Snowflake Inc. (SNOW) Earnings Call Transcript & Summary

Welcome to the Piper Sandler fireside discussion with Snowflake computing. My name is Brent Bracelin. I spearhead our research efforts in cloud apps and analytics. After this 25-minute discussion, we hope you'll have a better appreciation for Snowflake as one of the leading digital disruptors across our coverage universe. Joining me is Mike Scarpelli, CFO. Mike, welcome.

Good morning. Thanks for having me, Brent.

Absolutely. So let's start with really the vision around the business. Snowflake has quickly crossed over $1 billion revenue run rate, I think, last quarter. That's while sustaining triple-digit growth. Certainly, people appreciate the momentum, market cap in excess of $100 billion. So in your biased view as CFO of Snowflake, what makes Snowflake so special?

We really benefit from a number of underlying demand drivers today. Organizations are really prioritizing their data strategy as they recognize the importance of using data to drive decision-making, that is a key thing. And while data volumes are exploding, you hear about it all the time, the cloud provides the ability to actually harness and utilize all that data. And that's probably one of the biggest beneficiaries we have that make us so special. And both cloud-native and large legacy enterprise organizations fuel our durable growth trajectory. For many early cloud adopters, Snowflake consumption expands as their business grows. We have many, many, many of our companies. You can see their consumption on Snowflake correlates directly with their growth. For the larger enterprise accounts, at the outset of their cloud migration, we see the potential to unlock additional workloads and use cases. And I really don't see any of our customers that I would say are saturated. There's still such a long runway in front of us with our existing customers, let alone all the new ones we're talking to. So I think that's what makes it, and obviously, the technology and everything. But I think we're in the right place at the right time as well with people getting more comfortable with the public cloud, I will say. When we go into a -- especially a large Fortune 500 or Global 2000, the #1 concern they have, if they haven't already dipped their water -- or their toe into the cloud is security. And even if they are a company that has workloads in the cloud, they are still super focused on security. And that's one of the -- that's what really drags out the sales cycle for us a lot, is that whole security review.

Interesting. I think one of the points you made there around some of your cloud-native customers where their growth is correlated with data consumption. I think that just really exemplifies how the fact you're seeing just more and more of these data-driven business models, where their apps and the monetization is really directly tied to the data that they have and they're creating. So that certainly is interesting. Let's talk a little bit about that enterprise space. Clearly understand why some of the new digital cloud natives would turn to Snowflake. Enterprise side, as you well know, you've been selling enterprise for the better part of the last 20 years. It can be slow. It can be slower. But we're starting to see enterprises like Ansys, Baxter, Kimberly-Clark, Labcorp, T-Mobile, Williams-Sonoma, they're all starting to replatform on Snowflake, and it's for a variety of use cases, data warehousing, data lakes, some modern analytics stacks. What's resonating with enterprises as you think about their willingness to migrate off legacy?

Yes. I think the #1 thing that's motivating them to migrate to the cloud is really the cost benefit of elastic and scalable resources that the cloud offers. And this is a key thing with Snowflake. It's reduced complexity of a managed service. We -- it's self-managing, customer doesn't have to do anything. Talk to customers, it just works. And elimination of data silos that is the key thing. Because in an on-premise environment, the infrastructure really dictates data operations, you're limited by the size you have, and so it creates many data silos. With Snowflake, the inverse is true, the business rationale drives data operations and the unlimited compute and storage in the public cloud in one place, you have all your data in one place, you eliminate those silos, which is great for an organization because you can have people cross functional in different divisions, departments accessing that exact same data. And the nice thing about Snowflake and this is what resonates so well with our customers is the governance and security around that. You know who's looking at what, you give access to who you want. If anyone is doing anything with that data, you know exactly who's doing what.

That's interesting. And one of that, I think, misnomers out there is this idea that if you're in the cloud, it's easy to use and self-managed, but little surprised to hear customer feedback around Redshift, right, offering from AWS that still needs a lot of manual tuning. So how much of that is a differentiator. If you think about that self-tuning capabilities of Snowflake, is that one of the many reasons why customer -- enterprises are kind of converting over? Or are there other factors, like you said, price is, [ and the cost benefit... ]

Yes. It's not one thing, it's a number of things. I can tell you what is resonating in early sales cycles with customers the most is the ability to data share. It doesn't mean they're going to do that day 1. They'll have visions of data sharing. They all love the idea that you can be multi-cloud. You can move from one cloud to another if you choose to do that, you could have your primary in one and you could have -- you could do DR and Snowflake into another cloud. So they like that idea. They're not getting locked into one of the 3 hyperscalers out there. I'm not going to say 4. I'm going to say 3 because we really don't see that other one that was talked about yesterday.

