A Snowflake deep dive

Hey Fools,

Here is my deep dive on Snowflake’s architecture & platform. They are about to IPO - ticker symbol will be SNOW. (And I once again I have to say - it is so much nicer having embeddable images and drawings right in my posts. I hope the drawings makes it easier to grok tech architectures.)

https://hhhypergrowth.com/a-snowflake-deep-dive/

TL;DR:

Snowflake wants to be the engine that drives all analytics – both business intelligence (BI) and data science (ML/AI) – over the entire enterprise. They provide a turnkey cloud-native, scalable database platform with a heavy focus on enterprise concerns of security, governance & compliance, while having a rich ecosystem of integrations & tooling available for ingest, BI and analytics.

I was pretty amazed by the sheer number of use cases their platform enables. In particular, how the platform allows for the secure sharing & exchange of data between customers and outside users. This includes a Data Marketplace, which enables the rise of the Data Broker.

Customers choose Snowflake for its cost, performance, scalability, ease of use, secure sharing, and flexibility. Investors will choose Snowflake for its optionality and ever increasing TAM.

Data will always grow, and analytics are becoming more and more crucial. Snowflake = A Data + Analytics Cloud. It is a data cloud-within-a-cloud, that also acts like Hadoop-as-a-service, allowing data scientists to run analytical code directly inside their compute.


I have seen a lot of recent takes on Snowflake, including tech dives & investment theses, and… I wish I could sit with a bright red marker and highlight the mistakes, and especially the omissions. Maybe I will just berate the authors on Twitter.(Nah, being a jerk is not my style.)

Most takes continue to focus solely on Data Warehouse (where they started), and don’t mention Data Lake (the entire core of their platform!) or all the other product lines that spawned from their Data Lake + Data Warehouse architecture like Data Exchange or Data Marketplace.

To me, it’s like calling Fastly “just a CDN”. So I am glad to try to help the individual investors here @ Saul’s get their tech straight. Analysts want to gloss over the details and just get a boiled down version - or know just enough about the underlying tech that it causes them to look at the company with blinders on from pre-conceptions. Keep boiling it down too much and you cannot see the finer details, where often times the real value emerges in these highly sticky platforms!

So set aside an hour and dive in. (Or, maybe two or three times through for the luddites out there.) At a minimum, you will walk away understanding all the terms being bandied about in all the other Snowflake takes. At a maximum, you will come to understand the potential of this company well.

Just look at these numbers again:


Revenue 133M +121%
TTM Revenue 403B +137%
RPO 688M +211% 
Gross Margins 62% +1300bps
$NER 158% 
Custs 3117 +101%
- Custs >1M 56 +155%

And the prior 8 Qs of the top line:


| Period |  Rev |  QoQ |   YoY |
|--------|------|------|-------|
| Q221   |  133 | +22% | +122% |
| Q121   |  109 | +24% | +148% |
| Q420   |   88 | +21% | +138% |
| Q320   |   73 | +22% | +152% |
| Q220   |   60 | +36% |       |
| Q120   |   44 | +19% |       | 
| Q419   |   37 | +28% |       |
| Q319   |   29 |      |       | 

For a followup companion piece to mine, w/ a fantastic investment view of the numbers plus comparables to the SaaS industry (which I love), I highly recommend this piece by the Public Comps blog: https://blog.publiccomps.com/snowflake-s1-ipo-teardown/

Unfortunately Snowflake has a lot of attention (damn you Warren Buffet!), and valuation is due to be astronomical, plus this is a full backbench of investors skipping IPO day that are eager to buy on dips. Really wish I could have invested in them a year ago when I first noticed them!

But regardless of day 1 valuation, as we know from Saul - scaled hypergrowth + operational leverage = scaled profits. This company has all the hallmarks of a top-tier Saul stock. It is bound to execute extremely well, as customers are flocking to it, and are spending more and more. All this in an industry (Data + Analytics) that is on fire and growing for years to come.

Hope you enjoy.

-muji
@hhhypergrowth

PS I talk about AYX and MDB a bit too (and not exactly in a positive way).

175 Likes

That was an incredible deep dive!

It’s really going to give companies a leg up over their competitors - since they’ll be able to innovate more quickly than their peers just by the insane efficiencies they’re going to realize.

A few notable standouts as current customers.

Hubspot and Square.

I’d say I’m more interested in those two companies because they use this incredible software. Others will follow suit I’m sure but I wonder how much Snowflake is responsible for how well these companies are doing…?

1 Like

Great work. I appreciate that muji.

I think it was Brian Stoffel (TMFCheesehead) who mentioned he thought of ROKU in order to help with understanding what SNOW is doing. I do see the similarities.

With online video streaming you have huge competing providers, and a lot of subscribers that are coming to a neutral place, ROKU. ROKU helps subscribers get to the content, and helps providers find a greater audience for their content. But at the same time ROKU is in competition with the providers.

SNOW has this same relationship with the cloud titans. Amazon is in both spaces (Prime Video and AWS)

A couple of quotes from https://retailroadshow.com/ by the CEO Frank Slootman

“In some accounts we are the second largest line item behind the public cloud spend itself”. It would be nice to know which ones.

“We see the beginnings of powerful data network effects. Another reason to be on Snowflake becomes whoever else has data on Snowflake. Both data providers and data consumers need to be on Snowflake for this reason. That is the network effect.”

Thanks again for the write-up.
HN

3 Likes

Great write up muji,
I’ve been pondering something about, “Data” lately. I wonder if last year we weren’t at the peak of the gartner hype cycle for data, data analysis, data lakes, data warehouses, etc etc etc and we are now on the downslope until we hit the trough of disillusionment.

If any of you aren’t familiar with what I’m talking about here is a lovely little graph. https://images.app.goo.gl/bCDM4SaMyQp35dXT9.

The reason I say this is I keep reading about and talking to data scientists that many companies have been burned by there big data projects. They aren’t seeing the ROI they expected or are abandoning projects because they just aren’t useful. I think that snowflake has been in the right place in the right time, their growth and investments in them show this. However i’m wondering if that can continue going forward.

Having said that, I don’t see the need for data storage decreasing any time soon.

Thoughts?

8 Likes

I might have missed it if it was already mentioned here but it looks like SNOW increased their IPO price from a range of $75-85 to a range of $100-110, and it is likely to pop quite a bit above that as soon as it starts trading (tomorrow?)

https://www.marketwatch.com/story/cloud-company-snowflake-ra…

I wonder if this impacts the pricing of the shares that Berkshire got at $80, or if they still got that sweet deal

-mekong

2 Likes

Great write up.

A couple other things I gathered from the S-1 is their land and expand strategy is much more aggressive than other companies.

Up until a few quarters ago they were spending more on sales and marketing than they were generating in revenue. At first I viewed this as a negative statistic, but after second thought I think it is good with such high net retention. Any sale they get is going to be worth it.

Both the CEO and CFO are from ServiceNow during their growth years so leadership seems to have good experience in growing a company at this stage.

Also think the transfer of data from one cloud provider to another is a major leg up on the competition. A lot of enterprises do not want to just rely on AWS for everything in case AWS has an outage.

2 Likes