---
title: "Snowflake: The Data Startup That Split Storage From Computing"  
description: "The Snowflake startup story: its founders’ cloud-native data warehouse, separation of storage and compute, the 2020 IPO and its 2026 AI-driven growth."  
author: "Oliver Moore"  
published: 2026-11-13  
canonical: https://www.mindstick.com/startup/6507/snowflake-the-data-startup-that-split-storage-from-computing  
category: "Cloud Data Platforms"  
tags: ["Snowflake", "Data Warehouse", "cloud data", "Enterprise AI", "Frank Slootman", "Sridhar Ramaswamy"]  
reading_time: 4 minutes  

---

# Snowflake: The Data Startup That Split Storage From Computing

Companies collect vast amounts of data, from sales records to app usage. For years, analysing that data meant expensive, rigid systems that slowed down when many people used them at once. Snowflake rebuilt the data warehouse for the cloud and made it far easier to store, share and analyse information.

## Three Engineers and a Clean Sheet

Snowflake was founded in 2012 in San Mateo, California, by Benoît Dageville, Thierry Cruanes and Marcin Żukowski. Dageville and Cruanes were experienced database architects at Oracle, and Żukowski had co-founded a database company in the Netherlands. They believed that existing data warehouses had been designed for old hardware and could not take full advantage of the cloud.

They chose to build from scratch. Snowflake stayed in stealth mode for two years before launching publicly in 2014.

## The Key Idea

Snowflake’s main innovation was separating storage from computing. Data was stored cheaply in cloud storage, while computing power could be scaled up or down independently, even for different teams at the same time. A marketing team could run heavy reports without slowing down finance analysts working on the same data.

It also introduced consumption-based pricing. Customers paid for the computing they actually used rather than buying large fixed licences. This made it easy to start small and grow. Snowflake ran on Amazon, Microsoft and Google clouds, so customers were not tied to one provider. Features for securely sharing data between companies opened new uses, such as data marketplaces.

## Experienced Leadership and a Record IPO

Snowflake brought in experienced executives to scale. Bob Muglia, a former Microsoft executive, became chief executive in 2014. Frank Slootman, known for building ServiceNow, took over in 2019 and pushed hard on sales and growth.

In September 2020, Snowflake went public on the New York Stock Exchange in what was then the largest software IPO ever. Berkshire Hathaway was among the investors. The shares more than doubled on the first day, though the valuation later came down sharply as tech stocks fell.

## The AI Shift

In early 2024, Slootman retired as chief executive and was succeeded by Sridhar Ramaswamy, a former Google advertising executive who had joined Snowflake through the acquisition of his AI search startup, Neeva. Ramaswamy has focused on making Snowflake a platform for building AI applications on top of company data, with tools for AI agents, natural language queries and coding assistance.

## Where Snowflake Stands in 2026

Snowflake’s growth is accelerating. In the quarter ended July 2026, product revenue rose 37 percent from a year earlier to about 1.49 billion dollars, its third straight quarter of faster growth. The company raised its full-year product revenue forecast to about 6.07 billion dollars, up 36 percent, and said it had 828 customers spending more than one million dollars a year. Its AI tools are being adopted by thousands of customer accounts.

Snowflake still reports losses under standard accounting rules, largely because of stock-based pay. Its challenges include strong competition from Databricks and from the cloud providers’ own data services, and the need to keep costs manageable for customers running heavy AI workloads.

## How Snowflake Makes Money

Snowflake earns most of its revenue through consumption. Customers buy credits that are used up when they run computing workloads, such as queries, data processing and AI tasks, along with charges for storing data. Many large customers commit to spend a certain amount over one or more years in exchange for discounts, which shows up in Snowflake’s remaining performance obligations, a measure of future contracted revenue. Because revenue depends on usage, it grows when customers find more ways to use their data and can slow if they optimise costs. Snowflake also earns from its marketplace, where companies can share or sell data and applications to other Snowflake users. Its newer AI features and tools for building data applications are designed to increase consumption further. Professional services, such as help with migrations, add a smaller amount. The company’s large customer base, including hundreds of the world’s biggest companies, gives it a strong foundation for continued growth.

## Lessons for Founders

Snowflake shows that rethinking a product from first principles can beat incumbents with decades of experience. It also shows how pricing can drive adoption: pay-as-you-go made it easy for customers to try and expand. And its leadership changes reflect a common startup pattern, with technical founders building the core and experienced operators helping it scale.

Three engineers set out to build a better data warehouse. They created one of the most important platforms for how companies use data, and now AI.

---

Original Source: https://www.mindstick.com/startup/6507/snowflake-the-data-startup-that-split-storage-from-computing

Copyright © MindStick Software Pvt. Ltd. This Markdown version is provided for developers, AI systems, and offline reading.
