How Oracle Makes its Money: Revenue Breakdown (FY2025)

How Does Oracle Make its Money?

Oracle Corporation (NYSE: ORCL) generated approximately $57 billion in revenue in fiscal year 2025 (fiscal year ending May 31, 2025) — up from $53.0 billion in FY2024 (+7.5% YoY) — making it one of the world’s largest enterprise software companies alongside Microsoft, Salesforce, and SAP. Oracle’s revenue comes from three fundamental businesses operating at very different growth rates: Oracle Cloud Infrastructure (OCI), a fast-growing infrastructure cloud platform competing directly with AWS, Azure, and Google Cloud; Cloud Applications, a portfolio of SaaS enterprise software products for ERP, HCM, and supply chain management; and the legacy License Support business, a high-margin, slowly-declining annuity stream from customers running Oracle databases and applications on-premises.

Oracle was founded in 1977 by Larry Ellison, Bob Miner, and Ed Oates on the premise that IBM’s relational database research papers could be commercialized. For 40+ years, Oracle’s relational database engine (Oracle Database) was the essential infrastructure for enterprise data storage — essentially every major corporation ran its core transactional systems on Oracle Database, creating one of the largest installed bases of mission-critical software in enterprise computing history. This installed base is both Oracle’s greatest competitive asset (the switching cost to move an Oracle database to PostgreSQL or SQL Server is a multi-year, multi-million-dollar project) and its strategic challenge (cloud-native competitors are building new applications on competing databases, slowly eroding Oracle’s relevance in greenfield deployments).

Under CEO Safra Catz and executive Chairman Larry Ellison (who remains heavily operationally involved), Oracle’s strategic transformation has two planks: OCI (building cloud infrastructure to host workloads from Oracle’s installed base and attract new AI/enterprise customers) and cloud application migration (converting on-premises Oracle ERP and HCM customers to cloud-based Fusion and NetSuite subscriptions). The Cerner acquisition ($28B, closed October 2022) added healthcare IT — electronic health records for hospitals — as a third growth vector, though integration has been complex. The most significant recent development is OCI’s emergence as a preferred platform for AI training workloads, with major contracts from OpenAI, xAI (Elon Musk’s AI company), Meta, and others — positioning Oracle as an unexpected AI infrastructure winner.

Key Takeaways

• OCI growing 50%+ annually — Oracle Cloud Infrastructure is the fastest-growing major cloud platform by percentage growth rate; OCI’s growth (52% YoY in recent quarters) significantly outpaces AWS (~17%), Azure (~33%), and Google Cloud (~28%); OCI’s advantage in the AI workload market is its GPU cluster architecture — Oracle has invested in large-scale Nvidia H100 and H200 GPU clusters, offering AI companies the ability to rent thousands of GPUs in a single logical cluster; OCI is also priced approximately 30–50% below AWS/Azure/GCP for comparable compute, a critical cost difference for AI startups burning through training compute budgets
• Remaining Performance Obligations (RPO) of $130B+ — Oracle’s cloud backlog (contracted future revenue from multi-year cloud agreements) has surged from $98B in FY2024 to over $130B by late FY2025; this backlog represents signed contracts that will convert to recognized revenue over 1–5 years; at $130B+ vs. ~$57B in current annual revenue, Oracle has more than 2 years of current revenue in its signed backlog — already contracted, before any new wins
• License Support ($19B+, ~34% of revenue) is the ultimate cash cow — Oracle’s on-premises database and middleware customers pay annual maintenance fees equivalent to approximately 22% of the original software license cost; these fees are essentially non-negotiable (stopping maintenance loses access to security patches) and renew at ~98%+ rates; the gross margin on License Support is approximately 85–90%; this high-margin annuity stream funds Oracle’s aggressive cloud investment without requiring external capital
• AI infrastructure bet is materializing — the $130B+ RPO includes major AI company commitments: OpenAI signed a $1B+ OCI agreement, xAI (Grok) uses OCI for training, Meta is using OCI GPU capacity, and the US federal government has large OCI contracts; Oracle’s participation in Project Stargate (a $500B AI infrastructure initiative announced January 2025 with OpenAI, SoftBank, and the US government) would make Oracle a primary infrastructure provider for national AI capability
• Cerner integration ongoing — Oracle paid $28B for Cerner (healthcare IT, electronic health records) in 2022; integration has been slower and more costly than anticipated; Oracle is migrating Cerner to run on OCI and adding AI capabilities (ambient clinical voice documentation) to differentiate against Epic Systems; if successful, the $28B investment becomes highly accretive over a 5–10 year horizon
• Database moat is extraordinary but slowly eroding — Oracle Database’s installed base at Fortune 500 companies represents decades of switching costs; major banks, insurers, and government agencies cannot migrate without multi-year engineering projects; however, new applications are increasingly being built cloud-native on open-source databases (PostgreSQL, MySQL), reducing Oracle’s relevance in greenfield deployments
• Capital structure: heavily indebted but high FCF — Oracle carries approximately $80B in long-term debt from the Cerner acquisition and data center financing; however, Oracle generates approximately $18–20B in annual operating cash flow — comfortable debt service coverage; FCF of $12–15B after elevated CapEx funds both debt service and continued cloud investment