We won't go there. But it is interesting looking at their growth profile that...

And then with Redshift, the big thing is scalability. It just can't scale like Snowflake and it does it with concurrency. They just have architectural issues. We never see Redshift in a large Teradata migration. It's just not there. When it's Google, we see the most in Microsoft competing for those workloads.

Let's pivot to some of the momentum in the Fortune 500. You added what, 25 net new Fortune 500 logos last quarter. I think you're now at, at least a logo penetration of 212 of the 500. But obviously, just scratching the surface of share gain opportunity at those. Can you compare, contrast what are those lands with Fortune 500s look like today versus maybe 2 years ago?

Yes. I'm going to start with answering -- sharing some other data first. We started out disclosing Fortune 500 because we're really U.S.-centric and it was much easier. I don't want to get into all the details around it, but we didn't have the details around the Global 2000. I would say we're focused more on large enterprises. The Fortune 500 is really a U.S.-centric term. And so we're going to continue to disclose the Fortune 500, but it's really the Global 2000, which has its limitations too because it doesn't deal with large private companies. It's the 2,000 largest public companies in the world. That also doesn't deal with public sector. So I just want to get that point across because it's not just -- it's large enterprises organizations that we really go after, and we have a number of those. In Q2, we added 18 Fortune 500 customers. I want to say these are long sales cycles. They're going to be lumpy. You saw that in Q1 people were picky about it, and they kept saying it's not a problem. These are lumpy deals. There -- these -- some of these sales cycles are 1 to 2 years. They don't happen overnight. And a lot of it is because companies want to get through their security review when you're dealing with these Fortune 500. And then they also wanted -- they want to -- none of these people are going to procure Snowflake until they're ready to do that migration. And there's a lot of competing priorities and companies. I'm dealing with this today. I had a vendor that came to me and then one of my executive sponsorship on something, and I said, "you know we really love your technology, but I'm going to tell you it's not going to happen in September." And the reason it's not going to -- we're going to do it, but I won't let my people buy anything until we're ready to do the deployment and start the work. We can talk and have everything set, but we won't sign a -- and that's no different than any large company. It's irresponsible to sign a contract until you're ready to start to use it, and -- where you start -- where you have to pay money upfront. So the vast majority of our customer wins with these Fortune 500 what's driving it, and we'll get back to that, and large enterprise, is it's a kind of replacement of either a cloud or on-premise solution. The replacement of a cloud, think of a Redshift or something else, those tend to happen faster. The on-prem Teradata, the Netezza, the Hadoop and others take more time to do. The -- and I would say the larger the enterprise, it's more likely going to include some type of on-premise replacement that's going to happen. Because typically, these large companies have been around for years, and they have some type of legacy system. The newer companies, those guys are typically replacing some type of cloud solution. I would say the major shift that we've really observed over the last few years is the attitude towards the cloud adoption for these large companies are getting more comfortable with it. And most companies, if you talk to any of them, all have an intent on moving to the cloud. It's just a matter of when. So I think that's the -- hopefully, I answered your question.

Absolutely. We see that as well too kind of post-COVID, right, that there definitely feels like particularly in back office, middle office, we've seen that acceleration, right? We've seen multiple vendors really see accelerating growth in just the last 30 to 60 days, where it feels like that appetite to move to cloud is definitely changing in the enterprise. I guess for me...

Yes. I would say with that, too, the other thing that's important too, it used to be us convincing customers, you need to move to the cloud. Where we spend most of our time with customers now is not convincing them that they need to move to the cloud. It's really helping them. They want help with how do I move to the cloud. And that's why we're investing so much in our professional service organization. We're really spending a lot of time in training our SEs around that to help our customers. That's where the discussions are mainly now in sales processes. It's not about the benefits of moving to the cloud, it's how do I do it?

If you think about those 2 use cases, it's interesting. One use case, cloud replacement, the other use case more of an on-prem replacement scenario. As you think about those cloud replacement scenarios, as you think about more enterprises willing to kind of move to cloud, do they try the cloud version of a data warehouse first and then they kind of move over or walk me through kind of the...

Most of these people are doing POCs. When we will -- if we're doing an on-prem migration, we will do a POC against a Redshift, a Azure, a BigQuery. All of these companies do POCs, and they're testing for performance, and they will do small, medium, large data warehouses and testing processing speed and other things. But technically, we win virtually all the time. I don't want to say all the time because there are certain types of data that may be more akin for BigQuery or something, but the vast majority at a technical level we win in those things. So many of the times, we don't even spend that much time anymore on the technical side of Snowflake. We tend to spend more time and this is what we're evolving our salespeople to sell, what is the business value Snowflake will deliver because many of the times today, we're not talking to the CIO. We're not alienating going around them, but we are really talking to the lines of business and telling this is what Snowflake can do for you.

