Snowflake is one of the fastest growing tech stocks on the market, and at a quick glance, it doesn’t appear cheap. Its market cap is around $100B and its 1-yr fwd P/S multiple is 46x. However, Snowflake is different from other SaaS stocks because the company bills customers based on their consumption rather than a subscription. This is a relatively new approach to software billing, which makes it harder to model and forecast near term sales. However, there are signs that Snowflake’s forward estimates are likely conservative, which artificially increases its multiple and makes it look expensive relative to peers. I explain why in more detail below.

Benefits of Snowflake’s consumption model:

As data creation, ingestion and storage soar in the cloud environment, cloud software providers are starting to migrate away from subscription agreements, which are fixed, to a consumption-based pricing model, which are uncapped. Snowflake is one of the few cloud providers that is nearly 100% consumption-based.

Consumption-based pricing has a few drawbacks. For example, its less predictable than subscription revenue and there isn’t a ‘floor’ on revenue, because if consumption declines then so will sales. However, the flip side is also true, consumption billing does not have a ‘ceiling’ on revenue, so if customer consumption rises, so does sales. This uncapped revenue potential, and Snowflake’s leading market position, sets the company up well to execute in the near term.

What is unique about consumption billing is that it is non-linear, and its harder to predict. However, Snowflake’s sales have started to accelerate (Q3 sales increased 110% YoY up from 103% YoY in Q2) and there are signs that sales will continue to be robust in the near term. Snowflake states that customers generally accelerate spending once they are fully deployed on the platform. Specifically, Snowflake disclosed in its 10Q that:

 “Consumption for most customers accelerates from the beginning of their usage to the end of their contract terms and often exceeds their initial capacity commitment amounts. When this occurs, our customers have the option to amend their existing agreement with us to purchase additional capacity or request early renewals”accelerates from the beginning of their usage to the end of their contract terms and often exceeds their initial capacity commitment amounts. When this occurs, our customers have the option to amend their existing agreement with us to purchase additional capacity or request early renewals”

The company’s NRR metric supports management’s claim that customer spending ramps overtime. For instance, NRR recently reached 173% in the latest quarter, the highest level since it went public and above the level disclosed in its S-1 (169%). The improvement in Snow’s NRR metric highlights that customer spending continues to ramp and that there isn’t a ceiling on total spend. The company’s NRR of 173% is also well above other cloud (SaaS) leaders at around 130%, further highlighting the uncapped nature of consumption based spending relative to subscriptions.

We can also see that growth in enterprise customers has outpaced sales growth. In the most recent quarter, enterprise customers spending >$1m per year increased 128% YoY, faster than Snowflake’s 110% YoY rise in total sales. These customers are also getting bigger each year. For example, Snowflake disclosed that 53% of revenue came from enterprise customers, up from 46% in the year-ago quarter, implying that spending per enterprise customer increased 6% YoY to ~$3.4 million per enterprise customer. The fact that enterprise customers are growing faster than sales and also increasing their spending highlights the uncapped revenue potential of Snowflake’s consumption model.and also increasing their spending highlights the uncapped revenue potential of Snowflake’s consumption model.  

As new customers join the platform and ramp spending, this will be a tailwind for sales going forward. At the same time, existing customer spending continues to rise, leading to multiple tailwinds for topline expansion.

Another key metric that highlights Snowflake’s revenue potential is the amount of future sales under contract. When customers sign onto the platform, they purchase consumption at specified prices, which gets recorded as remaining performance obligations (RPO). RPO increased 94% YoY to $1.8 billion, with a weighted contract length of ~2.5 years.

Looking forward, RPO is equal to a third of the aggregate FY2023 and FY2024 sales estimate (~$5.1 billion). In other words, Snowflake has a third of its forward sales estimates already under contract.

Furthermore, Snowflake’s RPO metric is backed by cash, giving us more visibility into future sales. Since customers are paying for their contracts a year in advance with cash, this demonstrates the pricing power Snowflake has and the strong demand for its products. Generally, customers paying for a service upfront is a sign of strength.

