Snowflake Inc. (SNOW) Earnings Call Transcript & Summary

Good morning, everyone. My name is Brent Thill, and I will be one of your hosts this week. Thanks again for your support of the Jefferies conference. And we're here for you. If you need anything, reach out to us. I'm really happy to kick off with Jimmy at Snowflake. Katherine is here, too, in the front row. I asked her on stage, but she said she's going to sit in front row. But Jimmy has been with Snow for 6 years and was with ServiceNow for 7 prior to that, tremendous experience in the industry.

And Jimmy, coming off a really good quarter. Maybe we just start there and your thoughts about what surprised you, what maybe wasn't a surprise. Give us a sense of...

Yes. Yes, yes. So look, it was great to start the year in a position of strength. I think everyone remembers the last 12 to 15 months with the Sridhar transition and all of the changes that we've been implementing as a team. And I think you really saw us hit our stride in Q3, and being able to continue that momentum through Q4 was great. If we look at the actual drivers of outperformance, it was quite broad-based. It wasn't concentrated in an individual customer. We called out retail and technology being segments of strength. I think we're hitting our stride in EMEA. And then as it relates to just different kind of hyperscalers, you're seeing nice growth on Azure for us. But look, it was pretty steady throughout the quarter. And I think there was a moment there during the quarter, especially when we were on the road talking with investors where we're kind of waiting to see some disruption from some of the macro noise, and we really couldn't identify anything driving positively or negatively in our business. So I would say it was just a really smooth quarter on the consumption side. On the bookings side, I know we'll talk about it in a little bit, but we saw some large deals and 2 north of $100 million deals in the financial services sector. And we gave the call out in Q4 that some customers signed shorter duration deals, and we were expecting those to close in Q1, and so it was great to see those close. So I would say really just strength on the bookings and consumption side across the board, which is great to see.

Anything surprising you that you weren't expecting?

I think intra-period, we were waiting to see any changes in trends. I think if you go back a few years when there was some macro noise, everyone remembers, we did see some negative impact as customers were kind of tightening their belts. I think for us, what was positive intra-period was looking at the profile of our customers kind of skewing more large enterprise, and the growth curve for those customers is much smoother than our customer cohort from a few years ago. And so I wouldn't say that it was a positive surprise, but I think we were kind of waiting and seeing -- and as you know, we look at our consumption patterns on a per customer basis on a daily basis. And so I'd say that was positive. I think if you look at our free cash flow margin in the quarter coming in at 20%, that was where we had expected that to come in. I think that wasn't viewed as a negative from our perspective. But if you look relative to history, externally looking in, you'd probably think that could be viewed as a negative, but that was really the duration impact of the Q4 bookings that didn't close. Those would have been February collections. And so now you'll see our free cash flow be more back half seasonal, which was expected, but relative to history, was a little bit lighter on the margin front.

One of the comments you made to me just was more around diversity. Last time we went through an economic down cycle, it was tech heavy. And now you've got diversity in your base, and we can't see that diversity, but maybe describe that what you're seeing in terms of this customer base.

Yes. So I think the digital native cohort was a popular term a few years ago. I think like the food delivery companies or the cryptocurrency companies, those represented meaningful growth for us in calendar 2021. We accelerated growth sequentially in Q3 of calendar 2021, driven by a lot of those customers. They were all in the top 10 of our customers. Now if you look at our top 10 customers, they skew much more Global 2000, think large banks, large telcos, large health care companies. And their growth, albeit, I would say, slower in their earlier years, if you just look at the cohort, they have a much longer tail. And so like the terminal value of those customers to us is much higher. And they have much more operational rigor on purchasing and budget than I say some of those other companies did. And so we benefited from the massive growth of those companies a few years ago, but then they were the first to optimize because they were clearly growing inefficiently. So you can, we believe, make the argument that the large North American banks are scrutinizing budget every single period, not letting that get away from themselves. And so it's much more durable in our opinion.

