Gunjo · Business Intelligence for the AI Era
← Sticker Wall JOURNEY · DETAIL

Snowflake: Veteran Founders Bet Against Consensus on Cloud Data Warehouse, Halving After Largest Software IPO and Rebounding via AI

Founded: Benoit Dageville, Thierry Cruanes (former Oracle data architects), Marcin Zukowski (co-founder of Vectorwise) · Snowflake Inc. (NYSE: SNOW)

JOURNEY

Key Fields

FIELD STAMPS
IndustryCloud Computing
RegionUS
ScaleGiant
ChannelOther

Origin

In 2012, three database veterans founded Snowflake in the San Francisco Bay Area. The founders came from Oracle's elite architect team and Europe's columnar database research circle; they were seasoned professionals rather than viral hackers or open-source gurus. At a time when Hadoop and NoSQL were all the rage in Silicon Valley and public opinion had pronounced relational databases dead, they went against the consensus to assert that relational plus SQL wasn't dead—it was simply dragged down by legacy implementations. They believed that rebuilding it for the cloud would unleash massive pent-up demand from the enterprise market. From day one, they faced a dilemma: building a cloud-native service while piggybacking on AWS infrastructure meant sharing a table with their biggest competitor while selling products.

Milestones

2013
Contrarian Start Growth
Dageville and Cruanes spent years at Oracle as core database kernel architects; Zukowski participated in the columnar query engine Vectorwise in Europe. The trio judged that relational plus SQL was far from obsolete; the problem lay in on-premise implementations of legacy systems like Oracle and Teradata, which failed to meet the elastic demands of cloud-native enterprises. Peter Wagner of Wing Venture Capital led the seed round in January 2013 after meeting them, though he admitted almost no one around him believed in three things at the time: that relational still had a future, that they could beat Amazon, and that enterprises would dare put their data in the cloud.
2014
Emerging from Stealth Turning Point
In June 2014, former Microsoft executive Bob Muglia was brought in as CEO, taking over from the first CEO and Sutter Hill investor Mike Speiser. In October of the same year, the company emerged from stealth mode, debuting with 80 pilot customers while running on top of AWS. At that time, Amazon Redshift had already entered the market; Snowflake had to leverage AWS compute and storage to build its service while simultaneously competing head-on with AWS for data warehouse customers, which peers viewed as a suicide mission.
2015
Product Launch PMF
Formally launched its inaugural cloud data warehouse product while closing a $45 million funding round. Snowflake integrated storage-compute separation into a cloud-native service where customers pay-as-you-go and spin up virtual warehouses in minutes, offering an experience vastly lighter than legacy on-premise data warehouses that took months to deliver. A sales leader at the time described Snowflake as addictive, with customers continuously increasing usage once they started, driving Net Revenue Retention (NRR) steadily higher to become the capital market's most coveted core metric.
2016
Near-Failed Financing Failure
Snowflake burned heavily on sales acquisition to capture customers. Institutional investors focused on unit acquisition costs were unenthusiastic about the ratio of spending to newly signed ARR, and the 2016 funding round nearly collapsed, ultimately rescued by internal investors stepping in to backstop the deal. Wing later wrote candidly in a retrospective: none of the institutions posting S-1 breakdowns on Twitter and appearing at the IPO celebrations back then were shareholders; every single one of them missed this round.
2018
Unicorn Sprint Growth
Raised a $100 million Series D in April 2017, bringing total funding to $205 million; raised another $263 million in January 2018 at a $1.5 billion valuation to enter unicorn status, followed by a $450 million Sequoia-led round in October at a $3.5 billion valuation, bringing total funding close to $1 billion. NRR remained consistently high, and the consumption-based billing model caused customers to expand usage over time, transforming the platform from a single data warehouse into a cloud-neutral data foundation spanning AWS, Azure, and GCP.
2019
Leadership Change Turning Point
Suddenly changed leadership over a year prior to going public, with former ServiceNow CEO Frank Slootman replacing Muglia. Slootman was renowned for his aggressive sales culture and pacing of the IPO process, perfectly complementing the Snowflake founding team's tech-heavy, commercial execution-light weakness right before the IPO. The news of the CEO change shocked the industry, but in hindsight, it was the pivotal move that translated database technology accumulation into a capital market success story.
2020
IPO Pinnacle Growth
Listed on the NYSE on September 16, raising approximately $3.4 billion and setting the record for the largest software IPO to double on its first day. Warren Buffett's Berkshire Hathaway and Salesforce purchased hundreds of millions combined during the IPO, interpreted by the market as a top-tier endorsement. After the stock doubled on day one, its market cap surged past $70 billion, but this joint surge also meant the company was locked in as one of the most expensive software stocks, leaving extreme pressure for subsequent growth.
2021
Market Cap Peak Turning Point
Driven by the cloud and SPAC boom, SNOW's stock price briefly touched highs above $400, with its market cap approaching $100 billion. However, the company had consistently stated in its prospectus that it operated at a continuous loss, driven by high sales expenses. The lofty valuation corresponded to extremely demanding expectations for NRR to remain high over the long term, and the market began searching for validation of this optimistic pricing.
2024
CEO Departure & Drop Failure
On February 28 after market close, it was abruptly announced that Frank Slootman was retiring and Neeva co-founder Sridhar Ramaswamy would take over as CEO, causing the stock to plunge over 20% in a single day. The market not only had to digest the shock of the sudden leadership change but also reassess whether the new CEO could defend the growth narrative under the dual squeeze of Databricks and cloud providers. Ramaswamy faced an immediate dual test of confidence and velocity upon taking office.
2024
Data Breach Storm Failure
ShinyHunters and UNC5537 used credentials stolen by infostealer malware to attack a batch of Snowflake tenants lacking MFA, impacting over 160 organizations including Ticketmaster, Banco Santander, AT&T, Advance Auto Parts, LendingTree, and Neiman Marcus. Mandiant's investigation concluded that Snowflake's platform environment itself was not breached, with the root cause stemming from customer-side credential theft combined with failure to enable MFA. However, reputational damage was already done, and the boundaries of security responsibility for cloud data foundations were thrust into the spotlight.
2025
Trough Rebound Turning Point
Accompanied by tightening interest rates and a mocking slowdown in cloud growth, SNOW's stock price dropped sharply from its peaks, colloquially dubbed a fall from the clouds to the earth. After hitting a stumbling block with the $185 million acquisition of Neeva, Ramaswamy pivoted the strategic anchor firmly toward AI: Cortex allowed customers to invoke large language models directly inside the data cloud; Snowflake Intelligence used natural language to query structured and unstructured data while integrating with agents; and in June 2025, the company acquired Crunchy Data for roughly $250 million to bolster PostgreSQL capabilities.
2026
AI Era Rebound Growth
FY2026 Q4 product revenue reached $1.23 billion up 30% year-over-year, full-year product revenue hit $4.72 billion, RPO grew to $9.77 billion up 42%, and over 9,100 accounts used Snowflake AI features. In July 2026, the company launched the Snowflake Intelligence enterprise AI assistant with native Model Context Protocol support, repositioning itself from a data warehouse to a trusted data foundation for agents.

