What Does Oracle Do? The $59 Billion Mystery

Oracle Corporation made $59 billion in revenue last year. The company ranks among Silicon Valley’s oldest tech giants. Oracle stock jumped 40% this year, briefly making founder Larry Ellison the richest person in the world. Headlines trumpet Oracle’s involvement with TikTok, artificial intelligence, and cloud computing. Yet if you ask ten people what does Oracle do, you’ll get ten confused shrugs.

Try visiting Oracle’s website for answers. You’ll find the words “cloud, AI, applications, and infrastructure” arranged into dozens of combinations interspersed with acronyms like OCI, SCM, and ERP. The site promises to help people “see data in new ways, discover insights, and unlock new possibilities.” None of this explains what Oracle actually does or how it generates $59 billion annually doing it.

The answer is simultaneously simple and deliberately obscure. Oracle is a database company that discovered the most profitable products are the ones customers can never leave.

Databases Are Sticky

At its core, Oracle manages databases. That sounds archaic and boring because it is. Databases store and organize information like customer records, financial transactions, and supply chain data. It’s back-end infrastructure that keeps businesses running. Nothing sexy about it.

But databases have a crucial characteristic that Oracle learned to exploit. They’re sticky. Once a company’s data lives inside an Oracle database, extracting it becomes extraordinarily difficult. Not impossible, but expensive, time-consuming, and risky enough that most companies never attempt it. Migrations fail. Data gets corrupted. Business operations halt. The pain of leaving exceeds the pain of staying.

Oracle recognized this dynamic early and built an entire business model around it. First, sell companies database software. Then, once their data is locked in, sell them everything else. HR software, supply chain management, accounting systems, customer relationship tools. Oracle already has the data, so why not let Oracle provide software that uses that data?

This explains how a database company operates in seemingly unrelated markets. Oracle isn’t diversifying. It’s upselling captive customers who already committed when they chose Oracle databases decades ago. The company reports that 80% of Fortune 100 companies use Oracle products. Those relationships weren’t built last year. They’re decades-old dependencies that Oracle leverages relentlessly.

The Evolution Nobody Noticed

Oracle went public in 1986 focused entirely on database software. Through the 1990s, it expanded into enterprise software that helps businesses manage operations beyond just storing data. After a string of acquisitions in the 2000s including PeopleSoft and Sun Microsystems, Oracle shifted toward cloud computing with Oracle Cloud Infrastructure designed to compete against Amazon Web Services and Google Cloud.

Each evolution followed the same pattern. Oracle identified where computing was heading, then figured out how to bring existing customers along. When cloud computing emerged as the future, Oracle couldn’t let customers migrate data to Amazon or Google. That would break the lock. So Oracle built its own cloud and migrated customers there instead, maintaining dependency while appearing to modernize.

The company now offers products for virtually every business function. If you’re a restaurant needing customer data filtered, Oracle has a product. If you’re a city managing employee benefits, Oracle has a product. If you’re the United States Air Force managing supply chains, Oracle definitely has a product. The breadth makes Oracle seem unfocused until you understand that breadth is the strategy. More products mean more reasons for customers to stay.

Larry Ellison’s Wild Ride

Oracle’s reputation as aggressive and uncompromising stems largely from its founder. Larry Ellison flew fighter jets, raced yachts, used the pickup line “Can I buy you a car?”, hired private detectives to spy on Microsoft, joked about shooting another CEO, and owns most of the Hawaiian island of Lanai. His personal brand combines billionaire eccentricity with ruthless business instincts.

Ellison built Oracle through what competitors called hardball sales tactics. Oracle became notorious for stringent audits of customer licensing compliance, finding violations that forced customers into new contracts. The company would discover that a customer licensed Oracle for 100 users but actually had 150 accessing the system. Oracle would present a massive compliance bill and offer to settle through a new multi-year contract upgrade.

