In the high-stakes world of enterprise software, a clear divide is emerging. Winners. Losers. And Sridhar Ramaswamy, the chief executive at cloud storage giant Snowflake, clearly believes his company is on the favorable side of that chasm.
His conviction? Not unfounded. Snowflake just clocked a blowout first quarter, delivering a performance that obliterated expectations across the board. The market responded, violently. Shares rocketed up 36% this week, extending five-day gains past a staggering 50%. An eye-popping $6 billion commitment to Amazon over the next half-decade, earmarked for the tech titan’s Graviton chips, further signals robust demand. This isn't just growth. It's an affirmation.
These numbers arrived precisely when Snowflake, and indeed many software-as-a-service (SaaS) firms, desperately needed them. Investor jitters about AI replacing traditional software vendors had triggered a brutal slump. Many saw their valuations decimated. But Snowflake, among a cadre of entrenched players, is anchoring itself, having launched aggressive AI initiatives that weave agentic technology directly into its vast data operations. The Q1 results—revenue up 33% year-over-year, the fastest clip in two years—aren’t just good news. They validate a core strategy. Ramaswamy himself pointed out it proves the long-held consumption-based pricing model works, and that traditional software can indeed morph into AI compute powerhouses.
“It’s important to understand that all software companies are not the same,” Ramaswamy told Fortune, just ahead of Snowflake's San Francisco tech summit.
The critical difference, he maintains, is Snowflake's pricing model, consumption-based from day one. “We recognize revenue only when a customer actually uses Snowflake’s capabilities,” he explained. “We have to show value to make money.” A simple, powerful truth.
“We recognize revenue only when a customer actually uses Snowflake’s capabilities. We have to show value to make money.”
Agentic AI, an entirely new species, has exerted immense pressure on the industry’s antiquated enterprise seat-based pricing model. Ramaswamy predicts a future where companies clinging to seat-based income will scramble. They'll have to justify their premiums. Especially as AI empowers employees to accomplish what once took an army.
Ramaswamy assumed the CEO role at Snowflake in 2024, right as the AI boom exploded. His company’s wager: a foundational “infrastructure layer” supporting user-facing products, coupled with that resilient consumption model, positions Snowflake for the long haul. A shrewd bet, it seems.
Snowflake's AI journey began in earnest about two and a half years ago. They built Cortex Code, an innovative coding agent now deployed across more than 7,100 accounts. Snowflake Intelligence, another agentic application, has seen its user accounts more than double quarter-over-quarter. Now, the horizon features what Ramaswamy dubs the “control plane.” He sees it as a “cockpit of work,” where users don't just query data, but orchestrate complex tasks across disparate applications. “I liken it to the new browser,” he said, painting a vivid picture of the future.
This ambitious strategy leans heavily on Amazon. Snowflake is doubling down on the cloud provider, primarily for its chip performance. Amazon is, in fact, Snowflake’s largest partner, powering over 70% of its operations.
Other titans face their own battles. Salesforce CEO Marc Benioff recently touted “record levels” returned to investors, citing a colossal $25 billion share repurchase. Salesforce's AI product, Agentforce, shows promise, but the company still chases the growth that truly galvanizes investors. The so-called “SaaSpocalypse” fears persist. Yet, for deeply entrenched players, sentiment gradually brightens.
Ramaswamy shares Benioff's optimism, even as major labs like Anthropic experiment with highly autonomous systems, such as the much-hyped Mythos model. He wouldn't confirm if Snowflake had early access to Mythos. But he did stress the imperative for responsible companies to harness such potent technologies—perhaps for automated security scans on their own software. “You have to figure out how to harness the awesome power of these coding agents and put them to work in a responsible way,” he asserted. His personal take? “I’m also very paranoid about making sure that I actually know what it’s doing and give permissions to it.” A healthy dose of skepticism, in a world rushing headlong into AI.
He foresees a dramatic shift, away from countless “off-the-shelf” SaaS applications. A future with far fewer major applications, yes, but also a surge in bespoke, small-scale solutions. “There will be major applications that folks will continue to buy, but there will definitely be a consolidation.” The market, it seems, is only just beginning its brutal culling.
No comments yet. Be the first to share your thoughts!