Turning Points

  • Brought in Bob Muglia in 2014 to emerge from stealth, equipping a purely technical founding team with a commercial face.
  • Replaced leadership right before the 2019 IPO with Slootman taking the helm, translating technical assets into a capital market narrative.
  • Financing nearly failed in 2016 and relied on internal backstops, exposing high customer acquisition costs and the fragility of institutional confidence.
  • Slootman's sudden retirement in 2024 compounded by a major customer data breach storm plunged the company into a trough, forcing a strategic re-anchoring toward AI.

Failures & Pitfalls

  • The 2016 financing nearly collapsed as institutional investors collectively pulled out due to an overly high ratio of sales expenses to newly signed ARR, ultimately rescued by internal investors stepping in—driven by insider conviction rather than market enthusiasm.
  • The abrupt announcement of the CEO's retirement in February 2024 wiped out 20% of the stock price in a single day. The market interpreted this transition-less leadership change as a governance risk, evaporating far more than technical issues overnight.
  • Over 160 customers were attacked by ShinyHunters in 2024. Although Mandiant's investigation determined the platform was not compromised but rather customer-side credentials were stolen, Snowflake as a cloud data foundation was thrust into the public eye, fully exposing vulnerabilities in security boundaries and customer education.
  • Neeva, acquired for $185 million, failed validation in consumer search before being integrated into Snowflake. In hindsight, the market logic of this acquisition looked more like paving the way for Ramaswamy's takeover, with questionable commercial returns.