These tactics worked because of database stickiness. Customers couldn’t threaten to leave. Their data was already inside Oracle systems. Migration would cost more than compliance penalties. Oracle knew this, and Ellison had no qualms exploiting it. The approach generated revenue but also resentment. Oracle earned a reputation as the company that squeezes customers rather than serves them.

Yet this reputation didn’t prevent success. Oracle grew into one of tech’s largest companies specifically because Ellison understood that customer satisfaction matters less than customer dependency. Happy customers might leave for better offers. Dependent customers stay regardless of satisfaction. Oracle optimized for dependency.

The TikTok Wild Card

Oracle’s involvement with TikTok represents its highest-profile project and demonstrates how the company operates in national security contexts. Project Texas, launched in 2022, routes all US TikTok user data through Oracle cloud servers in Austin. The $1.5 billion initiative aimed to address concerns that Chinese parent company ByteDance could access American user data or manipulate content.

Under Project Texas, Oracle stores US user data, monitors data flows, inspects source code, and oversees content moderation systems. TikTok created a new subsidiary called US Data Security Inc. governed by an independent board that reports to the Committee on Foreign Investment in the United States rather than ByteDance. Oracle acts as the technical enforcer ensuring data doesn’t flow to China.

Congress dismissed Project Texas as insufficient and passed legislation requiring ByteDance to sell TikTok or face a ban. The proposed solution involves Oracle, along with investment firms Silver Lake and Andreessen Horowitz, acquiring majority control of TikTok’s US operations. Oracle would continue hosting data while ByteDance retains under 20% ownership and licenses its algorithm to the new American entity.

Critics call this arrangement “Project Texas 2.0” that doesn’t meaningfully address security concerns. The Justice Department argued in legal filings that Oracle lacks resources to adequately monitor TikTok’s massive, complex codebase. Security experts warn that as long as ByteDance licenses the algorithm, backdoors could exist regardless of where data is stored.

For Oracle, TikTok represents both opportunity and risk. The deal reportedly generates $1 billion in cloud revenue, making TikTok likely Oracle’s largest Oracle Cloud Infrastructure customer. If TikTok gets banned, Oracle loses that revenue. If the deal goes through, Oracle gains visibility as the company that “saved” TikTok while expanding its role in sensitive national security infrastructure.

The AI Gold Rush

Oracle has spent the past year redefining itself around artificial intelligence. The company announced massive data center projects, secured partnerships with OpenAI, and positioned Oracle Cloud Infrastructure as essential for training and running large language models. Larry Ellison appeared at events alongside Elon Musk and Mark Zuckerberg discussing multi-gigawatt data centers that would power the AI future.

Oracle stock exploded on this AI narrative. The company signed deals where OpenAI would use Oracle cloud infrastructure for training models. Oracle revealed partnerships with Nvidia providing computing power for AI workloads. Ellison claimed Oracle was building AI infrastructure at scale nobody else could match. Investors responded by pushing Oracle’s market cap above $500 billion.

The AI positioning follows Oracle’s classic playbook. Identify the next computing paradigm, ensure existing customers can’t migrate to competitors, capture new customers while they’re making initial commitments. If AI companies build on Oracle infrastructure from the start, they’ll face the same migration pain that keeps database customers locked in.

However, Oracle’s AI strategy faces serious questions about sustainability. Much of the announced revenue comes from projected future deals rather than realized income. OpenAI reportedly generates only $13 billion in revenue, nowhere near enough to cover the billions in compute it owes Oracle, Nvidia, and other AI infrastructure providers. This raises concerns about circular deals where OpenAI’s investors fund purchases from their other portfolio companies.

Oracle also faces practical challenges building the data centers it’s announcing. Securing land, obtaining power, finding local governments willing to host massive facilities, and then figuring out how to profit from them all remain uncertain. The AI narrative has inflated Oracle stock, but delivering on AI promises will determine whether that valuation holds.

