Snowflake and I faced a trial by fire but we're stronger now, says CEO Sridhar Ramaswamy

You took charge as CEO in February this year. Has it been a trial by fire?

Always excited and humbled to be back in Bangalore. I grew up and did my high school here.

I think Snowflake went through a trial by fire this year, as did I, but I think we are much, much stronger. Snowflake was a company that built an amazing product that was years ahead of its time, and they did it like eight years ago. And all successful companies, then start to believe that everything they do will be successful. They forget that you must keep earning that success. But companies are also very resilient.

My goal this year was to make sure that we got on a much faster cadence of just product development. And in the world of AI, things are changing so rapidly. Nobody would have predicted that in December this year, we would be at this point if we had made the prediction in January.

Snowflake stepped up when it comes to product delivery.

We invested a lot in go-to-market things—in how do we get product managers and engineers and marketers and salespeople to work together.

But in life, you must grind a bit, and then you get great things. And in Q3, we showed not only strength in our core business but also progress in new areas like data engineering and AI. And so, the stock market reacted in a positive way. But more importantly, I think there is just a bit of confidence in me and confidence in the team that they can shape this future.

And I try to be at a place where I am neither a little too cocky, too confident about driving outcomes, nor too timid. I feel very good about where we are, coming up on basically the end of my first year as CEO.

Also read: Meet IIT Madras alumnus Sridhar Ramaswamy, the new Snowflake CEO

Your recent Q3 performance — I think your shares rose 32 percent after results, which was actually the best day for Snowflake since it went public in 2020. Does that feel like validation?

I would put it slightly differently. I think Snowflake is feeling confident that it can be pretty much close to 30% growth year on year, and the stock market believing that we can continue to sustain high growth rates—I would say that's the positive news that came out of Q3.

Even the numbers, if you're growing at 40% year on year, you're doubling every two years. You're taking a business that's three and a half billion dollars, and you're saying you're going to double that every two-ish odd years. That's astronomical growth.

I would say that is more the thing that I feel very good about. It's the confidence in ourselves that we can do it, and it's the belief on the part of our investors and the external market that we can do it.

You've transitioned from founder mode to manager mode. Do you miss being a founder?

No. I thought a lot about this. So, I ran a team with over 10,000 people at Google, running a business that was over $100 billion. And then I decided to do a startup with two of my friends, Neeva, which was a search engine. And it was very much like a reset, a back-to-basics.

But as I was doing Neeva—I loved doing Neeva, it was a passion project of mine—but on the other hand, I also realized that there were many things that I was good at: setting strategies, building teams, working across complicated issues like regulation, new product, like what are the right companies to be acquired. I felt that a small company like Neeva didn't necessarily use all the things that I had learned.

And so, I feel much more at home at Snowflake. That doesn't mean that the inner hacker in me or the inner developer in me is going to disappear. I get joy out of trying new products, both ones that Snowflake built and ones that others built.

I tell our team that our job is to create great products, and that means that every person in the company understands it—certainly the product and engineering and the salespeople.

I think it's a benefit for Snowflake that a technologist is at the helm.

Obviously, we have a wonderful team of great product managers, of great engineers. But at the end of the day, I think the tone of the company is also set by the leaders. I feel it's a pretty good match for what I bring to the table.

Depending on who you talk to these days, they have very extreme views on software as a service broadly. People who feel that it's going to be RIP SaaS because it's going to be disrupted by AI. How do you see the SaaS space getting reshaped?

First, it's important to understand that there was a bit of irrational exuberance about SaaS companies during the pandemic.

In terms of the valuations?

In terms of valuations, the logic was that these were subscription companies. They could make very predictable revenue.

So, I think part one is really resetting that expectation. Because as things like interest rates go up, basically the cost of capital effectively is going up, and these companies have a higher bar to prove.

Now you're asking a very different question about whether the model itself will be under threat in the world of AI. I think there is a chance for disruption. So, for example, at Snowflake, we have a consumption model. We still sign contracts with people, but we can recognize revenue only when people use our product. That's one thing to keep in mind—how strong is the model of your subscription itself?

I think the second and more interesting question is about data. All our life, we bought software by buying boxes to run software.

In the world of SaaS, though, these services ran on the cloud, which was nice, but they were still virtual boxes. If you used some product—it doesn’t matter, Salesforce, something else—the data went into that virtual box, and it's hard to take it. But what we are seeing, for example, is increasingly a lot of Snowflake's customers basically sync data from all of these applications into Snowflake to get a global view.

This is where you bring this together so it can be for analytics. And this is the real threat. What AI is going to do is it's going to make analytics a lot easier because it's easier to compose queries.

It's easier for business users to get answers that go across ERP, that go across CRM, and so on. But I think AI can also help create new applications that bring together things. Right now, it is next to impossible for your company, let's say, to write an application that shows information from Salesforce and from Workday on the same screen and lets you update it. The SaaS model makes that very hard.

And so, I think that is the threat. Whereas data goes into basically open things like open data lakes, and open data estates, people are going to be come up with new kinds of applications powered by AI that they would not have thought about before. And it's potentially the combination of these AI-driven applications and a consumption model that can turn out to be a larger threat to the SaaS model.

But people are not sitting still. SAP is releasing better analytics products, as they should, and so is Salesforce when it comes to what you can do for analytics.

But I think there is absolutely a lot of disruption coming to the world of SaaS and software.