关键成功要素

  • Storage-compute separated cloud-native architecture allows customers to pay-as-you-go and scale within minutes, delivering an experience that crushes legacy on-premise data warehouses and providing the technical prerequisite for moving the legacy enterprise market to the cloud.
  • Persisting with relational plus SQL as the primary contrarian thesis, capturing millions of existing analysts and DBAs within enterprises alongside their entrenched ecosystem tools, bypassing the need to educate the market from scratch.
  • Cloud-neutral unified service across AWS, Azure, and GCP grants an extra layer of trust to enterprises fearful of vendor lock-in, forming a moat against Amazon Redshift.
  • Consumption-based billing drives addiction-like expansion and high NRR, where customers increase usage as they go, turning the data warehouse into a self-propagating billing engine—the core narrative anchor that commanded high valuations from the capital markets.
  • Brought in Slootman prior to the IPO to reinforce aggressive commercial execution, followed by Ramaswamy re-anchoring strategy onto the AI data foundation during the AI turning point; both critical leadership changes hit the right rhythm at the right time.

Lessons

  • The prerequisite for betting against consensus is deep foundation and domain expertise; Oracle veterans declaring relational databases were not dead carried more credibility than outsiders. Depth equals judgment.
  • Piggybacking on the infrastructure of your biggest competitor is not necessarily a dead end. As long as product strength and the cloud-neutral narrative are robust enough, you can turn Amazon—simultaneously a supplier and competitor—into a peer competitor rather than getting wiped out.
  • A high-NRR consumption-based model can sustain astronomical valuations during capital market booms, but it will equally face swift retribution when growth slows. Valuation cycles are harder to navigate than the business model itself.
  • The security responsibility boundary of a cloud data foundation is a grey area; a platform not being breached does not mean customers suffer no losses. Enforcing default MFA and baking customer security education into the product are no longer optional.
  • The damage inflicted by an abrupt CEO departure far exceeds operational handovers. A transition-less leadership change hits both valuation and morale simultaneously, and providing the market with stable expectations is more critical than who replaces them.

Core Data

  • FY2026 Full-Year Product Revenue：$4.72 billion (Company disclosed figure, as of 2026, independent review unverified)
  • FY2026 Q4 Product Revenue：$1.23 billion (Company disclosed figure, as of 2026, independent review unverified)
  • FY2026 Product Revenue YoY Growth Rate：30% (Company disclosed figure, as of 2026, independent review unverified)
  • FY2026 Remaining Performance Obligations (RPO)：$9.77 billion (Company disclosed figure, as of 2026, independent review unverified)
  • FY2026 RPO YoY Growth Rate：42% (Company disclosed figure, as of 2026, independent review unverified)
  • Accounts Using Intelligence Features：9,100+ (Company disclosed figure, as of 2026, independent review unverified)
  • IPO Funds Raised：Approx. $3.4 billion (Company disclosed figure, as of 2026, independent review unverified)
  • Slootman Departure Single-Day Drop：20%+ (Company disclosed figure, as of 2026, independent review unverified)
  • Data Breach Impacted Organizations：160+ (Company disclosed figure, as of 2026, independent review unverified)

Competitors / Peers

Competing head-on with proprietary data warehouses from the three major cloud providers: Amazon Redshift, Google BigQuery, and Microsoft Azure Synapse; contending in the independent track with Spark-origin Databricks for dominance in lakehouse and AI workloads; similar consumption-based models also face fragmentation from emerging contenders like Firebolt, Starburst, and Dremio.

Sources

Related

Same industry / scale
Cloud ComputingUS

CoreWeave: The AI Compute Rising Star Transforming from Crypto Mining to GPU Cloud, Tied to NVIDIA

Cloud ComputingUS

Cloudflare: From a Harvard Business School Project to the World's Largest Edge Security Network

Cloud ComputingUS

Databricks: From a 'Worst-Ever' Pitch Deck to a $188 Billion Data and AI Platform by Seven Berkeley Scholars

Cloud ComputingEurope

Nebius: From Yandex Asset Spin-Off to European AI Cloud Computing and Autonomous Driving Infrastructure Giant

⚠️ This entry draws on self-reported accounts and public reporting, not independently verified, and is not investment, startup, or legal advice; revenue figures are case-specific and not representative of typical results.