Oracle (ORCL) Business Model

Oracle operates through the SaaS Business Model for its cloud applications and a unique lock-in infrastructure model for its database and OCI businesses. The economic architecture has three distinct layers:

Layer 1 — Database Lock-In (License Support, ~34% of revenue, ~85–90% gross margin):

Oracle’s most profitable business is the maintenance and support fees from on-premises database customers. The lock-in dynamic: enterprises spent years and tens of millions implementing Oracle Database systems; their business processes, custom code, and data schemas are tightly coupled to Oracle’s specific SQL dialect and procedural language (PL/SQL); migrating to an alternative database requires rewriting application code, re-testing all business logic, and re-training IT staff; the migration risk (data loss, system outage during a critical business process) is existential for many companies; therefore, Oracle customers pay 22% of the original license value annually in perpetuity rather than bear migration risk — and do so with ~98% renewal rates.

Layer 2 — Cloud Application Migration (SaaS, ~65% gross margin):

Oracle is converting its on-premises ERP (E-Business Suite, JD Edwards), HCM (PeopleSoft), and database customers to cloud versions: Oracle Fusion ERP, Oracle NetSuite, and Oracle Autonomous Database. The conversion economics are favorable for Oracle: the on-premises software license generates revenue once; the cloud subscription generates recurring revenue indefinitely. Fusion ERP (cloud) charges approximately $200–600 per user per month versus a one-time license; a 1,000-person ERP deployment converts from a $5–10M one-time license to a $2.4–7.2M annual subscription — a total cost increase for the customer but a massive recurring revenue improvement for Oracle. NetSuite (cloud ERP for mid-market companies, acquired 2016 for $9.3B) is growing 20%+ annually and is Oracle’s most successful cloud application business.

Layer 3 — OCI (Infrastructure Cloud, fastest growing, ~40–50% gross margin):

OCI competes with AWS, Azure, and Google Cloud as a platform for hosting compute, storage, networking, and databases in the cloud. OCI’s differentiation:

• Price: OCI charges approximately 30–50% less than AWS/Azure for equivalent compute; for AI training (which requires sustained, high-throughput GPU compute for weeks), this price difference can mean tens of millions of dollars on a single training run
• GPU cluster architecture: OCI offers “bare metal” GPU instances where customers rent full physical servers without virtualization overhead — important for large-scale AI training where communication between thousands of GPUs must be extremely fast
• Oracle Database native integration: Running Oracle Database workloads on OCI is easier, cheaper, and better-integrated than on AWS or Azure — a built-in advantage for Oracle’s installed base
• Multi-cloud partnerships: Oracle has partnerships with Amazon (Oracle Database@AWS), Microsoft (Oracle Database@Azure), and Google Cloud, allowing enterprises to run Oracle databases within a hyperscaler’s environment; this lowers adoption barriers and captures customers who want hyperscaler infrastructure but need Oracle databases

Oracle Competitors

Cloud infrastructure:

• Amazon Web Services (AWS) — the dominant cloud infrastructure provider (~33% global market share); AWS has the largest ecosystem, most services, and deepest geographic presence; OCI competes on price and GPU cluster architecture but AWS’s breadth (200+ services) and enterprise relationships give it a structural advantage for general-purpose cloud workloads; Oracle is growing faster than AWS in percentage terms but from a much smaller base; see Amazon Revenue Breakdown
• Microsoft Azure — second-largest cloud provider; Azure’s primary advantage over OCI is its Microsoft 365/Teams/Office integration and enterprise software sales relationships; Azure OpenAI Service (exclusive access to OpenAI models via Azure) is a significant differentiator for enterprise AI applications; Oracle and Microsoft have a partnership (Oracle Database@Azure) that makes them partial allies in the market; see Apple vs Microsoft for enterprise software competitive context
• Alphabet Google Cloud — third-largest cloud provider, growing rapidly; Google’s AI differentiation (proprietary TPU chips, Gemini models, DeepMind research) competes with Oracle’s GPU cluster approach for AI infrastructure; Google BigQuery competes with Oracle’s database products for analytics workloads

