SaaS / Enterprise SoftwareUS
← Sticker Wall
JOURNEY · DETAIL
Oracle: How Ellison Turned a CIA Contract into a Database and Cloud Acquisition Empire
Founded: Larry Ellison, Bob Miner, Ed Oates · Oracle Corporation
JOURNEY
Key Fields
FIELD STAMPSIndustrySaaS / Enterprise Software
RegionUS
ScaleGiant
ChannelOther
Origin
In 1977, Larry Ellison—a two-time college dropout working odd jobs in Silicon Valley writing code—read a research paper on relational databases by IBM researcher Edgar F. Codd and learned that the CIA had an outsourcing requirement for a database project code-named Oracle. Alongside Bob Miner and Ed Oates, he pooled $2,000 ($1,200 from Ellison) to found Software Development Laboratories (SDL). They secured the CIA contract, named their product Oracle, and gambled on the window of opportunity created by IBM publishing papers rather than building products, using the SQL standard to pioneer the commercialization of relational databases.
Milestones
1977
Founding and First Contract Turning Point
In 1977, the trio founded SDL with $2,000, surviving on the CIA Oracle project contract. In 1979, they launched Oracle Version 2 (naming it V2 directly to project maturity), making it the first commercial SQL relational database. Backed by the CIA and military intelligence agencies from the start, the company established its DNA in direct enterprise and government sales. This phase lasted from 1977 to 1979.
1986
IPO and Near-Death Crisis Failure
Oracle went public in 1986 with roughly $55 million in revenue. However, to meet performance targets, the sales team aggressively recognized revenue ahead of time by booking multi-year license fees all at once. In 1990, the company was forced to restate earnings, resulting in a quarterly loss and a single-day stock plunge of about 30%, pushing the firm to the brink of bankruptcy. Ellison faced a class-action shareholder lawsuit, settling for roughly $100 million. This marked his biggest personal failure and turning point. This phase lasted from 1986 to 1990.
1992
Oracle7 Comeback and Internet Boom PMF
In 1992, Oracle launched Oracle7, an engineer-led release that resolved quality and stability issues, returning the company to growth. Riding the internet wave in the late 1990s, Oracle championed the slogan that software would disappear into the network. By around 1999, revenue approached $10 billion, and at the peak of the dot-com bubble in 2000, Ellison's net worth briefly surpassed Bill Gates to become the world's richest person. This phase lasted from 1992 to 1999.
2003
Hostile Takeover Addiction Turning Point
In 2003, Oracle launched an 18-month hostile takeover of PeopleSoft, ultimately closing the deal for approximately $10.3 billion. This was followed by the acquisition of dozens of companies such as Siebel and BEA. In 2009, Oracle announced a $7.4 billion acquisition of Sun Microsystems, unexpectedly capturing Java and MySQL and expanding its software dominance into hardware. M&A integration became Oracle's core growth engine. This phase lasted from 2003 to 2010.
2016
Difficult Cloud Transition Turning Point
Facing the rise of AWS, Oracle made a forced push into the cloud in 2016 by acquiring cloud ERP vendor NetSuite for $9.3 billion. However, its homegrown cloud infrastructure (OCI) was long mocked as a tier-three player with single-digit global cloud market share. Ellison's strategy shifted toward tying up major clients via multi-cloud interconnects—such as partnering with Microsoft Azure to integrate database services—slowly migrating legacy database customers to the cloud. Growth hovered in the single digits for years during this transition. This phase lasted from 2016 to 2023.
2024
AI Compute Gamble and Peak Growth
Beginning in 2024, Oracle made massive investments in building AI data centers and signed a compute partnership for OpenAI's Stargate project. In September 2025, earnings reports revealed that Remaining Performance Obligations (RPO) surged to roughly $455 billion, sending the stock soaring roughly 36% in a single day and pushing the company's market cap near $1 trillion. 81-year-old Ellison saw his net worth exceed $380 billion, briefly surpassing Elon Musk as the world's richest person, completing a stunning leap from legacy database giant to AI pick-and-shovel provider. This phase lasted from 2024 to 2025.