And since you took charge, you've infused AI in your product offerings. You recently unveiled Arctic LLM. Did you think that this had to be done with a sense of urgency? What kind of mindset shift did it require internally?

All great companies are driven by a sense of urgency, are led by people that will tell you, "I don’t understand why this wasn’t done yesterday." I worked a lot with people like Larry Page and he was famously impatient.

I’m a very motivated person. And I’m not always proud of it, but I’ve never taken any time off between my jobs. I pretty much finish on Friday and start on Monday.

I bring that mentality to Snowflake, and I drive myself. I drive my team relentlessly hard. And to be honest, this is not to take away from things like family absolutely comes first. Your sanity comes first. But I don’t see these as either-or. This is like, we can be sane and look after your family and work hard.

And equally bluntly, there are a set of people that will basically be like “I just want to work like six, seven hours, take a long lunch break, and not have to be so driven and so motivated." I have nothing against them. I’m not sure Snowflake is the right place for them.

The big questions when it comes to integrating AI into enterprises is around reliability, cost, security. CFOs are also asking if they’re spending too much too soon.

We try very hard to make sure that the products that we build are easy to use. We try very hard to make sure that they are efficient. You don’t have to spend a lot of money. We also place a lot of emphasis on trustworthiness: that this AI application should do what it says it does.

The more people internalize tools, the more they’ll be able to come up with, "Here are business situations that are helpful for the company. With AI, it’s pretty much down to what all applications you can think of that either help you make more money or that help you spend less money.

Do you see a consolidation of the big AI models? Where will the horizontal players score and where will the vertical players score?

These are very hard questions. And they seem to change every six months.

As of the end of 2024, there are already clear winners. It feels like the gap—there is a very large and widening gap—between the likes of OpenAI, Anthropic, and Google. These are pretty much the only teams that seem to be at the cutting edge of AI.

No one knows if this is going to change in the next six months.

But there are also new ways to create value that is not quite going to be how incumbents think about value. For example, companies like MidJourney—they're doing very well. Applications of AI into different verticals—companies like Harvey or legal—are doing very well.

So, it is a combination of deep expertise in a particular area, and knowing the workflows for information within the area, and it increasingly looks like all these companies would be better off using one of the best models in the world as opposed to also needing to be a great foundational model company.

You can argue that much of the value that was created by the internet, at least in the United States, didn't go to—I mean, some of it went to consumers—but a lot of it went to a handful of companies. And so, I'm hoping that there's a much larger set of companies, and effectively people, that realize the value of AI.

What do you think Perplexity got right? And where did they succeed, where Neeva did not?

At Neeva, we spent three years creating a search engine from scratch, where we did the crawl, we did the indexing, and we did a lot of heavy work. Perplexity came quite a bit later, and they simply relied on APIs and produced a great product. That really was what drove the positive momentum for the company.

All startups have limited shelf lives in terms of how much you can keep going. My team in the last year sort of was tired of the uncertainty of being a startup and the relentless number of fundraising rounds that you must do. Perplexity created a great product, and they've been very quick to innovate. I give Aravind (Aravind Srinivas, founder)—also an IIT-Madras graduate, a lot of credit. What they have done is truly remarkable.

In terms of your plans for India—hiring—I think you also have a CoE (Centre of Excellence) in Pune. What can we expect in 2025?

We have a great team in Pune, and it has a variety of functions. We have folks in IT, and folks in finance, and they're generally somewhat internally focused. We're hoping to have a product team in Pune.

Independently, we’re staffing up a very large team that will work with GCCs. This will include a lot of our solution architects, for example, the technical people who make Snowflake come alive. Increasingly, more and more companies are setting up GCCs in India.

Sometimes, abbreviations don’t convey the spirit of what is going on. What is really going on is that many companies are effectively setting up very large technology centres in India. But the interesting thing about these technology centres is that they drive business. This is no longer back-office work. This is no longer coding work. This is the heart of banks, of other kinds of financial institutions.

So, these are very important segments being set up in India, and that’s the opportunity we see as well. The most progressive account executives and solutions architects on my team in New York now are traveling to India because they realize increasingly that their customer is in India, and the decision maker is in India.

These are ambitious, qualified, amazing folks who are responsible for entire subsystems, the software, and increasingly the processes that drive business. It used to be that software was like a little thing that you used on the side, but software now for a bank or a financial institution is its business. And so, we are headed into a very different role, and Snowflake will absolutely be leaning in. We will invest a lot more in India.

You got the Distinguished Alumnus Award from IIT-Madras. Do you still have close ties with the institute?

My wife and I have a foundation, and we contribute to things like student scholarships in IIT Madras. I obviously have a lot of friends from IIT Madras. Those are connections that will never go away in your life. I’m very excited for what the institution has done and will do, and being able to do more meaningful things with them is certainly among my wishlist of things to do. But I’m also a big believer in doing just like one or two things but giving all of my energy into it.

And as far as my personal life is concerned, 2024 has been the year of Snowflake, but all my energy and time has gone into it.

But look, somebody like me would simply not have gotten to this spot without the education, without the selfless sacrifice that many of those professors who could have chosen to live in any place in the United States, teaching at any university, they chose to teach us. And I'm very grateful for what they did.

My high school in Bengaluru, Hombegowda Boys High School was a Kannada medium high school, but there was this amazing math teacher that I had, whose name was Indira, who encouraged me to learn more, to study more. I'm very cognizant of how much this country has contributed to the prosperity in my life.