Enterprise software:

• Salesforce — competes with Oracle’s CRM and customer experience cloud products; Oracle has been losing CRM market share to Salesforce for 20 years, and Salesforce’s Agentforce AI capabilities are accelerating that trend; see Salesforce vs Oracle for detailed comparison and Adobe vs Salesforce for enterprise software market context
• SAP — Oracle’s primary rival in enterprise ERP; SAP dominates large enterprise ERP globally (particularly in manufacturing and process industries); Oracle Fusion competes directly with SAP S/4HANA in the cloud ERP migration market; Oracle tends to win in financial services and retail verticals while SAP is stronger in manufacturing; SAP is not separately covered on this site
• Snowflake — Snowflake’s cloud data warehousing platform competes with Oracle’s analytics and data management products; Snowflake is growing faster than Oracle in new data workloads, though Oracle’s Autonomous Database product attempts to compete; see MongoDB vs Oracle for database competitive dynamics; see Palantir vs Snowflake for enterprise data platform context
• MongoDB — MongoDB’s document database and Atlas cloud platform compete for new application development workloads that Oracle historically would have captured on its relational database; MongoDB’s developer-friendly model and flexible schema are winning greenfield applications away from Oracle Database; see MongoDB vs Oracle for direct comparison
• Epic Systems — Cerner’s primary competitor in hospital EHR systems; Epic holds approximately 35% of the US hospital market and is generally considered the quality leader in clinical workflows; Oracle must compete against Epic’s well-entrenched customer relationships and superior reputation; Epic is privately held and not separately covered on this site

Revenue Breakdown

FY2025 figures estimated from quarterly reports through Q3 FY2025. Full year ends May 31, 2025. FY2024 figures from Oracle Annual Report (10-K).

OCI — The AI Infrastructure Play

OCI’s acceleration from approximately $7B (FY2024) toward $10B+ (FY2025) is Oracle’s most important revenue story. The growth is driven by AI company demand for GPU cluster compute:

GPU cluster economics: Training a frontier AI model (GPT-4 class or above) requires running thousands of Nvidia H100 or H200 GPUs continuously for weeks to months; at $2–3/hour per GPU for 10,000 GPUs, a single training run costs $15–20M+ in compute; the AI companies (OpenAI, Anthropic, xAI, Mistral, and others) are running dozens of training runs per year; Oracle offers these companies dedicated GPU clusters at lower per-GPU pricing than AWS/Azure while maintaining the throughput performance required for large-scale training.

Why Oracle vs. AWS for AI: Oracle has invested specifically in InfiniBand networking (a high-bandwidth, low-latency interconnect) between GPU servers, which is the critical factor for distributed model training performance; additionally, Oracle offers “dedicated” GPU capacity (the customer owns the hardware for the contract duration vs. sharing a shared pool) which provides training performance consistency that a shared cloud pool cannot guarantee.

Remaining Performance Obligations from AI: Oracle’s $130B+ RPO includes substantial multi-year AI infrastructure commitments; the Project Stargate commitment (if materialized at scale) could add $50B+ to Oracle’s cloud backlog — potentially the largest single contract in enterprise technology history.

Cloud Applications — SaaS Migration Progress

Fusion ERP (cloud ERP for large enterprises): Oracle’s highest-margin cloud application; Fusion ERP is a complete rebuild of Oracle’s on-premises E-Business Suite and PeopleSoft for cloud deployment; enterprise ERP implementations take 2–5 years, creating sticky relationships — once live on Fusion ERP, the switching cost to SAP or Workday is another multi-year project; Fusion revenue growing ~15% YoY as Oracle’s on-premises customer base completes cloud migrations.

NetSuite (cloud ERP for mid-market): Oracle’s most successful organic cloud growth story; NetSuite is a cloud-native ERP targeting companies with $5M–$500M in revenue; growing 20%+ annually with approximately 40,000+ customers globally; NetSuite is the de facto standard for mid-market cloud ERP with limited direct competition at its price point and feature depth.