As shown below, deferred revenue is rising with RPO. However, deferred revenue relative to RPO did decline YoY from 48% of RPO to 43% of RPO in the latest quarter. Mgmt said this was due to customers migrating to quarterly billings, away from annual billings. This is a form of payment term extensions, which can temporarily juice sales (if you require less upfront cash, it’s a better deal for customers and incentivizes them to sign up).

While the lower deferred revenue to RPO is a trend to watch, it is not yet concerning in my opinion. In fact, Snowflake is a unique position of commanding upfront cash payments for consumption-based spending. For example, New Relic (NEWR) recently transitioned to a consumption-based billing model and is paid mostly in arrears (or after the fact), while Snowflake is pre-funded with cash. Being paid upfront helps pay for working capital and can be a significant advantage in the long run.

Snowflake is effectively getting the best of both worlds; uncapped revenue potential with consumption spending and upfront cash payments (usually reserved for subscription billing models). This trend demonstrates the demand for its products and improves the quality of revenue, which deserves a premium multiple.

Finally, there are signs that Snowflake’s RPO metric may be understated.understated. As mentioned above, Snowflake states that new customers accelerate their usage after deploying and “often exceeds their initial capacity commitment amounts”. often exceeds their initial capacity commitment amounts”. This statement, which is backed up by the company’s robust NRR metric of 173%, and rising spend per enterprise customer, suggests that there is upside potential in Snow’s RPO metric and future sales growth.

The fact that RPO growth is back-end loaded supports our thesis that forward estimates are likely conservative. This is because customer’s often go over their initial contractual amount, meaning that 1) RPO is likely understated and 2) analysts estimates are likely conservative (assuming they extrapolate forward estimates from trends in RPO). Because of the uncapped nature of Snowflake’s revenue model and the tendency of its customers to accelerate overtime, this makes it difficult to compare Snowflake to SaaS peers, which report more linear growth. I discuss the company’s valuation in more detail next.

Valuation

Snowflake has guided to $10B in sales by FY2029 and its expected share count dilution is forecasted to be <3% per year. This implies a ~10x P/S multiple on FY2029 sales. Sales are expected to rise at a CAGR of 36.5% for the next seven years to reach $10B, and the company’s multiple compression is expected to be ~26.5% per year. There is upside to the company’s valuation if sales grow faster than 36.5% through FY2029 (assuming a 26.5% multiple compression per year).

As of today, Snowflake trades at a 84 fwd P/S multiple (based on FY2022 sales ending in January 2022) and a 46 fwd P/S multiple (based on FY2023 sales ending in January 2023), which is a premium in the tech space. On a two-year forward basis, Snowflake trades at a 29x fwd P/S multiple (FY2024 sales), highlighting the large multiple compression forecasted by the Street going forward. Annual sales growth is expected to slow from 104% in FY2022 down to 66% in FY2023 and 56% in FY2024.

Given the non-linear nature of consumption spending, comparing Snowflake to SaaS peers may not be totally appropriate. Analyst estimates are mostly linear, since they have to be prudent with their estimates and its difficult to predict consumption. However, Snowflake’s robust NRR metric and likely understated RPO metric suggest that there is upside to forward estimates and that current estimates may be too low.

Are analyst estimates conservative?

The company’s strong metrics discussed above highlight the potential upside in future sales growth. For instance, the company’s contracted revenue (RPO) is already equal to a third of the aggregate FY2023 and FY2024 sales estimate, and RPO is backed by cash, further increasing the quality of the metric. With evidence that existing customers are ramping spending (with NRR rising above 170% and spending per enterprise customer also increasing), the argument can be made that RPO is likely understated. This is impressive, considering RPO grew by nearly 100% in the most recent quarter. 

Furthermore, consumption based spending is inherently unpredictable, which makes it difficult to model near term revenues. I believe there is a degree of conservatism priced into forward estimates due to the unpredictable nature of consumption spending, which makes Snowflake appear more expensive. Yet, there are trends that improve the quality of Snowflake’s forward sales, such as its RPO metric discussed above, which may be understated, its cash support backing RPO, and the rapid expansion in customer spending over time.