Back to the two elephants, as we call them, the $200 million deals. Tell us, what does a customer do with $100 million worth of Snowflake?

Yes. So traditionally, analytics are our bread and butter. Those two companies are in the financial services sector, massive on-prem, think data warehouse estates and been on the platform for a few years. And it's really just a matter of time. And so kind of the Snowflake journey with these types of companies is really start with one on-prem data estate and then slowly tick those down because you have the ability to take smaller bites of the apple upfront. And so that follows with massive PS engagements over a multiyear period. And so it's really the traditional Snowflake data warehouse migration that's driving those. And customers will just slowly purchase those over time because they are really cumbersome projects. And so you've heard Sridhar talk a lot about a huge focus for us and for him is making those migrations more seamless, especially through the use of technology like AI. And so if we can make those easier, we think that should accelerate those migrations, which would be more revenue for us sooner.

You mentioned during the quarter, you were looking for the static and you didn't really -- you heard it, but you didn't see it. And so everyone is asking us, what is inherently in your forecast and your view? Are you embedding any static? Are you saying, "Hey, we're not seeing this, we're not forecasting in our numbers." How do you...

Yes. Same way we forecasted looking at usage patterns we've seen to date. And so whether or not it is impacting our customer base, I'd say we can't identify anything that's driving our customer usage, positive or negative from a macro behavior, but we're just still rolling forward what we're seeing from the customer base today. So there isn't some sort of plug for, I would say, like macro, positive or negative.

And maybe go back kind of 40,000-foot view. I mean, one of -- this is our thesis. When Sridhar came in, it seemed like you had nailed the go-to-market, maybe some of the product things needed to be tweaked. And it seems like -- he seems -- he has both skills, go-to-market and product, but everyone viewed him as being more of a product individual and kind of shoring up some of those capabilities. So many ask us, where do you think you're at now with Sridhar's arrival to now where you're on in some of the technical aspects of where you want to be?

Yes. I think he would like call it continuous improvement. And so interestingly, last year, when he came in, the belief was that he was going to spend a little time on go-to-market, a lot of time on product. I would say that if he was sitting here, he'd tell you he spent most of his time on go-to-market last year. I think the beautiful thing about Sridhar is that he doesn't have any religion on enterprise go-to-market because he had never managed it before. And so this idea around accountability at a very granular level is something that he has brought to the table, which is measuring activity and productivity on a much shorter duration, like a weekly basis, not a quarterly basis. And so it kind of up-levels the expectations that every single employee has at the company regardless of what group they're in. Like Katherine and I joke that he spends every day talking to us, to marketing, to sales, to product. And so he's really dialed in across the entire organization. So I would say on the go-to-market front, we feel good. Mike Gannon, obviously, coming on board was a change for us with Chris Degnan departing the business, and I think he'll bring another level of rigor. And then on the product side, I think the elevation of Vivek to run engineering, working closely with Christian and then rolling out leads for each of our different workloads. I think we have the structure in place that will scale us to the next level. And I think a lot of that is driven by Sridhar's identifying places where we could have improved.

There's a lot of new products that he has been focused on. I guess if you laid out a couple of the ones that you're most excited about?

Yes. Yes. So the four buckets we think of in terms of our product set is data engineering, analytics, AI/ML and then apps and collaboration. You'll hear Christian and Sridhar talk about this next week at Summit. And within those, there's a number of different features. The Cortex suite of products rolls up under AI/ML. We're super excited about that, going from basically 0 customers 15 months ago to over 5,200 customers using that on a weekly basis is really exciting. Modest contribution from revenue, but ultimately, we believe adoption is the leading indicator for future revenue. I think the investment that we made a few years ago in fully supporting Iceberg was a big shift for us culturally. That not being a net headwind and actually being a net tailwind to date since last June going GA is super positive for us and allows customers to have choice. But then I think more tactically, data engineering is the largest revenue contributor outside of core analytics. So think like moving upstream in the data management life cycle around the data transformation. So that would be the features that you've heard us talk about like Snowpark and Dynamic Tables. The contribution metric we gave in Q3 was $200 million run rate, obviously outpacing the business, and we'll disclose more of those metrics on a milestone basis. But we want to simplify the way that we're talking about the products with investors because I realize when we're releasing 125 features a quarter, it's hard to keep up.