Oracle Health (formerly Cerner): Healthcare IT segment contributing approximately $1.5–2B in annual revenue from hospital EHR contracts; integration has been slower than planned; Oracle is rebuilding Cerner’s clinical applications to run natively on OCI and adding AI capabilities including ambient clinical voice — AI that transcribes and structures physician-patient conversations into clinical documentation automatically; success in healthcare AI could differentiate Oracle Health from Epic Systems and justify the $28B acquisition price over a 5–10 year horizon.

License Support — The High-Margin Annuity

License Support revenue (~$19.5B in FY2025, ~34% of total) is Oracle’s most profitable segment but faces structural slow decline as customers eventually migrate workloads to cloud-native stacks. The decline is gradual because:

• Mission-critical Oracle Database systems take years to migrate safely
• Oracle offers incentives for customers to add cloud workloads alongside on-premises rather than requiring full migration
• The customer base is enormous — hundreds of thousands of enterprises globally
• Oracle’s support fee increases (approximately 3–5% annually) partially offset volume attrition

License Support gross margin of approximately 87% means each dollar of License Support revenue generates $0.87 in gross profit — the highest-margin revenue stream in enterprise software. As this revenue slowly declines and Cloud Services (lower gross margin at ~60–65%) grows as a share, Oracle’s blended gross margin will compress modestly — a deliberate trade for higher absolute revenue and better revenue quality (recurring vs. episodic).

Revenue Trend (3-Year)

The cloud services mix shift (31% → 37% → ~45%) tells the strategic story: Oracle is successfully converting its business from license-and-support to cloud subscriptions. The RPO surge ($65B → $98B → $130B+) is the leading indicator — Oracle’s cloud revenue in FY2026–2027 is already substantially contracted.

Oracle (ORCL) Income Statement

FY2025 figures estimated from Q1–Q3 actuals and Q4 guidance. Final FY2025 results publish June 2025.

Gross margin compression (63.0% → ~61.8%) reflects the mix shift from high-margin License Support (87% gross margin) toward Cloud Services (60–65% gross margin); this compression is expected and manageable — cloud shifts increase revenue quality (recurring vs. episodic) at slightly lower initial margins that improve with scale.

Operating margin stability (~35.8% vs. 36.2%) reflects Oracle’s ability to hold operating leverage as it scales cloud infrastructure; R&D and sales costs are growing slower than revenue, partially offsetting gross margin compression; Oracle’s cost discipline is a differentiator vs. SaaS peers that often operate at 15–25% operating margins. See gross margin vs operating margin for context on why both metrics matter.

Oracle (ORCL) Key Financial Metrics

• Gross Margin: ~61.8% — high for enterprise technology, reflecting the software-heavy revenue mix; the License Support segment operates at ~87% gross margin, pulling the blended average well above hardware-intensive peers; OCI’s infrastructure business operates at lower margins (~40–50%) but the blended result remains strong as SaaS applications (~65%) and License Support dominate the mix
• Operating Margin: ~35.8% — among the highest in enterprise software; compare to Salesforce (~25% operating margin) or ServiceNow (~30%); Oracle’s higher margin reflects its older, more efficient cost structure where decades of mature license business fund growth investments at scale
• Free Cash Flow: Oracle generates approximately $18–20B in annual operating cash flow and $12–15B in FCF after capital expenditures; CapEx has risen sharply (from ~$3B annually pre-AI boom to ~$7–10B as Oracle builds GPU-dense data centers); FCF is deployed toward debt service (~$5B annually), dividends (~$3.5B), and selective buybacks
• Operating Leverage: Oracle’s fixed cost base (R&D, corporate overhead) creates meaningful operating leverage — once cloud infrastructure is built, incremental workloads run at minimal marginal cost; as OCI occupancy rates improve (data centers moving from 30% utilized toward 70–80% utilization), operating margins on the infrastructure business expand dramatically
• Remaining Performance Obligations (RPO): $130B+ — the single most important forward-looking metric Oracle reports; at $130B+ vs. ~$57B in current annual revenue, Oracle has more than 2 years of current revenue in signed backlog; RPO is growing faster than revenue recognition — meaning the backlog is compounding, not shrinking; this is the primary justification for Oracle’s premium valuation
• Net Debt: ~$80B gross, moderate net — Oracle carries substantial debt from the Cerner acquisition and data center financing; the debt-to-EBITDA ratio is approximately 4x (elevated but manageable for a stable software business with ~$18–20B in annual operating cash flow); Oracle has committed to prioritizing debt reduction, and the FCF profile supports gradual deleveraging while maintaining cloud investment

Is Oracle Profitable?

