also highlights the migration of Twilio’s SendGrid marketing product to run on top of Snowflake. Instacart is a recent customer highlight, building their retailer insight product using Snowflake’s data infrastructure. I wasn’t aware of those two.
Participants in the Powered By program not only get access to the data platform, but Snowflake provides additional support through access to technical resources, architectural design and co-marketing. The benefit to Snowflake, of course, is that these companies generate incremental utilization of their underlying compute and storage engine. Consumption of these resources is how Snowflake generates revenue.
I think this is a smart strategy on Snowflake’s part as it expands their TAM. Their primary customer base is still those large enterprises that want to modernize their data analytics capabilities by migrating to Snowflake’s cloud data platform. In this case, the target customer department is the data analytics team. With the Powered By program, the customer is using Snowflake’s platform as the foundation of their product infrastructure. This makes Snowflake more similar to a hyperscaler or PaaS offering than a traditional data warehouse or analytics vendor. This strategy is similar to the Salesforce Platform, which allows partners to build CRM adjacent services on top of Salesforce’s infrastructure.
In this way, it opens up a new revenue stream for Snowflake. It also allows them to get exposure to application market segments, which might otherwise be an avenue for expansion. Rather than trying to launch their own products in consumer marketing, observability or financial services to grow revenues, Snowflake is enabling category specialists to build their own businesses on top of Snowflake’s data infrastructure and management services. These companies benefit from the scale and operational efficiency that Snowflake has already built, presumably realizing faster time to market and less staffing than if they tried to spin up their own infrastructure.
As part of the Powered By program, Snowflake has identified 5 target market segments for their offering:
- Customer 360: Marketing or sales automation applications that benefit from a complete view of the customer. This view is used to drive targeted customer interactions, like email campaigns. With the move to first party data, these capabilities are becoming critical for all retailers to have. Offerings from Twilio SendGrid and Instacart fall into this segment.
- IoT: Enables the consumption of time series data from fleets of smart devices and sensors. The advantage to Powered By customers is that ingestion, analysis and storage of large amounts of device data requires enormous compute and storage resources. Additionally, data pipelines can be complicated to set up and integrate with data sources. Snowflake has done most of the heavy lifting here.
- Application health and security analytics. This one does what it sounds like, with the target use cases being observability and SIEM. I don’t think companies building these solutions on top of Snowflake will displace Datadog, Splunk, Crowdstrike and other established players. But, they could focus on sub-segments within these markets. Examples might be in observability of specific closed systems, like cellular networks, fraud detection or manufacturing lines, where the broader observability platforms may not created a customized solution.
- Machine learning and data science. Combined with popular machine learning tools (who are technology partners), Powered By companies could build sophisticated prediction services, like product recommendations, retailer optimization, supply chain management, vehicle maintenance schedules, etc. While a company could spin up this infrastructure themselves, by using Powered By, users can take advantage of Snowflake’s data sharing and clean room capabilities to minimize data copying.
- Embedded analytics data apps: Customers can deliver rich visualizations and dashboards directly from Snowflake’s data storage and processing engine. I think this segment is less about a stand-alone company building a product offering, and more about highlighting Snowflake’s move up into the application stack. Normally, to deliver a data rich application, an enterprise engineering team would provision a separate database and app delivery infrastructure. By reducing query times and increasing concurrency, Snowflake is making a case that these data applications could be run directly on top of Snowflake, eliminating the extra hosting infrastructure and data copying. The Streamlit acquisition further supports this move by offering a development framework for building these data-intensive apps. The difference from standard OLTP databases is that the write load would be minimal for these applications.
At the end of Q4, Snowflake reported 285 companies participating in the Powered By program. As we look forward, utilization of the Powered By program should support the durability of Snowflake’s overall revenue growth. It would accomplish this by creating a new revenue stream that is separate from Snowflake’s traditional reliance on data warehouse migrations and powering analytics workloads for enterprise customers. Along with growth in data sharing relationships and the data marketplace, the Powered By program provides another vector to increase utilization of the Snowflake platform. Growth of participants in this program will provide us with another measure of Snowflake’s progress.
NOTE: This article does not represent investment advice and is solely the author’s opinion for managing his own investment portfolio. Readers are expected to perform their own due diligence before making investment decisions. Please see the Disclaimer for more detail.
Thank you very much for another insightful post.
It looks like market is entering a ‘recession’ mode so I wonder how do CTOs like you view Snowflake workloads if you have to maintain/cut IT budget. I feel database is very core function for sure but some analytics workloads might just be ‘nice to have’?
Once again thanks for consistently sharing very insightful ideas.
Good question and thanks for the feedback. Overall, I think Snowflake utilization would remain stable during a recession. Much of the usage drives reporting, insights and business functions that would be considered important in a challenging environment. If a CTO were forced to reduce budget, they could cut back on consumption of some of the ad hoc querying that occurs by analysts on Snowflake. This was estimated by management as being about 30% of usage (versus scheduled jobs) during the last earnings call. However, you could argue that this type of querying would still be needed as departments (like sales, marketing, operations) make requests of analysts to run different business scenarios or generate insights.