SnowflakeDB’s S-1: The Fine Print (Whiteboard Confessional)

Episode Summary

Join Pete Cheslock and me as we continue the Whiteboard Confessional series with a look at SnowflakeDB’s recent S-1 filing. We touch upon the genesis of Snowflake and how it’s raised over $1 billion in venture capital, how Snowflake’s growth can be attributed to shortcomings on the AWS side, how the product is sticky and usage grows over time, the fascinating conflict of interest buried in Snowflake’s S-1 with respect to the CEO’s private jet, how it’s unlikely that “Susie from accounting” will make out like a bandit in the IPO but we’re rooting for her anyway, and more.

Episode Show Notes & Transcript

About Corey Quinn
Over the course of my career, I’ve worn many different hats in the tech world: systems administrator, systems engineer, director of technical operations, and director of DevOps, to name a few. Today, I’m a cloud economist at The Duckbill Group, the author of the weekly Last Week in AWS newsletter, and the host of two podcasts: Screaming in the Cloud and, you guessed it, AWS Morning Brief, which you’re about to listen to.


Links


Transcript



Corey: This episode is brought to you by Trend Micro Cloud One™. A security services platform for organizations building in the Cloud. I know you're thinking that that's a mouthful because it is, but what's easier to say? “I'm glad we have Trend Micro Cloud One™, a security services platform for organizations building in the Cloud,” or, “Hey, bad news. It's going to be a few more weeks. I kind of forgot about that security thing.” I thought so. Trend Micro Cloud One™ is an automated, flexible all-in-one solution that protects your workflows and containers with cloud-native security. Identify and resolve security issues earlier in the pipeline, and access your cloud environments sooner, with full visibility, so you can get back to what you do best, which is generally building great applications. Discover Trend Micro Cloud One™ a security services platform for organizations building in the Cloud. Whew. At trendmicro.com/screaming.



Corey: Welcome to the AWS Morning Brief: Whiteboard Confessional series, where I am joined once again by my colleague, Pete Cheslock. Pete, thanks for taking the time to tolerate my slings, arrows, and other various forms of cynicism.



Pete: You know, I didn't take that much offense to the fact that I have an MBA, so I decided to come back and see if we can make use of that investment.



Corey: So, fun story. The last one of these that we did was talking about—who's one was that? They're starting to run together at this point.



Pete: That was Sumo Logic’s.



Corey: That's right. And it was, “Oh, let’s talk about what they're doing.” And then throughout the day, I think five tech companies all filed to go public, which is just bizarre. So, we're going to take a couple more episodes to slice and dice a couple more that were of interest to us.



Pete: Yeah, absolutely. We're going to chat about one that I was honestly been waiting for because of the hype and the myths around this company. But it's a big data company called Snowflake.



Corey: They're very special and unique.



Pete: They're very special. I think—I often will listen to CNBC in the background, it's kind of interesting to get little words, and sometimes tech pops up into a CNBC broadcast. When Snowflake filed. I think one of the announcers had said something to the effect of, “I don't know why you'd want to be called Snowflake.” [laughs]. So, I had a good chuckle at that one.



Corey: Because they've been around longer then that's been a disparaging term used by jerks.



Pete: [laughs]. Exactly, exactly. So, they filed their S-1 in that flurry with a whole slew of other companies—which we will definitely get to at least one more of those—and honestly, this company, I've never worked there. I did at one point go through a sales process which I can share some of my thoughts and opinions there, but the reason why I was so excited to see this one is because of the sheer amount of VC money that this company has raised, well over—I don't know if ‘well over’ but definitely over a billion dollars of VC funding raised. It's crazy.



Corey: My comment at one of the big tech conferences last year—back when that was a thing we went to—was I was walking around their booth, and I noticed that they had this mock-up of a race car suspended in the air. And then I realized, “Oh, my God, that isn't a mock-up.” Which told me at that point that if you're paying retail pricing for Snowflake, you're probably doing something very wrong.



