Quick S-1 Teardown: Snowflake

The Snowflake IPO is shaping up to be particularly exciting.  Their S-1 shows very impressive metrics across the board, including explosive revenue growth at scale (growing 174% annually to $264.7 million for the fiscal year ended January 31, 2020), and “land and expand” motion (169% net revenue retention in 2020), making Snowflake one of the fastest growing enterprise companies ever.

In addition to the intrinsic merits of the company, this is yet another example showing how gigantic the market is for data technologies (storage, analytics, machine learning, etc.).  Snowflake estimates its addressable market at $81B.  

We’ve had the pleasure of hosting Snowflake’s former CEO, Bob Muglia, a couple of times at our Data Driven NYC event of the years (see videos below), and it’s been really fun to watch the company grow. 

Here’s our quick take (with Avery Klemmer) on the IPO and company.

HIGH LEVEL THOUGHTS

Some highlights, big and small:

  • Competing head-on with the cloud giants
    • In the early years, Snowflake built their entire business on top of AWS (S3 in particular)… while competing directly with an AWS product (Redshift).  Over time, they added Azure and GCP, all of which have competing products.  
    • This is a remarkable example of the fluid friend/foe relationships you sometimes see in the tech industry – ultimately the cloud platforms benefit a lot from having Snowflake customers use their underlying cloud computing resources, although Snowflake competes with some of their own products. 
    • As many software infrastructure startups and investors worry about getting crushed by the cloud giants, Snowflake shows that it can be done – but it will require an ability to raise very large amounts of venture capital money ($1.4B over 8 rounds) 
    • Worth noting that Snowflake has been able to build a business with solid gross margins (currently at 61%)  – contrary to the perception that infrastructure businesses built on top of public cloud vendors are ultimately just “resellers” of cloud resources.  However, the S-1 makes it clear that the key to this is a lot of scale: “Our improved gross margin during the last four quarters presented is primarily attributable to higher volume-based discounts for purchases of third-party cloud infrastructure, increased scale across our cloud infrastructure regions, and improved platform pricing discipline.
  • Top down GTM velocity: while there may be some level of survivor bias in this statement, Snowflake has grown remarkably quickly (ultimately, this is just an 8 or 9 year old company, founded in 2012) with a fairly classic “top down” enterprise sales model (AEs, inside sales, partnerships).  While they offer 30-days trials, they don’t use much of the GTM tactics that are increasingly regarded as necessary for rapid market expansion for enterprise companies:  bottoms up, free tier, open source
  • Consumption-Based Pricing model: Snowflake has a different take on pricing, as they believe that standard SaaS (subscription) business model “often results in customers paying for unused software. We believe that business models are evolving from a fixed to a utility model, where customers pay only for the resources they consume.”
  • Network Effects: it is fairly rare for enterprise software companies to have network effects.  While I don’t have full clarity on how developed this truly is, Snowflake has a very interesting take around the old idea of a “data marketplace”, where they encourage the sharing of data among Snowflake customers.   “The more customers adopt our platform, the more data can be exchanged with other Snowflake customers, partners, and data providers, enhancing the value of our platform for all users. We believe this network effect will help us drive our vision of the Data Cloud.
  • Some violation of the Silicon Valley narrative:
    • For a youth obsessed culture, it’s worth noting that folks in the Snowflake management team and on the board are in their 50s and 60s. For this type of play, experience matters.
    • Contrary to the founder-friendly narrative of the founder-led public company, Snowflake has had multiple CEOs over the years.  The company’s first CEO was Mike Speiser, a venture capitalist at Sutter Hill Ventures, which incubated Snowflake. From June 2014 to May 2019, the company was run by former Microsoft exec Bob Muglia. Since May 2019, the company has been run by Frank Slootman, formerly of ServiceNow.

Here are the two chats we had with Bob Muglia at Data Driven NYC – really interesting to travel back in time and see how Snowflake thought of itself and the market over the years.

Bob Muglia at Data Driven NYC in April 2015

Bob Muglia at Data Driven NYC in April 2018

S-1 NOTES:

  • Product = storage, compute, cloud services
    • Storage can take in structured and semi-structured data
    • Compute provides concurrent access to data across a team and performs data tasks in low latency way
    • Cloud services is “the brain” that optimizes each use case
    • Multi-cloud across 22 regions
  • 507M avg. daily queries in 2020
  • Estimated TAM = $81B
  • Business model 
    • Charge for capacity — either annual capacity arrangements or on-demand — and recognize as compute is used
      • “On-demand” = initial customer onboarding and overage consumption < 10% of total revenue
      • If customers stay under, they can rollover some credits
    • Product revenue = storage, compute & data transfer    
    • Revenue recognized on consumption, which creates challenges in forecasting precisely when contracted revenue will come in
  • Customers
    • 3,117 customers as of July 31 2020, up from 1,547 as of July 31, 2019
    • 7 of Fortune 10 (4% revenue)
    • 146 of Fortune 500 (26% revenue)
    • Capital One = 11% of revenue for FY 2020, down from 17% in FY 2019
  • Key business metrics:
    • Product revenue (FY 2019, 2020): $95.7M, $252.2M (164% YoY growth)
      • Ended Q2 at $908M run rate (126% growth)
    • Net retention (YE 2019, 2020): 180%, 169%
    • Remaining performance obligations (YE 2019, 2020): $128M ,$426M
      • RPO accounts for all the consumption capacity they have sold but not yet delivered
  • GTM
    • Continue to invest in large accounts (more likely to be undergoing cloud transformation, have a large budget) because they have proven $1M+ ACV 
    • Increased $1M+ ACV customers from 22 Q2 2019 to 56 Q2 2020
  • Financial Results:
    • Revenue (FY19, 20): $96.7M, $264.75M
      • Prices went up 12% (better discounting discipline) 
      • $1M+ ACV customers represent 47% of revenue
    • Gross Profit (FY19, 20): $44.9M, $148.2M (46%, 56% blended gross margin)
      • Product gross margin: 57%, 62% (gross margin improvements came from discounting discipline, higher volume-based discounts with third party cloud providers)
    • S&M Spend (FY19, 20): $125.6M, $293.6M (130%, 111% of revenue)
    • R&D Spend (FY19, 20): $68.7M, $105.2M (71%, 40%)
    • *Note: Fiscal year ends Jan 31
    • Ownership pre-IPO:
      • CEO, Frank Slootman: 5.9%
      • Altimeter: 14.8%
      • ICONIQ: 13.8%
      • Redpoint: 9%
      • Sequoia: 8.4%
      • Sutter Hill Ventures: 20.3%

Leave a Reply

Your email address will not be published. Required fields are marked *