Interesting. So how does that correlate to average deal sizes as you think about kind of the lands today versus kind of maybe a year ago?

The average deal size has remained pretty consistent in the initial, what we call Cap 1, capacity 1 purchase typically is around $100,000. And -- but you have to understand we add a lot of small companies that will sign up at $40,000, $20,000 and others to start. It's unheard of that you would do a Cap 1 with a Fortune 500 starting out at $2 million a year. I'm not going to say it can't happen. But generally, they're sub-$1 million. And then when you average it out with all those other smaller ones, we get to $100,000. But what happens though is once they buy, on their first renewal, they end up buying more, in the second one they do -- I'll give you an example. There's 1 Fortune 500 I was just dealing with yesterday on internally. They started out at $300,000 last year. They consumed close to $1 million, but not enough. They did a number of purchases throughout the year. And now I'm talking with a $6.8 million 3-year contract with them. Whether we'll do that, they're still trying to decide do they just buy annually or they go. But I'm just showing that's the type of growth, and there's still a lot more opportunity in that account.

So do you think that's a good anecdote for what you typically see that starting out with a small land, it might be 2 or 3x consumption in year 1? Or is that maybe a more atypical...

Typically, a customer will consume their first amount in year 1 and could easily be 2 plus x year 2 and grow. But it really, really varies. I can give you another example; 2 years ago we sold a contract to a large retailer in Europe. We did a $5 million deal. It took them 9 months to really do their migration, 9-plus months. In the fourth quarter of that deal, they only consumed -- they consumed close to $1 million in that year where they did 5. They still renewed, and so they had 9. This past year, they were consuming close to $8 million. It would not surprise me if they're close to $20 million next year with that specific customer use case. And by the way, that was a -- almost a 2-year sales cycle to get into that account. That's why I'm just telling these are long -- it takes a long time before you get to -- and this is revenue I'm talking about.

Those are some big numbers. But certainly...

But it varies by customer. It really varies so much by customer.

Yes. So on that example, this idea you can have a relatively small land, 2-year process, they get to $8 million and have a line of shot to $20 million. Where are we at in penetration rate? As you think about the enterprise market, the Fortune 500, I know we can calculate a logo penetration, but it doesn't really tell us much, right? So what do you think penetration rate maybe with just the existing installed bases today, best guess?

Yes. We really don't have a clear sense of penetration rate within our customer accounts because there's so many use cases we just don't know. And customers are finding these use cases. Many of our largest and most mature customers are still growing. And so it's really, really hard to say that. What I would say is I can't find a single customer that I could stand up and say, unequivocally, they're fully penetrated. It doesn't mean all those customers are going to grow. It may mean they want to split some of the work with other things, but there is more opportunity in these customer accounts, especially with data sharing. And as we get more and more partners to be putting data into our marketplace where these people can access. At our Analyst Day, we provided some insight. We have 116 customers with trailing $1 million product revenue run rate. We expect that number to grow to over 1,400 by 2029. So clearly going to grow a lot. And I will tell you, we have a lot of customers that are at $1 million run rate that's not reflected in that 116 right now because it takes time to ramp them. We do expect that spend per million dollar customer. If you look at the average, it's going to increase. It's $3.4 million today. I think it's going to get to we're forecasting $5.5 million in 2029. So that just tells you the growth we think we're going to see on average across our big customers. And we do believe that large organizations, I don't want to say Fortune 500 because there's a lot of other companies internationally are our greatest opportunity to reach that -- to reach these targets. And that's where we're focused on those large customers. But I want to stress, the Fortune 500 only account for about 25% of those million dollar plus customers. Those smaller customers can be meaningful customers. We have a lot of customers with sub -- a couple thousand employees that are driving considerable revenue consumption on Snowflake. We have some with sub-500 employees that are driving a lot of consumption on Snowflake. So don't get too wrapped up in the...

Just enterprise. Yes.

Yes. We are very focused on it, but it's not just that. We're also focused on the lower end of the market. And that's the one thing that really impresses me in my entire career. I've never seen a product that can scale up or scale down. It's the exact same product whether you have 20 employees or you got 200,000 employees. It's the same product that you're using. And the beautiful thing about our consumption model is we can offer it to everyone. You don't have to do a $20 million deal or a $5 million. You can do on demand, and that's the exact same product you're getting. It all comes down to what type of pricing you're going to get though.

So you have a really good product, has lots of applicability technically kind of winning in POCs against the big cloud vendors. But it does take time to consume. And so we're all trying to figure out, okay, well, should we look at RPO metric, should we look at consumption metrics? What's the right metric, right, to look at here. So maybe just talk about that consumption, RPO kind of connection here? How long is the lag from a contract signing to consumption as we think about trying to think about RPO as a leading indicator for kind of the momentum of the business.