Yes. Oracle reported approximately $11.6 billion in net income on ~$57B in revenue in FY2025 — a 20.4% net margin. The company has been consistently profitable for decades. Operating income of ~$20.4B (35.8% margin) reflects Oracle’s highly efficient cost structure. On a free cash flow basis, Oracle generates $12–15B annually after elevated capital expenditures for AI-era data center buildouts — one of the strongest FCF profiles in enterprise software.

The primary earnings headwind is $5B in annual interest expense from Cerner acquisition and data center financing debt, which depresses reported net income relative to operating income. As Oracle’s debt gradually deleverages over 2025–2028, net income should grow faster than operating income — a structural tailwind for EPS.

Oracle (ORCL): What to Watch

1. Project Stargate and AI infrastructure contracts — the $500B US AI infrastructure initiative (announced January 2025 with OpenAI, SoftBank, Oracle, and others) is the largest potential contract in Oracle’s history; even 10–15% of $500B over 5 years would add $10–15B to Oracle’s annual revenue; watch for definitive contract announcements and Oracle-specific contract value disclosures; the scale of GPU cluster buildout required for next-generation AI training makes OCI capacity a strategic national resource
2. OCI growth rate sustainability — 50%+ annual OCI growth at $7–10B scale is remarkable; can it sustain as the base grows larger? Watch quarterly OCI revenue disclosures and the RPO trajectory (signed contracts awaiting revenue recognition); a slowdown below 30% would signal that OCI is facing more intense Amazon AWS and Microsoft Azure competition for new workloads
3. RPO conversion pace — Oracle’s $130B+ RPO is valuable only if it converts to recognized revenue on schedule; watch whether quarterly revenue recognition keeps pace with RPO growth; if RPO grows faster than revenue (which has been the trend), the backlog is compounding; if revenue outpaces new RPO additions, it suggests fewer new contract signings — the key inflection signal
4. Cerner / Oracle Health integration milestones — the $28B Cerner bet requires Oracle to demonstrate competitive parity with Epic Systems in hospital EHR quality; watch for new hospital wins, Cerner customer renewals (or cancellations), and progress on Oracle Health AI capabilities; any large hospital system defecting from Cerner to Epic would be a significant negative signal
5. License Support attrition rate — License Support’s slow decline is manageable at 1–2% per year; if attrition accelerates to 5%+ annually (customers migrating to non-Oracle cloud databases faster than expected, including to MongoDB’s Atlas or Snowflake’s data cloud), the loss of 87%-margin recurring revenue would create a structural earnings headwind that cloud growth might not fully offset; watch License Support’s year-over-year growth rate each quarter — flat to -3% is expected; -5%+ would be concerning
6. Data center CapEx and OCI margin trajectory — Oracle is spending $7–10B+ annually on data center buildout; this elevated CapEx compresses near-term free cash flow but should translate to higher OCI revenue as utilization improves; watch OCI unit economics in earnings calls — management has begun disclosing OCI capacity vs. demand metrics; improving utilization rates are the leading indicator of OCI margin expansion toward 50%+ over the next 3 years

Oracle (ORCL) Financial Summary

Oracle Corporation (NYSE: ORCL) generated approximately $57 billion in FY2025 revenue (+7.5% YoY), earning ~$11.6 billion in net income (20.4% net margin) with an ~35.8% operating margin and $130B+ in remaining performance obligations representing over 2 years of contracted future revenue. Oracle is simultaneously operating three distinct businesses at different lifecycle stages: the high-margin License Support annuity (slowly declining, funding everything at 87% gross margin), Cloud Applications (growing 20%+, NetSuite and Fusion ERP as the anchors), and OCI cloud infrastructure (growing 50%+, AI training workloads as the catalyst). The $28B Cerner healthcare acquisition remains a long-term strategic option on AI-powered clinical workflows if Oracle can close the gap with Epic Systems. The AI infrastructure thesis (Project Stargate, GPU cluster agreements with OpenAI/xAI/Meta) is the most significant potential revenue catalyst in Oracle’s history — if $50B+ in additional AI infrastructure contracts materialize, Oracle’s cloud revenue could reach $50–60B by 2030 from approximately $25B today.

For enterprise software competitive comparisons: Salesforce vs Oracle, MongoDB vs Oracle, Adobe vs Salesforce. For peer revenue breakdowns: Salesforce Revenue Breakdown, Snowflake Revenue Breakdown, MongoDB Revenue Breakdown. For sector context: Enterprise Software Sector.