Why Being Confusing Works

Most companies try to clearly communicate what they do. Oracle thrives on confusion. When customers don’t fully understand Oracle’s products, they rely more heavily on Oracle consultants and support. When procurement teams struggle to compare Oracle against alternatives, they default to the incumbent. When audits reveal licensing violations customers didn’t know existed, Oracle captures additional revenue through compliance.

The complexity serves Oracle’s interests. Enterprise software intentionally uses jargon and technical specifications that exclude non-experts from evaluation. Terms like ERP, SCM, and OCI sound important without conveying meaning. This forces buyers to trust vendors rather than independently assess capabilities. Oracle has perfected this dynamic over four decades.

Oracle’s website reflects this strategy. Instead of explaining products clearly, it arranges buzzwords into combinations that sound sophisticated while remaining vague. “See data in new ways” could mean anything. “Unlock new possibilities” promises nothing specific. This isn’t accidental. Clear communication would enable easier comparison shopping. Confusion maintains information asymmetry that favors Oracle.

The approach works particularly well in enterprise sales where buying decisions involve committees, multi-month evaluations, and risk-averse procurement departments. Nobody gets fired for choosing Oracle. The company has existed for 47 years, serves most Fortune 100 companies, and offers comprehensive support. Those factors outweigh whether anyone fully understands what they’re buying.

The Stickiness Problem

Oracle’s business model depends on customers staying even when better alternatives exist. This creates a fundamental tension. Oracle needs to provide enough value that customers don’t attempt migrations despite the difficulty. If products become too inferior or pricing too aggressive, even sticky customers eventually leave.

Evidence suggests Oracle may be testing those limits. Customer complaints about Oracle’s audit practices and aggressive licensing enforcement are common. Performance comparisons show competitors often deliver better results at lower costs. Yet migration remains painful enough that most customers tolerate dissatisfaction rather than switch.

Cloud computing potentially weakens Oracle’s stickiness advantage. Moving data between cloud providers is easier than migrating on-premise databases. Competitors like Amazon, Google, and Microsoft offer migration tools and incentives specifically targeting Oracle customers. If the barriers to leaving decrease, Oracle’s core business model faces pressure it hasn’t experienced in decades.

Oracle’s response has been expanding into adjacent services that create new dependencies. If customers use Oracle databases, Oracle HR software, Oracle supply chain management, and Oracle cloud infrastructure, migration complexity multiplies. Each additional Oracle product represents another integration to untangle. The strategy aims to make leaving so complicated that customers never seriously consider it.

What Does Oracle Do?

The simple answer is that Oracle manages databases. The accurate answer is that Oracle built a $59 billion business by making those databases impossible to leave, then selling everything else to customers who had no choice but to listen. The company evolved from pure database software to enterprise applications to cloud infrastructure to AI positioning, but the fundamental strategy never changed.

Oracle succeeds by creating dependency, exploiting that dependency through aggressive sales and licensing practices, and continuously expanding product offerings to deepen entrenchment. The company doesn’t need to be the best at what it does. It needs to be good enough while remaining impossible to escape. Four decades of success prove this model works.

Whether Oracle’s strategy continues working depends on several factors. Can the company deliver on AI promises or will that hype collapse? Does TikTok remain a customer or get banned? Do cloud competitors successfully lure away database customers with easier migration paths? Can Oracle maintain pricing power as alternatives improve?

For now, Oracle remains one of tech’s most consistently profitable companies precisely because most people don’t understand what it does. That confusion isn’t a problem. It’s the business model

Sources

1. Wall Street Journal: Oracle Business Model Analysis
2. CNBC: Oracle TikTok Project Texas Deal
3. Lawfare: Project Texas Technical Details
4. Yahoo Finance: Oracle TikTok Role Analysis
5. PYMNTS: Oracle TikTok Acquisition Talks
6. eMarketer: TikTok US Future Analysis
7. Oracle Investor Relations
8. TechCrunch: Oracle Cloud AI Strategy

Ex Nihilo magazine is for entrepreneurs and startups, connecting them with investors and fueling the global entrepreneur movement