Looking forward, the Street is pricing in a rapid multiple compression. If growth can remain above trend for the next few years, there is upside potential to its valuation. It is important to remember that consumption growth is non-linear and uncapped, and the company’s metrics suggest that growth will remain robust in the near term. For instance, sales accelerated, NRR increased to over 170% and spending per enterprise customer also rose. Finally, the company is paid upfront for its consumption contracts, which is unique and highlights the strong demand for its platform.

With the fundamental explosion in data creation, ingestion and storage in the cloud environment as tailwinds, Snowflake’s uncapped revenue model is well positioned to benefit from these massive secular trends. The company’s key metrics suggests that sales will remain robust in the near term and we believe that Snowflake is well positioned to outperform going forward.

*Here's Beth's most recent editorial below*

Snowflake Accelerates in Revenue while Tech Growth Sells Off

The company was listed in September 2020 and the shares more than doubled on the day of the listing. It was one of the biggest tech IPOs of all time raising roughly $3 billion with a road show that attracted risk-adverse Warren Buffet.

In my deep dive published on Forbes, I noted Snowflake’s sky-high revenue growth of 173% in the year prior to the IPO. Another key metric that led to the success was the net retention rate of 158%, which was the highest for any cloud company at the time of listing; this metric is even higher now. Snowflake closed the opening day with a market cap of $70.3 billion that was more than five times its last private valuation of about $12.5 billion. Snowflake has been public for over a year and now trades at a market cap of $92.5 billion for a gain of roughly 33%, at time of writing.

Below, we revisit the product and the company’s financials now that it’s been a public company for a decent length of time. The information in this analysis is partly why we have decided to build a position with more information on what makes Snowflake stand apart provided to our premium members.

Snowflake’s Rare Acceleration in Net Retention

There are a few key metrics that Snowflake discloses that help investors better understand the demand for its products and the cadence of its growth going forward. One of the key metrics is its net retention ratio (NRR), which increased YoY from 162% to 173%, the highest level since it went public and a notable (rare) acceleration.

Since Snowflake uses a ‘land-and-expand’ sales strategy, growth with existing customers is critical to scale its business. An increase in its NRR metric demonstrates that customers sign on and then rapidly ramp spending as they fully deploy on Snowflake’s platform.

However, CFO Michael Scarpelli, cautioned investors that NRR will decline going forward, but nevertheless will still remain well above 140% for a “very long time”. Specifically, he stated that, “I'm not going to guide long term. It's hard to do that. I'm just going to reiterate again what I said to Derrick is we will [keep NRR] above 160% for this year. And I do expect longer term as our customer base gets bigger and bigger and more mature, that number will come down, but I still think it will be well above 130%, 140% for a very long time”.

Source: Investor Presentation

While Snowflake uses a “land -and-expand” sales strategy, it also uses a consumption billing model. For instance, Snowflake bills customers based on the amount of data they store and transfer and what resources they use.  Accruing revenue based on consumption rather than a ratable subscription model decreases the predictability of quarterly revenue, but it leaves revenue uncapped. This provides revenue upside, because if consumption soars, then so will revenue. We can see with the NRR metric discussed above that existing customers are ramping consumption. Furthermore, this consumption is also contracted, meaning that a portion of forward topline growth is locked in which provides visibility into future sales.

The company’s robust NRR metric of 173% discussed above also backs up management’s claim that customers often exceed their original contract amounts. Furthermore, since sales are accrued under a consumption model rather than a subscription model, there doesn’t appear to be a ceiling on customer spending. For instance, Snowflake disclosed that customers spending over $1 million (enterprise customers) grew 128% YoY to 148. Moreover, these customers accounted for 53% of total revenue, up from 46% in the year ago quarter. This implies that spending per enterprise customer increased 6% YoY to $3.4 million. While the outsized growth in enterprise count is impressive, it is also great to see spending per enterprise customer rise as well, signaling that large customers keep getting bigger.