Frequently Asked Questions

How does Oracle make money? Oracle makes money through three primary revenue streams. (1) Cloud Services (~45% of FY2025 revenue, ~$25.5B) — including Oracle Cloud Infrastructure (OCI) which rents compute, storage, GPU clusters, and networking to enterprises and AI companies at 50%+ growth; and Cloud Applications (SaaS products including Fusion ERP, NetSuite, and Oracle Health for healthcare). (2) License Support (~34%, ~$19.5B) — annual maintenance fees paid by customers running Oracle databases and applications on-premises; these fees are approximately 22% of the original license cost, renew at ~98% rates, and carry ~87% gross margins — Oracle’s most profitable revenue stream. (3) Cloud License and On-Premise License (~8%, ~$4.8B) — upfront fees for on-premises software, declining as the market shifts to subscriptions. Plus Hardware (~5%) and Services (~8%). Total FY2025 revenue is approximately $57 billion.

Why is Oracle’s OCI cloud growing so fast? Oracle Cloud Infrastructure (OCI) is growing 50%+ annually primarily because it has become a preferred platform for AI model training workloads. Three factors drive OCI’s AI infrastructure advantage: (1) Price — OCI charges approximately 30–50% less than AWS/Azure/Google Cloud for equivalent compute; for AI training runs consuming thousands of GPUs for weeks, this price difference can be tens of millions of dollars; (2) GPU cluster architecture — OCI offers bare-metal GPU instances and InfiniBand high-bandwidth networking between servers, critical for the fast inter-GPU communication required during large-scale distributed training; (3) Dedicated capacity — OCI offers dedicated GPU clusters (not shared pools) providing consistent performance for training workloads. Major AI companies including OpenAI, xAI, and Meta have signed large OCI contracts. Oracle’s installed base of enterprise customers (running Oracle databases) also provides a natural OCI pipeline — Oracle Database workloads migrate to OCI more easily than to AWS or Azure.

What is Oracle’s database moat and is it still relevant? Oracle Database is one of the most powerful examples of a software switching cost moat in enterprise technology. For 40+ years, major corporations have run their most critical transaction systems on Oracle Database. The switching cost is enormous: Oracle Database uses a proprietary SQL dialect and procedural language (PL/SQL) that does not transfer to other databases; application code must be rewritten; business logic must be retested; and the project carries existential business risk. This creates a situation where Oracle customers pay approximately 22% of the original license cost annually in perpetuity rather than bear migration risk — at ~98% renewal rates. The moat is still highly relevant for existing deployments. However, it is eroding at the edges: new enterprise applications are increasingly being built cloud-native on open-source databases (PostgreSQL, MySQL, MongoDB) that have no Oracle licensing cost. Each new greenfield application that chooses PostgreSQL over Oracle represents a lost future License Support customer.

How does Oracle compare to Salesforce? Oracle (~$57B FY2025 revenue) is significantly larger than Salesforce (~$38B FY2025 revenue). Oracle’s revenue mix is roughly: 34% high-margin database maintenance, 45% cloud services (OCI infrastructure + SaaS applications), and 21% declining legacy. Salesforce’s revenue is almost entirely SaaS subscription (CRM, service, marketing, commerce). Profitability: Oracle’s operating margin (~36%) significantly exceeds Salesforce’s (~25%) because Oracle’s mature License Support business funds growth at much higher margins than Salesforce’s pure-play SaaS model. Competition: Oracle and Salesforce compete in CRM and customer service applications — a market where Salesforce has been gaining share at Oracle’s expense for 20 years. Oracle tends to win enterprise customers who already run Oracle ERP (offering an integrated suite), while Salesforce wins on pure CRM functionality. See Salesforce vs Oracle for full head-to-head analysis.

What is Oracle’s RPO and why does it matter? Oracle’s Remaining Performance Obligation (RPO) is the total value of contracted cloud agreements signed but not yet recognized as revenue — over $130 billion as of late FY2025. RPO matters as the primary leading indicator of future revenue: when a large enterprise or AI company signs a multi-year Oracle cloud contract, the full contract value is added to RPO; revenue is then recognized ratably as services are delivered. A growing RPO means Oracle is signing new contracts faster than it recognizes revenue from existing ones — implying cloud revenue will continue growing for years even without new contract wins. Oracle’s RPO grew from approximately $65B (FY2023) to $98B (FY2024) to $130B+ (FY2025) — implying Oracle’s cloud revenue should roughly double over the next 3–5 years from conversion of existing contracts alone. The RPO surge is primarily driven by multi-year AI infrastructure agreements (OCI GPU capacity) from OpenAI, xAI, Meta, and the US government.