Source: Bloomberg

TL;DR

Snowflake is a heavy-hitting data warehouse company that’s poised to make a big splash when it goes public in the next few months. Can it out-compete the AWS, Google Cloud, and Azure goliaths?

The upcoming S-1 should shine some light on the company and its prospects, and there are a few things I’m very curious to see.

Setting the stage

For analytics geeks, it’s been awhile since we had an IPO to gawk about.

2017 saw Alteryx and MongoDB go public in pretty close succession, but since then, it’s been a bit of a dry spell.

I know lots of us are excited to see Snowflake go out. It’s the little-non-CSP engine that could. It’s the standalone David fighting an incredibly strong fight against the Azure, AWS, and GCP Goliaths.

Well, it’s now been two months since news of the IPO first leaked… and I, for one, am getting impatient waiting for my S-1 notification. While we wait, I thought I’d break down a little bit about Snowflake, what’s driven their success, and the metrics I’ll be looking out for in the S-1.

What Snowflake is

Snowflake is a SQL-based cloud data warehouse. It was founded in 2012 by Thierry Cruanes, Benoit Dageville, and Marcin Zukowski.

Cruanes and Dageville had been long-timers at Oracle, and they decided it was time to rethink what it meant to be a data warehouse.

They left Oracle, went back to the drawing board, and dreamt up an entirely new cloud database architecture from the ground up.

This is an important point — there’s absolutely no element of porting over a legacy on-prem product, unlike some of their competitors. Snowflake truly has had Cloud in its DNA from day one.

Source: https://www.snowflake.com/blog/beyond-modern-data-architecture/

It was built on the thesis that the cloud’s virtually unlimited storage and incredible reliability could enable so much that was previously unfathomable.

Where they’re at today

They’ve built an astounding business in just eight years.

Since its founding, Snowflake has grown and raised capital fast — wicked fast. Launching in 2012, they absolutely nailed the timing on the enterprise cloud hype cycle, and they’ve been reaping the benefits.

In addition to massive customer traction and becoming a household name in the enterprise data world, the pace and magnitude of their funding rounds speak for themselves.

Source: https://news.crunchbase.com/news/snowflake-gets-479m-reaches-decacorn-status-with-12-4b-valuation/

After achieving a $12.4 billion valuation earlier this year on the heels of their Series G funding round, sources say their go-public valuation could be as high as $20 Billion.

Heading into the IPO, Snowflake has a proven CEO at the helm — Frank Slootman, previously CEO of ServiceNow which he took from startup to IPO and beyond. His resume also includes taking Data Domain from early-stage to IPO.

How Snowflake differentiates

There are many things Snowflake claims to do better than its competitors. In my view, their success boils down to two main themes:

  1. The Architecture

The architecture is just really, really smart. Specifically, the praise I hear most often about Snowflake is how they were the pioneer in separating Storage and Compute — something most databases before them hadn’t done well.

The separation of Storage and Compute is a key enabler in so much of what they do.

The most touted benefit is pricing —“pay only for what you use” — but this architecture is also critical in enabling product features that set them apart, including:

  • Instant Scaling and Concurrency
  • Zero copy cloning
  • Data sharing

That said, there are questions around whether this engineering advantage is sustainable in the longer-term.

2. User Experience

A lot of people would roll their eyes at calling out UX a differentiator — but in the data infrastructure space, I really think it is.

Fantastic, user-driven design is unfortunately still far from table stakes. Developers and Analysts often grin and bear it with applications that are, quite frankly, pretty miserable to work with.

Building a delightful, easy-to-navigate product has been key to Snowflake’s success.

Source: https://docs.snowflake.com/en/user-guide/snowflake-manager.html

They’ve made a conscious attempt to really figure things out for the Analyst and Developer, rather than putting the onus on them — and it shows.

Also, their documentation is fantastic. It’s approachable even to less technical folks — it feels like their documentation speaks English, while other enterprise software vendors speak Data.

All in all, they’ve massively driven down the end-user complexity of a highly complex system by putting user experience front and center. I think the importance of this is hugely understated in most public discourse about the company.

Also, Multi-cloud…

While I won’t formally add it as a #3, I’d be remiss if I didn’t at least mention Multi-cloud.

Until now, the fact that you can run Snowflake on any combination of AWS, Azure, or GCP has been a differentiator.

Although this has helped them stand out in the past, this is beginning to change. I expect Multi-cloud to become the new data warehouse standard, and with Google’s BigQuery Omni announcement last month, that’s starting to come to fruition.

Metrics I’ll be looking out for in the S-1

Services as a % of Revenue:

The first thing I’m very curious to know is how Services-heavy Snowflake is. Database technologies typically come alongside substantial Professional Services (implementation, 24/7 customer support, etc.) that are one-time, often non-recurring revenue streams that tend to be very low margin.

It is clear Snowflake offices services — and the first thing I’ll look for in their Income Statement is “how much”. This will have an impact on the next metric, Gross Profit.

Source: https://www.snowflake.com/snowflake-professional-services/

Gross Profit:

I expect Snowflake to come out of the gate with higher gross margins than the other shops are running. It never had an open-source strategy, it’s a true cloud-first SaaS platform, and while I could easily be proven wrong, I expect its product numbers to way outnumber its services numbers.

Comparable companies’ Gross Margins for comparison:

Sources: 2019 Annual Reports — MongoDB, Teradata

Unfortunately, of course, we can’t compare gross margins of the major cloud platform data warehouses (e.g., Amazon doesn’t report Redshift-specific financials, unfortunately) — so this is the best we got.

Net Loss / Revenue:

How far is Snowflake from reaching profitability? Are they going to be another Uber that goes public and has no clear path to cashflow breakeven, or will they break the 2020 tech mold and come out EBITDA-positive?

In these times of economic uncertainty, this is something to keep a watchful eye on.

What next?

We sit back, grab a beer, enjoy our summers, and wait. Once the S-1 comes out, you can bet I’ll be all over it.

If you’re thinking about the Snowflake IPO or the evolution of the analytics & data infrastructure space in general, I’d love to chat. I’m @AnnikaSays on Twitter.