As a result of the improvement in the metrics disclosed above, the company’s revenue accelerated by 110% YoY to $334.4 million. The topline growth was very strong and has been above 100% for at least five quarters in a row (since the company went public). Growth was led by financial, media, technology, and retail customers.

The company has a growing customer base. As mentioned above, customers with trailing 12 months product revenue greater than $1 million were 148, up 128% YoY.  Furthermore, total customers increased 52% YoY to 5,416 customers, while Fortune 500 customers grew by 30% YoY to 223.

Source: Company Website

The company is also growing internationally and the growth is higher than the company’s total growth. International revenue which was 14% of the total revenue in the Q3 FY21 has increased to 18% in the Q3 FY22.

In the earnings call, Frank Slootman said, “We continued our international expansion with product revenue from EMEA and Asia-Pacific outstripping the company's year-on-year growth, up 174% and 219% respectively. We recently launched operations in three new countries Israel, Korea, and the United Arab Emirates.”

Another key metric, remaining performance obligation grew by 94% YoY to $1.8 billion. This represents revenue that is contracted but not yet realized.

Of the total $1.8 billion the management expects about 55% to be recognized as revenue in the next one year. Some of the notable large multi-year deals in the recent quarter include a $100 million three-year deal to an existing customer and additional five eight-figure multi-year deals. This is a positive trend that the company has been able to win large contracts.

The company’s margins continue to show improvement. Total gross margin is 64% and adjusted gross margin is 71% when compared to 58% and 67% respectively, for Q3 FY2021. Adjusted product gross margin came in at 74.6% when compared to 73.6% in the previous quarter and 70% in the same period last year.

Net loss came in at $154.9 million or ($0.51) per share compared to $168.9 million or ($1.01) per share for the same period last year.

The company’s free cash flow improved to $9.5 million from a free cash outflow of $37.9 million for Q3 FY2021. Adjusted free cash flow came in at $21.5 million compared to adjusted free cash outflow of $37.1 million in the same period last year. The company has maintained a strong balance sheet as it has cash and investments of about $5.1 billion.

Source: Investor PresentationInvestor Presentation

Product:

Snowflake’s decoupled architecture allows for compute and storage to scale separately with the storage provided from any cloud provider the customer chooses. By processing queries using massively parallel processing (MPP), where each node in the cluster stores a portion of the data set locally, the virtual warehouses can access the storage layer independently so as not to compete for compute power. With the competitors, such as Redshift, where compute and storage are coupled, more time is spent reconfiguring the cluster. 

Snowflake calls this offering a virtual data warehouse where workloads share the same data but can run independently. This is crucial because Snowflake’s competitors combine compute and storage and require customers to size and pay based on the largest workload.

Data warehouses are centralized data repositories that collect and store information across many sources that are both internal and external. The raw data is ingested into the data warehouse and processed to answer queries. One key product differentiator is that Snowflake is not built on Hadoop, rather the company uses a new SQL database engine with cloud-optimized architecture. Overall, this translates to faster queries and also reduces costs by scaling up or down for both capacity and performance. This also allows the shift to the cloud while still honoring traditional relational database tools. Just like cloud infrastructure does not require you to hold server space for peak times year-round, a cloud data warehouse does not require you to plan, acquire or manage resources for peak data demand (i.e. elasticity).

The need for resources could change by either increasing or decreasing (scaling up or down). Customers that have a need for storage but less of a need for CPU computations do not have to pay up front and can shrink the environment dynamically. Users either pay for terabytes or are billed on a per-second basis for computations. As discussed above, Snowflake charges by execution-based usage and is not a cloud SaaS-company that charges by subscription.

Snowflake has a multi-cluster architecture which is unique from single cluster databases. The multi-cluster approach allows the clusters to access the same underlying data yet to run independently. This allows for heavy queries and operations to run very quickly and with fewer errors because the queries are not accessing the same data warehouse.