You mentioned the conference next week, which is June 2nd through the 5th, and Sam Altman is on the agenda.

Yes.

So everyone would love a teaser.

I don't know what Sam is going to say. I'm sure he and Sridhar will talk about very forward-leaning AI use cases. OpenAI is a great partner of ours. And I think they believe data is at the center of a lot of positive outcomes for AI. And so we're well positioned. But I don't have the inside track to what Sam is going to say.

Okay. No teasers. I thought I'd try. Many ask us, is Snowflake an AI story, more of a traditional data story?

Yes. Yes. I think majority of our business is traditional data. We're not standing still on that front either. And so we're investing and still delivering a market-leading product in core analytics. But if you're a data company, in order to be relevant longer term, you have to have a product road map that supports next-generation use cases, whether or not enterprises are willing to adopt them right now. And so when we think about customer conversations, what are they going to do with Snowflake 5, 10 years out, you need to be able to speak to those use cases and be ready for when customers want to use those features. And so I would say in terms of enterprise AI, we believe we're best-in-class. I mean the Cortex family of products, being able to access market-leading frontier models natively in the service without data leaving the service is incredibly attractive. And I think the adoption would support that of those use cases. But in terms of revenue contribution, no surprise, majority of that is coming from the analytics and data engineering.

And when do you think AI would be more meaningful to the business? Is it ' 26, '27? How do you...

Yes. I mean, look, we guided to 25% growth this year to, call it, $4.3 billion of product revenue, majority of that being core. When the core business is growing that quickly -- what does material mean? So it's hard to exactly say. You saw Snowpark and Dynamic Tables and data engineering get up to a couple of hundred million dollar run rate as of last Q3. I don't view that number as material as it should be. So I think I'm not going to make any predictions on next year, but I know that we have high aspirations for it to become material.

Just on $4 billion plus of rev, everyone says single-digit margin? Come on. It's a little pitchy. But what I say is you've got the best margin upside of anyone in large cap because no one -- you put a lot of investments in the last year.

Yes.

Do you believe that there's now a natural floor set where hopefully, we don't go back to the low, but...

Yes. I think you've heard Mike and Sridhar talk a lot about the commitment to showing leverage. Last year, you recall taking, what, a few points off the product gross margin line to invest in a lot of the AI infrastructure. And then subsequently seeing operating margin down last year was investing ahead of the curve. The 8% guide for this year, we feel very comfortable with. And the levers where we can get more margin expansion will come from, I would say, one, R&D. If you think about how we've invested in R&D historically, we've really built out a strong management layer. So I would say, very talented, expensive employees from the likes of Google and Meta at the more managerial level. If you look at where we're hiring now on the R&D side, it's much more early career, so less expensive and kind of supporting that management layer, and that's a really, really big focus for Vivek and Christian. And then our own internal use case for AI and how we're using some of these productivity tools, Sridhar said it publicly many times that he thinks we'll get a lot of leverage in that line. And then beyond that, the revenue growth supports, I would say, attractive operating income that will flow directly to the bottom line. So absolutely, we feel comfortable with the ability to show operating margin leverage.

We get the question -- and maybe I'll spend more time in the next week talking through this, but the growth and margin framework, is there an easy way we should think about this over the next 3 years and how you think of it?

Yes. So growth comes first, but we're committed to efficient growth. And so if there's opportunity, we are going to go after it. And I think that starts with sales and marketing kind of accountability, and you'll hear Mike Gannon talk about this. Productivity has been -- trended well for us over the last year, and you saw us bring on a lot of sales and marketing heads in the first quarter of this last reported period. And that's as a result of us seeing strong productivity and believing there's a massive opportunity to go after. And as we see these newer features contribute over time, we'll continue to invest more in those resources. But I think that we can do that while still showing operating margin expansion, which is how we think about kind of the medium- to long-term profile.