Pete: Yeah, absolutely. I think to dive into one of my favorite Snowflake stories, at a previous company, we were checking Snowflake out—we got connected with them via some connections our head of product had and some success that we heard that Snowflake had with helping, you know, a data warehouse. That's what it is: it's a data warehouse technology. If you're in the Amazon ecosystem, you might be using Redshift. Snowflake can do some of those things, it can do some other things.



Corey: Why would I use something like Snowflake instead of Redshift? I mean, for starters, naive approach as well, okay, this is in a different Amazon account, so at minimum, I'm going to be paying data transfer in and out on both sides. But again, we're talking data warehousing, the data transfer is usually something of a rounding error compared to all the extra cost goes into that.



Pete: And this is where I think a lot of their growth in the early days came from a lot of the deficiencies in Redshift. In technologies, in the investment that Amazon was doing there, Snowflake could do a lot of things just simply better. I think additionally, too, they were probably taking a lot of business from Oracle shops and things of that nature. But I do know a friend of mine at his company, they had a well over a million dollars a month in Redshift spend, and they actually moved over to Snowflake as a cost-savings initiative. It was significantly cheaper. 



But what’s, I think, so fascinating, when I heard that I was like, “Well, hold on a second. You know, Snowflake runs inside of Amazon.” So, I'm always curious of how that relationship exists with Amazon where you've got some account manager who's going to lose on some big spend of an Amazon customer by their Redshift spend going down dramatically, but then whoever the account manager for Snowflake must just be super excited by that because obviously their spend is going to go up.



Corey: Yeah, on some level, if you're running a data warehouse on top of AWS, from the high-level AWS perspective, well, is it spend that’s going to happen on your account, or is it spend that’s going to happen on Snowflakes account? It's not likely that you're going to be building everything on top of AWS, and then Snowflake is going to be running its stuff on another provider. The data transfer charges there become exceedingly non-trivial.



Pete: Yeah, absolutely. One of the things that is interesting about how Snowflake works, at least from my recollection a few years ago, is that you can stream your data into S3 which is very cost-effective. Snowflake can actually ingest your data from S3, and what they basically do is they put it into their S3. And you pay the same S3 pricing. I remember the sales guy. 



He was like, “Yep, it’s just pass through pricing. You pay what we pay.” But then I said to him, I was like, “Okay, cool. Well, how is the rest of Snowflake charged?” And he said, “Well, most people just buy credits because it's all consumption-based, right? The more you use, the more you spend.” I was like, “Oh, okay. Well, how does the credit work?”



Corey: So, nondeterministic is what I'm hearing there.



Pete: Yeah, like, “How is it priced? How many credits?” “Eh, most people just buy, like, 100,000 in credits and see where it takes them.” And I'm like, “Okay. So, I just going to cut you a $100,000 check not knowing if that's going to last me one month, 12 months, 24 months?” It did not go well. [laughs].



Corey: Just pour some money in and when it runs out, well pour some more money in.



Pete: Exactly. So, it felt a little weird from that perspective. We ended up not going with it for other reasons, more so we needed something a little bit more real-time, and we had a definitely a tradition at that company to build a lot of it ourselves.



Corey: It happens to the best of us.



Pete: You know, build or buy. Sometimes you just want to build stuff I guess.



Corey: So, do we know what they're built on top of? Is it something that is disclosed? Is it their own special secret sauce? In other words, what makes a Snowflake snowy?



Pete: What is the special Snowflake behind the avalanche of data? I'm available for marketing, too, Snowflake needs some help. But the rumor that's been out there is I think it's been reported on, and so it’s, again, hard to say of what they're doing today, but at least in the past, I had heard and definitely read things online about this, that there was some FoundationDB was actually powering some of their metadata layer. FoundationDB was actually a closed-source database, and the only other company I was aware of that was using it for their SaaS product was a company called Wavefront, which VMware acquired; it was a monitoring company. And from people that I know who have used it have said extremely good things about it. What ended up happening to Foundation is that Apple actually acquired it but then I believe they open-sourced it, so it might actually be an open-source product now. But it's definitely was in the past used by it, and my assumption is like, most databases it’s probably still the case today.