Beyond the value proposition of separating storage from compute for speed, and also scaling up or down to reduce costs, the third takeaway is that Snowflake is also much easier for customers to use as it’s designed to remove the role of a database administrator for monitoring and/or to tune query performance.

The end goal of choosing Snowflake is that you load data, run queries, and do little else – which is an immense value proposition due to the amount of time wasted prepping, balancing, tuning and monitoring traditional data warehouses originally built for on-premise.

Snowflake is capitalizing on the multi-cloud trend and growing rapidly with customers who want a choice in public cloud provider despite the cloud giants having their own data warehouse systems, such as Amazon Redshift, Azure Synapse and Google Big Query.

In our first article written at the time of Snowflake’s public listing, we discussed competitors Google’s Big Query and Amazon’s RedShift. Big Query has a strong following of about 2X customers compared to Snowflake, growing at 40% and also offers separate storage and compute. The differences between BigQuery and Snowflake include pricing structure where Snowflake is a time-based pricing model where users are charged for execution time and BigQuery is a query-based pricing model, where users are charged for the amount of data returned from the queries. Redshift has growth of 6.5% and is not as competitive due to coupling compute and storage.

In 2020, The Enterprise Technology Research study showed 80% of AWS accounts plan to spend more on Snowflake in 2020 relative to 2019 with 35% adding Snowflake as new compared to 12% adding Redshift as new. In Azure, 78% plan to spend more on Snowflake with 41% adding new. On Google Cloud, 80% plan to increase spending on Snowflake.

Granted, this study was in 2020 but this helps drive home why Big Tech owning the data centers is not a deterrent for Snowflake’s rapid adoption. Judging by Snowflake’s revenue growth, these preferences are likely still intact.

The company also launched support for unstructured data earlier this year, which is another strength compared to the SQL legacy competitors. Due to the increasing use of unstructured data, there is demand to support unstructured data for big data analytics.

Data Sharing and Data Marketplaces

Snowflake allows businesses to share their data with other external businesses on the platform. Data Marketplace allows free or monetized data sets to be exchanged. This has helped Snowflake break into new industries with use cases that other data lakes and competitors do not currently offer.

For example, earlier this year, Snowflake announced support for Unified 2.0, an open sourced and transparent identity framework that will help publishers, advertisers, and its partners identify users. When browser providers like Google plan to eliminate third party cookies, Unified 2.0 is seen as one of the potential replacements by ad tech firms.

In the Q2 earnings call, Jeff Green, the CEO and Founder of The Trade Desk, mentioned, “I think Snowflake adopting UID2 is one of the biggest headlines that has happened for UID to date and not enough has been said about it. I don't think most people understand why this is so big.” He further added, “So in the same way that Wix made it really easy for companies to build websites, Snowflake makes it really easy for companies to put their data to work.”

The management has maintained since its IPO that the opportunity in data sharing is substantial and largely untapped. In the recent earnings call, the CEO mentioned, “Generally I agree with what your assessment that we are just seeing the tip of the iceberg. Snowflake was built from the ground up as a data sharing platform and we've been at it from the beginning. You see a lot of other players following our lead in this regard, but we are in the beginning.”

The company also follows a consumption model, which makes investment decisions easier for its customers to decide which business units need the workloads. The management gave an example of the financial sector in the earnings call. The CEO mentioned, “That really mitigates the sticker shock, people can make investment decisions as they go along and as it warms it, we're seeing with some of our large banking customers as they went from recomputing loan rates on a monthly basis to doing it every night, while they had a business case for.”

In the most recent quarter, Data Marketplace grew 41%, which is “steady” but expected to could expand at a “meteoric rate” due to the non-linear way data sharing expands. The company recently introduced two industry data clouds: Financial Services Data Cloud and Media Data Cloud. The customers include companies like Allianz, Blackrock, New York Stock Exchange, State Street, Disney Advertising Sales, The Trade Desk, and Experian, among others.