Just on that point on the sales team, Mike Gannon just joined as CRO in March. Everyone -- any time you bring a new CRO and everyone on our side goes, "Wait, what's he going to do to the sales team? Is he going to tweak it too much?" Are we going to have a quarter off where things go offline for a moment and then comes back online? And how do you think about what Mike is going to do?

So he's spent the last couple of months -- I guess, he joined 2 months ago -- meeting with a lot of customers, so kind of listening and hearing, and I think he's really happy with what he's hearing. In terms of the actual structure, he has a really strong background in focusing on the GSIs. I think we believe we can do a lot better on that front in having joint alignment and goals. And we realize that to get into the largest enterprises and be meeting for the largest enterprises, you need to have a stronger presence not only like with us with them, but also with their investment in their Snowflake practices. So he's highlighted that. You'll hear him talk next week a lot about kind of owning the entire customer journey end-to-end from sourcing, all the way to the customer support. And so I think, look, his -- the difference between he and Chris is Chris got us to where we are, and Chris knew every deal, very detail-oriented, kind of just through the ranks of growing with Snowflake, that was his personality. Mike is bringing much more of an operational focus to how you enable, how you structure different go-to-market territories and then how you partner in the field. And so we're looking forward to that, and I think people will be very impressed when they hear from him next week.

Just speaking of partnerships, you talked about the Microsoft partnership. You're working with others. Just give us a sense of what's happening, what's new?

Yes. On Azure, I would say the partnership -- Christian, Mike and Sridhar said this. It's never been better. We're working with them on making that relationship even closer. I think everyone understands that if you want Microsoft to play nicer with you, their reps need to get fairly compensated for selling your product, whether that's through their marketplace or direct. And so we're working on that. I think us becoming a larger customer of theirs with bigger commits will help that over time as a larger percentage of our business is on Azure. But then we've announced the integrations with Cortex and accessing SharePoint and Teams and all of those features definitely benefit the relationship longer term. AWS is still 70-plus percent of Snowflake's business. Azure is the fastest growing at low 20s, and then the other balance being GCP at 3%. And so I think we view Azure as a very important partner for us longer term.

There's a new kid on the block called Salesforce, and they have this little thing called Data Cloud. Everyone asked us, it was interesting because Databricks said less about you and said, well, we're really going after Salesforce because they're trying to take our workloads and they're easy to fight. Like, we can beat them. What do you think about what Salesforce is doing with Data Cloud?

Yes. We're a close partner of Salesforce. We're also a close partner of Informatica. Ultimately, customers choose where data goes. And so fortunately, for us, Snowflake is a destination that customers want to put their data because they want to combine that data with all sorts of other data, whether that's Workday data or ServiceNow data or AWS data. And so that's not a trend that we see changing. But clearly, Salesforce has big aspirations with the Data Cloud, and we hear a lot about it. In practice, in the field, it's not something that we see competitively when we look at our competitive opportunities. It's still very traditional for us. But I completely understand why Salesforce is doing what they're doing.

Yes. Any questions from the audience? We can take a few from the room. I know there's -- Fred's got to have a question. I can keep going. Yes, sir.

Expand the Databricks-Snowflake back-and-forth?

Yes. Databricks was a partner for a very long time. They played more upstream in the ETL space before data landed in Snowflake. Both of us have been encroaching on each other's territory. They started on the left side of the data management life cycle. We started on the right. And so we released our Snowpark product a few years ago. They released their SQL product a few years ago. And so we are now, I would say, going after a lot of the same business in our existing accounts. We have a ton of customer overlap. In practice, if you look at our largest customers and their largest customers, we coexist, and that's clear through the growth rates in each of those accounts for both of us. But longer term, we're absolutely becoming more competitive. I think our key differentiators are the fact that we're highly performant and very easy to use. And so historically, they have addressed a much more technical audience, and we've addressed a less technical audience. And so that's kind of been our differentiator as well as performance. And so look, we hear about them a lot. They're clearly growing very quickly. We're clearly growing very quickly. But I think all data management vendors, you're hearing us say very similar things on being a single pane of glass.