Corey: It's an interesting problem. I feel like on some level, the thing that probably annoys AWS the most is that they have their Snowmobile, and their Snowball, and their Snowcone, and Snowflake would fit in so perfectly, except there's a whole company there that frankly, does way better marketing than Amazon seems to. So, I feel like that is probably a sticking point over there.



Pete: Well, so I think that's a good segue into the fun bits of the S-1, which is, how did it look? You got this company that has been claiming having this stratospheric growth, raising a billion dollars; you got to have some sort of success to raise that kind of money. They had hired this legendary executive out of Microsoft, Bob Muglia, to be the CEO for a period of time, although he's no longer there. And diving into it, some of the high-level stuff that I saw was actually pretty impressive. Their revenue growth is still over 100 percent year over year, and they're doing hundreds of millions of dollars. 



I mean, I think they are on track by next year to be doing half a billion dollars in sales at a top-line revenue. And that is growing at 120 percent or something, I think, with some of the things that I saw. It's pretty impressive where you can grow at that rate. It's very similar to what I was seeing with Datadog, and they had a pretty monster IPO, a big company with a big wind behind it. Much different fundraising round. I'm pretty sure Datadog raised like, what, maybe 100 million or less. 



So, very interesting how companies are capitalized. But I think the other thing, too, that I was really blown away by—but it should make sense—is they have this—their retention of their customers, called net dollar retention, it's over 100 percent. It basically means that they have to spend a lot of money to bring these customers on, but once the customers are there, they expand their usage. And again, if you've ever used a SaaS service, you can probably understand how easy it is to grow over time. And this seems to be a very sticky service and that means that a lot of their clients they bring on board, they're spending more by the next year, which is always a good sign.



Corey: This episode is sponsored in part by ChaosSearch. Now their name isn’t in all caps, so they’re definitely worth talking to. What is ChaosSearch? A scalable log analysis service that lets you add new workloads in minutes, not days or weeks. Click. Boom. Done. ChaosSearch is for you if you’re trying to get a handle on processing multiple terabytes, or more, of log and event data per day, at a disruptive price. One more thing, for those of you that have been down this path of disappointment before, ChaosSearch is a fully managed solution that isn’t playing marketing games when they say “fully managed.” The data lives within your S3 buckets, and that’s really all you have to care about. No managing of servers, but also no data movement. Check them out at chaossearch.io and tell them Corey sent you. Watch for the wince when you say my name. That’s chaossearch.io.



Corey: And it seems like it's definitely everything's up into the right. And also, let's not kid ourselves, data warehouses don't get smaller over time. It's very sticky revenue, it doesn't go away on its own, and when WeWork on AWS bills, one of the questions we ask that we already know the answer to when we ask it is, “Do you need all of this old data?” And the data science team rages, “Of course, we need this very detailed transaction logs from 2012. There's the future of our business in there as soon as we figure out how to unlock it.” I have my own skepticism, but okay, I get it. The industry clearly disagrees with my perspective on these things.



Pete: Yeah, no one likes to delete data. But it's funny, data is a liability as much as it is an asset. In your emails that you send out talking about Amazon S3 Breach Awards—what is it—the S3 Negligence Award, there are a lot of liabilities that come with having data. And for a company like Snowflake, they have all this data in S3. S3 and their, obviously, metadata layer. But there's a risk there that people have to think about as the complexity of where your data lives and who owns it. But you're right, I think, once you have it, you're going to keep using it, and you're going to use more of it.