Yes. Well, it all depends upon the contract and I do think RPO, and current RPO, more importantly, is important, along with revenue. And the reason revenue is the best indicator of the future because that's what customers are consuming. And once you consume, you don't really decrease your consumption. Yes, you could run certain things. But -- so we give the metric that we expect that 60 -- or sorry, 56% of our total RPO will be recognized over the next 12 months. That was up from, what was it, 54%, 55%. I think it was $85 million increase in current RPO quarter-over-quarter. That is what is a good metric to look at in conjunction with revenue because you got to remember, if I end up beating my revenue in the current quarter, say, by, I'm just throwing numbers out, $10 million, that took out of my current -- my current RPO that I'm reporting at the end of that quarter as well, too. So it's important to look at both and how those are both growing over time.

One thing about the data space is it's in constant change, constant innovation and constant VCs putting money to work. So we're going to go into a little bit of a lightning round on competition. Friend or foe. Let's start with Databricks. Friend or foe?

Many mutual customers rely on Snowflake as their data platform and use Databricks for data engineering workloads. We really don't see Databricks compete as a comprehensive data platform. So it really appears that Databricks sees more of a threat from us.

Kind of different part of the stack. Confluent, friend or foe?

Correct. Correct. Sorry, what was that?

Confluent.

Confluent, they're partner. We don't directly compete with Cloudera. We replaced a lot of many -- we're constantly replacing Hadoop. So I don't -- they're definitely not a friend, but they're not a foe because we don't see them as competition, one that...

You got Alteryx, I get questions of Alteryx...

They're a partner, a good partner.

Okay. Great. Very cool. Last couple of questions here for me. As we think about futures, Snowpark venture is -- obviously is putting some money to work. And as we think about the appetite for Snowflake, you kind of own that kind of data platform, data sharing kind of concept. Any appetite to move up the stack into applications over time. Walk us through kind of the vision of if you were to expand, build, buy, where do you have appetite going?

We want to be extremely partner-friendly. We want to be an open platform. We really don't have aspirations to build applications today. You can -- there's other companies out there, when I talk to them why people want to partner with us is because they see those other companies as being more competitive with them. Our goal is really to provide the infrastructure that is applicable to a variety of use cases and verticals. We're going to continue to invest in tools and features that make no use cases possible. We're really focused on making it easy for our partners to build applications on Snowflake. We're adding some observability functionality to empower a number of observability and security vendors on the Snowflake platform, not to market a single proprietary solution. We really don't want to do that. That's not our business.

Good. Very clear. Two last questions for me as we wrap up here, Mike. One, Snowflake was one of the early companies to talk about moving to flexible work environment. How is corporate life in Bozeman and is it helpful attracting and retaining employees, one. And then two, I'd love to hear what you're most excited about thinking about next year.

The work of the future is definitely changing. We're not going to go back to the way it was. When you talk about Bozeman, I want to -- we're a global company. We have people all over the world. They have very few people in Bozeman today, but we haven't really marketed Bozeman to hire a lot of people yet because our office is not ready, and we're confined in space right now. That it will be a market that we will go after people. And it's a market that a lot of people want to relocate to as well too we're hearing. So we will return to the office at some point in time, but it's not going to be the way it was in the past that you're going in 5 days a week just because that's your work home. It's going to be a place where you go with purpose. I do think technologies like Zoom and stuff are here to stay. I know -- I'm a huge believer in virtual meetings. They're the most efficient thing. It really is going to extend my work life if I can continue to do these, and I know a number of other people as well, too. And so I'm excited about what the future is going to look like from a work perspective there.

And next year, what are you most excited about thinking about Snowflake in 2022?

I'm really excited about a lot of the -- the impact of a lot of the announcements we had this year, Snowpark powered by Snowflake. We're seeing some of the customer relationships that we've signed up this year and where we see them going, data sharing. With Snowpark, we did announce that that's really enabling application developers to harness the data and Snowflake to write applications. We did release a public preview on AWS support with Java and Scala that's going to go into GA in Q4. We have Scala coming out -- or sorry, Python coming out, which is going to be a real key thing because it seems like that's the most common language everything is moving to. So we're just excited about what we're seeing within our customers and how they're growing. I also think it's going to be -- it's interesting to see what the impact of the support for unstructured data is going to look like next year for us, which is going to come out later in the year. And then also once we get FedRAMP high, that there's a lot of opportunities within the federal space. So I think we're in a great space, and I'm pretty excited just to continue to do what we've been doing.

Sounds great. Well, Mike, thank you so much for taking time today and sharing us a little thoughts on the business.

Thanks for having me.