Developers Building Apps with Snowpark

Snowpark offers the ability to migrate business logic with popular programming languages Python, Scala/Java Virtual Machine or Java. The library and DataFrame API allow querying and processing data without having to move data to where the application code runs. This extends programming functionality for ML model training and allows data processing to run natively in the data cloud.

Prior to Snowpark, code deployment required separate infrastructure. Building applications that interact with Snowflake’s virtual warehouses minimizes processing time and lowers the learning curve/broadens adoption of complex data pipelines by removing the need to move or copy data into other systems to overcome working with SQL.

The recent announcement of adding Snowpark for Python is key because of Python’s widespread popularity among developers. With the Snowpark Accelerator, Snowflake is courting developers to build more applications and this is likely to help Snowflake maintain a competitive advantage with a newer class of machine learning startups. The company had 23,000 developers register for the last Snow Day event.

As stated, unstructured data has recently become available in public preview, and this is being leveraged through Snowflake’s newer programmability as customers can now store new data types.

Risks

The company’s revenue growth has been exceptional. However, the company is undergoing losses. There is no clarity as to when the company will be profitable on a GAAP basis. In last year’s Investor Day presentation, the company has laid its roadmap to reach $10 billion in annual product revenue in the FY 2029 and adjusted operating income margin of 10%. So, it suggests that the competition is very high for its bottom line to improve significantly.

The company’s current revenue growth rates might not sustain long-term. The management expects long-term product revenue to grow by 30%. Overall revenue growth is down from 174% in FY 2020 to 124% in FY 2021, and for this year, analysts expect revenue to grow about 104% YoY.

Another key risk we will be monitoring is the reduction in payment terms, as Snowflake is migrating from annual upfront invoicing to quarterly upfront invoicing. This reduces the amount of cash customers have to pay upfront, which can temporarily juice sales. We will need to monitor this trend going forward to ensure that growth will be sustainable as customers fully migrate.

Another risk is the company’s consumption billing model, which is inherently unpredictable. This can make growth lumpy and some quarters may disappoint the Street. Investors should expect increased volatility in growth from Snowflake in the near term as new customers ramp consumption. However, management does expect revenue growth to smooth and become more predictable in the aggregate as customer consumption scales and matures on the platform.

Valuation

Considering the company’s strong metrics discussed further above, it makes sense that Snowflake trades at a premium multiple compared to other high growth companies. At time of writing, it’s 1-year fwd P/S multiple is 46x and Snowflake trades at a 29x 2-year forward multiple.

 

Looking forward, management guided Q4 product sales to increase 95% YoY to $348 million at the midpoint and for FY2022 product sales to increase 104% YoY to $1.1 billion. One year forward, the Street expects FY2023 sales to increase 66% YoY to $2.0 billion and then to further deaccelerate to 56% YoY growth in FY2023 and reach $3.1 billion. EBITDA is also expected to turn positive in FY2023 and then rapidly expand to over $260 million by FY2024.

Conclusion

Snowflake separates compute and storage which allows companies to store large amounts of data while running complex queries at high performance. The company also drives down costs for customers newly onboarded due to its pricing model where you pay for only what is used. Snowpark now allows data processing to occur natively on the data cloud instead of external Spark clusters and is opening up complex data pipelines with popular programming languages, such as Python.

The company is disrupting legacy databases while keeping a strong focus on how to lower the barrier of entry for data applications and machine learning workflows. The product is not where the company is often disputed by investors rather it’s the valuation. Those on the sidelines for the past 1.5 years have only given up 30% in gains (in terms of market cap) and have saved themselves a rather rocky, volatile ride. Snowflake has always been a strong company yet the last earnings report was a perfect 10. We think the company may be finally gathering its strength to truly earn its valuation once and for all. The I/O Fund is officially on board, per the disclosure below.

I/O Fund analysts Royston Roche and Bradley Cipriano contributed to this analysis.

Please note: The I/O Fund conducts research and draws conclusions for the Fund’s positions. We then share that information with our readers. This is not a guarantee of a stock’s performance. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. The I/O Fund plans to initiate a position in Snowflake in the next 72 hours.