Just on that question, I mean, they haven't been afraid of doing $1 billion deals. I think they've nicked off 3 in the last 3 years. Why not get more aggressive in rolling some of these smaller AI stories up? They've been more aggressive, maybe -- who knows if it's working, but...

Yes, yes. I think we have a very high bar for what we would acquire. It needs to be a very strong team. It needs to have very differentiated, I would say, proprietary technology. And it needs to accelerate our product road map, and it needs to fit the right price. And so I think we're very thoughtful in how we want to approach inorganic growth, and it needs to check a lot of boxes. And so that's kind of our process. And so we'll continue to go down that path. But yes, it needs to make sense on a lot of different fronts for us.

Okay. Great.

Cloud migration is obviously a huge top of funnel for you. You talk a lot about that picking up in the enterprise in the last earnings call. Is the driver of that AI is the carrot? Is it a lot of the big on-premise estates, SAP, VMware, Oracle are finally starting to move? What do you see as the biggest tailwinds for that?

Yes. Interesting, we've tracked a lot of this commentary over the years, and sometimes it's correlated to us, sometimes it's not. It's kind of -- it's been very steady for us over the last 5 years regardless of quarter. And unfortunately, we're at the mercy of a lot of renewal cycles for on-prem data estates. And so like the example that I gave on the $100-plus million -- one of the $100-plus million deals, they've been on the platform, it will be 4 years in the fall. They signed an 8-figure PS engagement at the end of last year to take down more on-prem data estates as they came up for renewal. And that isn't just the Teradatas of the world. That's also the Hadoop environments. And so that's kind of why we benefit. It's hard to convince someone to start a migration and double pay for a long period of time unless that renewal is on the horizon.

Just in terms of next week for investors, anything -- you're hosting Mini Day, maybe recap what's happening? And we're doing a partner dinner at your favorite restaurant, Spruce. So if anyone wants to come kind of look me up, we got a...

I might be there.

Easen up. He's now a GSI.

So for next week, the keynote's Tuesday morning. You'll hear at our event, it's 1 to 3 in the afternoon on Tuesday. It's going to be a couple of hours. You'll hear from Sridhar. You'll hear from Christian, and then you'll hear from Mike Gannon, our CRO. We've been messaging to everyone that -- Mike has a family obligation at [ Scarpelli ], so he's not going to be there. And so we're not going to have a financial section. We feel like we need to host that event once we get through a CFO transition and have someone identified to kind of bless that more longer-term outlook. And so don't expect much from the finance section, but you'll hear how we view the data management life cycle and where we play and how we view that playing out longer term.

On the Salesforce dynamic, you seem too discerning because [indiscernible]. Can you just elaborate on why you're not [indiscernible]?

Yes. I think that they have a very strong foothold in the go-to-market office. I think we see a lot more data in Snowflake outside of go-to-market as well. And so when you think about a performant market-leading analytics product, that's something we've been developing for 13 years and invest a ton behind. In terms of the technical pillars of that, like everyone or most people know about our data sharing capabilities and how you can access third-party data without actually moving it across cloud, across regions. So all these different pillars that we've been working on for a very long period of time incentivize customers to put all types of data in Snowflake and enrich it and then run analytics on it. And so it's not that we don't think what Salesforce is doing is interesting. It's just not something that we're seeing our customers do -- choose us or them.

I'll give you a real-time view. Jefferies has Snowflake, Databricks and Salesforce. And Salesforce is not going to be the landing spot for AI. Snow and the other partners will. So because of all of our data lives in so many different systems, it's better populated in these systems to then do the work on top.

Yes.

So that's our strategy internally. We're not moving to Agentforce and others. Jimmy, thanks so much for joining.

Thanks for having me, Brent.