Corey: One of the interesting things that I always like to look at in S-1s, personally, is conflicts of interest. And on some level, again, everyone has some and I get that, but we promote family values here almost as often as we promote family members. And the air travel stuff that was buried in there S-1 is fascinating to me.



Pete: Yeah. This one was, uh, as I roll through it—again, they have to disclose the stuff. You don't want to not disclose it; there's going to be real issues if you don't, but the newest CEO, who started when Bob Muglia left is named Frank Slootman and he has an aircraft. Like, he owns a plane. Again, I don't know what kind of plane this is. Is it a jet, is it—I don't know anything about private airlines, but it is an aircraft that's in a pool of aircraft by this CTP aviation company, and they're a charter aircraft company. 



So, in the S-1, it talks about how the business, the company actually books charter air travel services for the CEO and other employees through the same company, and they even say, “Hey, from time to time, we actually use the CEOs plane for these business trips as well.” And because of that, because of this arrangement that they have, they essentially have to pay a certain amount of money for these flights. And they even listed out how much they pay, were—I think they listed it—let's see here… for the fiscal year ending 2020, it was nearly $300,000 they paid for the use of this private plane. Crazy.



Corey: And it's funny to me—and I'm not calling this out as any form of grift or whatnot—they paid this guy in total comp last year alone $60 million. Which again, given what he's done to the company seems to be reasonable, I guess, I don't know. Most of that's equity-based. But I'm not looking at that and thinking, “Oh, clearly there's maleficence at work.” But it does raise questions to the point where if your CEO owns a private plane and then sells it back to the company, maybe you don't use that particular plane for private travel. Maybe you go with something else, just to keep the ethical issues clear. Am I just naive on that?



Pete: I would agree. I mean, again, it's not a money thing, it's a time thing. I mean, why would you fly private? Why would you charter? Because if you're the CEO of a company, that's probably going to be valued at tens of billions of dollars, your time, super valuable. 



You should get to where you're going as quick as possible. And so, you're right though. Does it look questionable when stuff like this happens? I mean, it's buried in there. I'm sure someone might say something like we do but doesn't really change anything? It's nothing like what was seen in some of the other S-1s with kind of strange monetary deals between the WeWork stuff, for example.



Corey: Yeah, the weird disclaimers on a lot of their compensation is they wound up having a specific line item that their president slash head of product received specific compensation for certain travel expenses of his spouse in connection with a company-sponsored event. And I look at that, oh okay, how much money are they talking? The total other compensation line that's referencing is 887 bucks last year, and this guy makes $2.2 million in terms of total comp as the president of products. It's on some level, just eat the expense yourself personally, to avoid the appearance of conflict of interest because we're talking about this now, we're probably not the only people going to highlight this. And it’s just, is the grief worth it?



Pete: Yeah, exactly. It seems silly in a lot of ways, but I'm sure at the time, it was just like, “Oh, yeah, I put my expense report in and got it taken care of,” or whatever. I'm sure it was just a line item, but of course, like, you got a lot of lawyers that are going through a lot of details to create these S-1s to make sure that they disclose everything. I mean, to your point as well, always trying to find interesting things, is that the son of the president of products, who is also a member of the board of directors, but the son was hired as an account executive, that's like entry-level salesperson, cold-calling person. And I feel really bad for his son because they had to list the per year compensation that this person received while employed there, which, sure, you're a member of a family that is going to be entering into a large amount of generational wealth.



Corey: Yeah, in 2019, he made just under $65,000, total cash comp and commissions. Last year, because the company is taken off, he made the princely sum of $126,000. Which, again, I'm not saying that that's nothing money. Far from it, I get it. But by the same token, that's not embezzlement money. 



That is the kind of money, you pay a sales executive who's hitting their plan, presumably, for this stuff. In fact, many SaaS salespeople make significantly more than that. And it's clear from those numbers at least that this guy isn't having a whole lot of favors done for him, but I guarantee you you're right, everyone is going to assume, “Oh, it's because your dad is the president of products that you have this job. You're probably making five times more than we are.” It's got to be rough.