Turning Points
- The 1990 collapse triggered by aggressive revenue recognition forced Oracle to shift from a sales-driven culture to a dual-engine model of engineering and financial discipline.
- The 2003 hostile takeover of PeopleSoft established Oracle's playbook of buying growth through M&A and absorbing competitors' customer bases.
- The 2009 acquisition of Sun Microsystems gave Oracle unexpected control over Java and MySQL, expanding its ecosystem moat from databases to the entire enterprise tech stack.
- The 2024 all-in bet on AI data centers and partnership with OpenAI positioned a 47-year-old company as an AI-era compute landlord in the second half of the cloud computing race.
Failures & Pitfalls
- Aggressive revenue recognition in 1990 led to an earnings restatement, a roughly 30% single-day stock crash, near-bankruptcy, and hundreds of millions in legal settlement costs.
- Despite early rhetoric around network-centric software, Oracle missed the SaaS wave, losing the CRM market to Salesforce, founded by its own former executive.
- Homegrown cloud OCI started late with a hesitant strategy, remaining long suppressed to single-digit market share by AWS, Microsoft Azure, and Google Cloud.
- Around 2016, Oracle heavily relied on layoffs and price hikes to maintain profit margins, drawing industry criticism for squeezing legacy customers through legal and audit penalty pricing tactics.
关键成功要素
- Anchoring early startup contracts with government clients like the CIA, leveraging intelligence agency backing to unlock trust in the enterprise market.
- Naming the first product Version 2 and beating IBM to market by implementing SQL according to IBM's paper standards, seizing technical definition rights through marketing audacity.
- Introducing professional financial controls and betting on Oracle7 engineering quality after the 1990 crisis, completing a self-redemption from deceptive sales to a reliable vendor.
- Eliminating competitors and absorbing their customer bases through over a hundred acquisitions, turning M&A into a more efficient growth machine than in-house R&D.
- Going all-in on AI data centers in later years, locking in future revenue upfront via long-term RPO contracts, and betting on becoming one of the AI era's top pick-and-shovel providers.
Lessons
- Great starting points often combine someone else's research paper with a government outsourcing contract; execution speed is worth more than original invention.
- High growth fueled by aggressive accounting will eventually be audited; the 1990-style crash serves as a warning for all revenue-chasing companies.
- Missing a generational technology wave does not mean being knocked out; buying a ticket via acquisitions allows late entrants to leapfrog ahead.
- Locking into a hyper-demand customer (the early CIA, today's OpenAI) is the shortest path for a small company to disrupt a large market.
- When a founder in their seventies dares to stake their balance sheet to gamble on a new technological paradigm, organizational inertia proves to be an established company's true enemy.
Core Data
- 创始资本：Startup seed capital of $2,000 (based on public sources, independent verification pending)
- 1986年营收：Revenue of approximately $55 million at the 1986 IPO (based on public sources, independent verification pending)
- 2003年收购仁科对价：Approximately $10.3 billion spent acquiring PeopleSoft from 2003 to 2005 (based on public sources, independent verification pending)
- 2010年收购太阳微系统对价：Approximately $7.4 billion spent acquiring Sun Microsystems from 2009 to 2010 (based on public sources, independent verification pending)
- 2025年剩余履约义务：Remaining Performance Obligations (RPO) of approximately $455 billion in September 2025 (based on public sources, independent verification pending)
- 2025年股价涨幅：Single-day stock price increase of approximately 36% on September 10, 2025 (based on public sources, independent verification pending)
- 2025财年营收：Fiscal 2025 revenue of approximately $57.4 billion (based on public sources, independent verification pending)
Competitors / Peers
Long-standing rivals in the core database arena include IBM (DB2) and Microsoft (SQL Server), locked in a four-decade three-way battle. In the cloud era, Oracle's OCI directly challenges Amazon AWS, Microsoft Azure, and Google Cloud; despite a clear scale gap, it is carving out openings via massive AI compute orders. In enterprise application software, Oracle competes head-to-head with SAP following the acquisitions of PeopleSoft and NetSuite. Additionally, Salesforce—founded by former executive Marc Benioff—has long suppressed Oracle in SaaS, while Snowflake and Databricks nibble away at its traditional database stronghold in cloud data warehousing.