So you've been CEO of Snowflake for about 11 months now, correct? And something I get told about people who take over any organization, whether it be a corporate organization or maybe a nonprofit one, etc., is no matter how much you think you know about the organization and you had been there for a little while, you show up, you get behind the big desk, and you learn some big things you really didn't know. What is the biggest thing you learned about Snowflake that you had no idea about until you were officially CEO?
Look, I was actually at Snowflake for 10 months before I became CEO. I probably knew more about the company. Frank was nice enough to have me on the exec staff, was part of planning, you know, all of that is fine. But then when the rubber hits the road and you're like, okay, I am really going to look at what earnings next quarter is going to look like, stuff like that, there's that kind of cadence that you absolutely need to get used to. I'd say the biggest thing that I didn't quite expect was I knew that I had to accelerate product velocity. That was part of why I was brought in to run the company. But then changing how the sales team and the product teams work together to take products out to market, just that basic motion you need to do to take new products, was not quite there. That's what we worked on. And you know, 10 months later, it's a much better machine.
So you get to Snowflake because your company gets acquired by Snowflake. Before that, you were at Google. So now you're a public company CEO. David Solomon, CEO of Goldman Sachs, the other day said being public's not fun. The term he used was 'fun'. You have a lot of startup founders who have companies that could absolutely go public today but don't because they think it is more of a hassle than it's worth. Is it fun?
Yeah, it's a great job. Is it easy? No. But I also tell people, you know, all the high-impact easy jobs are gone. I don't really know why. And so, you know, you get to do what's left. Is it constrained? Absolutely. But I always believe in sort of creativity within constraints. And I think the rigor of actually having to show your books every quarter makes you do responsible things, makes you earn business responsibly. Yes, it's a constrained environment, but it's still Snowflake is a company making three and a half billion, growing at 30% year on year. Great place to be. Could be worse.
I like how you say it's the thing left. Is that what I should tell my kids someday? Look, if your dreams don't pan out, maybe you can fall back on being a CEO of a giant multi-billion dollar company. Is that what I should tell her?
That's fine. She can have low expectations.
You know, you talk about how when you got in, you had to accelerate product development. And there was a perception, I think maybe beyond perception, that Snowflake, which had been around for a while, when it came to AI, was behind the eight ball. Why do you think it was, and what was kind of the first thing you feel you had to change?
You know, Snowflake is a database company. There's a certain mindset that comes when you're a database company. It's all about big iron, heavy infrastructure. You got to be super careful. Even now, we can't afford to have a single data incident, a single security incident. There's a whole bunch of things that just have to be absolutely perfect on. And that's kind of how database engineers think about stuff. AI is just the opposite. As you know, every three months it seems like we have to revise all of our principles about what this thing is going to do, where it's going to be applicable, who has the best models. And so it's a much more iterative kind of environment. And so I've dealt with both kinds of teams. And so it is more of there's not a perfect product, you got to iterate your way to a perfect product. But teams are remarkably good at adapting. And I think we have done a fabulous job in the last year of being at the cutting edge of what AI and data can be.
You, I think you said, or Snowflake, you said the company, about, I want to make sure I get it right, that about half of your customers are now using your AI products. It was essentially non-existent the year earlier. Within that, because there's a lot of them, what specifically is getting the most traction?
It's the simple stuff. So for example, pretty much all of our customers write SQL with Snowflake. It's a programming language. It's typically considered easier than other languages like Python or Java or Go. And what we did was make it super easy to do things like sentiment detection, translation, summarization right from within SQL. So any analyst now can also be an AI analyst. That gets used a lot. The rest is honestly chatbots. But our secret sauce is we help people develop reliable chatbots.
Yeah, so when you use ChatGPT, you're always in like a guessing game of did this thing make up everything I'm seeing or can I actually rely on it? When you want a business use, that's not acceptable. So for example, we take great pains to make sure that we cite sources when it's looking at things like documents. Similarly, when you want answers to financial information, I can't put up an app in front of my CFO and say like, yeah, this is wrong 10% of the time. It's not something that's going to go. So we have developed basically a software engineering framework for how do you make that reliable. These are the classes of applications that are getting used by thousands of customers. Yes, we have the fancy stuff like an agentic platform that can do more. But I think getting the ingredients right has been the key unlock.
Do you find yourself ever getting ahead of your customers in the sense of, on the AI side, offering them things that they're really not internally equipped to really use yet or use efficiently? You have 10,000 customers, spans the spectrum.
So, you know, there's a whole bunch of people that be like, I want to be at the cutting edge, I want to check stuff out. We always work with them because they haven't got their data house isn't really in order yet. There are a ton of companies that, you know, that absolutely tell us we need to get these other things done first before we can think about AI. I generally try to not get into like an either-or sort of conversation. There are things that AI can do for you even if the entirety of your data estate is not perfect. It's our job, and this is a distinguishing factor of Snowflake, to make tech easy to use. And we're very, very proud of being able to do that.
So I was in San Francisco last month with the CEO of Databricks. And this predates you, this is your predecessor, but he and your predecessor had a lot of back and forths, public back and forths. And he said to me that I asked about competing with Snowflake and he said that is quote 'behind us'. Are you guys not competitors anymore? And does that mean he laughed you off?
Oh well, you know, Snowflake is the best AI data cloud that there is. And last I checked, you know, that's the product they were trying to create. So I'm not quite sure what Ali means. But Ali also said they wanted to compete with, you know, be more like Salesforce. So maybe they're doing SaaS software.
He did say that. I did ask him though, because they used to have a program called Snowmelt, I believe, and I asked if he had some sort of clever name for a Salesforce competing product. He said not anymore.
You know, look, there's not, first, it's a very big market. It's going to be, by some estimates, the data market, the data platform market is going to be $500 billion in 10 years. So there's plenty of business to go around. But when it comes to what Snowflake is, we prize simplicity and ease of use and how quickly people can get value from what they do. Their origin is more one of serving software developers. There's a lot of complexity, you have to spend a lot of people time and maintenance to do this. In that, we actually feel very good about where we are as a company. And I hear this repeatedly from customers. I spend a lot of time on the road and that's the consistent message that I get. Absolutely, we have accelerated in places like AI where honestly we were behind a year ago. So I feel very good about where we are.