Pete: Yeah, I think between that and just like, I've never had my salaries published publicly so that two idiots on a podcast can talk about it, before. That in itself—



Corey: Yeah, right?



Pete: That can't feel very good.



Corey: Oh, yeah. If I wind up disclaiming how much I make in any given year, there's no good answer here. If it's low, it’s, “Wow, he's not making any money at all.” And if it's high, it’s, “Wow, there's good money in that making fun of Amazon business.” There's no upside; there's a lot of downside and I do feel for it, especially since this guy's very clearly relatively new to his career, and this is kind of going to, I guess, cast a shadow to some extent. I feel for the guy, I really do.



Pete: Yeah, yeah, exactly, exactly. But the last thing that I wanted to call out in this one is because we get these startups with this insane growth, you really want to hope that there's people the quote-unquote “normal people” that have been there, just grinding it out day in and day out over a long period of time, you want them to get their win, right? They worked hard, and hopefully they get their shot at getting a slice of this huge success. And so what's cool is they actually listed out all of the various tender offers, secondary offerings, price per share over a period of time. Now granted, they'll only list who sold shares who was director within, so like, they're not going to tell you if Susie from accounting sold some of their shares or whatever, but it does list the share value of those people who sold. 



And honestly, I think the only interesting thing is that you're seeing a company whose share price was continually growing at each valuation round, which then makes a lot more sense why they raised so much money if the money was there and the valuation was good. But the other reason I like to bring this up is because this is something that not a lot of people who go to a startup realize, is that there are exit ramps for people who are higher up who are founders that exist, hopefully to the whole business, but oftentimes, it's just to those early founders, those early people. I mean, for example, even though the former CEO, Bob Muglia, is no longer there, in the February and March timeframe this year, he sold about $80 million worth of that stock. Is it better than what's going to IPO for? Who knows, but what I do know is that he got a check for $80 million earlier this year that he can do whatever he wants with it. Are other people internally going to have that kind of win? Probably not. But it's one of those extra things—



Corey: He also left and was either forced out or resigned, it's unclear. Part of the story on some level is it's not often considered appropriate or good optics for a former executive who's no longer there to hold an outsized proportion of the company's value. So, it may very well have been a negotiated drawdown of stake.



Pete: Yeah, I think you're exactly right there, which is he doesn't have a management interest in the company anymore. So, why would you have so much of your net worth, I guess, tied up into there.



Corey: We saw that in some respects with the Bezos divorce, where it was very clear that Jeff Bezos’s ex-wife was going to get a fair chunk of Amazon stock, but she didn't have a role there, so how her being effectively a massive minority shareholder with voting rights was an issue of some question. Is this going to change the way that Amazon is run? In practice, no. And there's a whole story around that, but I don't think that there's anything necessarily egregious going on as far as the previous exec selling out. But I bet it does rankle a lot of the employees who, “Wait, the guy that just left gets the cash out $80 million dollars and we're still being told that the future is bright, but we have no actual dollar figures coming in?” I don't know.



Pete: Yeah, exactly. You know… pays to be a CEO of a billion-dollar company. [laughs].



Corey: It sure does. Thank you so much once again for joining me on this one, Pete. In our next episode, we'll probably tear apart Palantir S-1.



If you've enjoyed this podcast, please leave a five-star review on Apple Podcasts, whereas if you hated it, please leave a five-star review anyway on Apple Podcasts, and thank you for listening.



Announcer: This has been a HumblePod production. Stay humble.



Newsletter Footer

Get the Newsletter

Reach over 30,000 discerning engineers, managers, enthusiasts who actually care about the state of Amazon’s cloud ecosystems.

"*" indicates required fields

This field is for validation purposes and should be left unchanged.
Sponsor Icon Footer

Sponsor an Episode

Get your message in front of people who care enough to keep current about the cloud phenomenon and its business impacts.