The OCI-only deal structure is the cleanest Oracle Cloud procurement model and the one Oracle's sales team most reliably tries to talk buyers out of. The OCI-only structure means an Annual Universal Credit commit standing alone — no bundled on-premises Database licences, no Java SE Universal Subscription, no Fusion Cloud SaaS commitment, no Support Rewards interlock with the on-premises support stream. The contract is a pure cloud commit benchmarked against AWS and Azure. The structure protects buyer optionality — at the end of term, the buyer either renews the OCI commit, reduces it, walks away, or migrates workloads to a hyperscaler — without unwinding entangled product commitments. Oracle prefers bundles because bundles create switching cost. The buyer prefers OCI-only because OCI-only preserves the right to walk.
The OCI-only structure is the right model for buyers who are evaluating Oracle Cloud on its merits — performance, price, specific Oracle workload (Autonomous Database, Exadata Cloud, OCI GenAI), or strategic interest in Database@AWS/Azure/GCP — but who are not committed to the broader Oracle ecosystem. It is also the right structure for buyers exiting an Oracle ULA who want to land at a cloud commit without taking on Oracle's broader Fusion SaaS or Java pricing. This article walks the buyer-side blueprint: the structural elements, the negotiable clauses, the BYOL carve-outs, and the optionality protections.
What an OCI-only deal actually looks like
The OCI-only structure has five core elements.
Element 1: Annual Universal Credit commit. A single annual dollar amount the customer commits to spend on OCI services. Universal Credits are flexible — they apply to any OCI service (compute, storage, network, Autonomous Database, Exadata, OCI GenAI, OCI for Government, etc.) at the published rate-card minus the negotiated discount.
Element 2: Term length. 1, 3, or 5 years. The discount scale increases with term length and commit size.
Element 3: Discount tier. The negotiated discount off published OCI rate-card pricing. Achievable: 35 – 60% at $500K+ annual commits, up to 70% at $5M+ commits with 5-year terms.
Element 4: Service-level commitments. SLAs for compute uptime, Autonomous Database availability, storage durability, network performance. Default OCI SLAs are workmanlike; negotiated SLAs add credit mechanics for missed availability.
Element 5: Order Form scope. Single Order Form for the OCI commit; no cross-references to perpetual on-premises licences, no Java subscription, no Fusion Cloud SKUs. The clean Order Form structure is the optionality protection.
Why Oracle resists the OCI-only structure
Oracle's sales playbook prefers bundles. Three reasons.
Bundle discount allocation. Oracle's internal commission and discount allocation favours multi-product deals. A pure OCI deal carries less Deal Desk political capital than a bundle that includes Database, Java, and Fusion. Sales reps push bundles because their compensation incentivises bundling.
Switching cost creation. An OCI-only customer can walk at end of term. A customer with an interlocked OCI + Java + Fusion + on-premises support bundle faces unwinding cost that effectively locks them in. Oracle's customer-retention metrics favour interlocked customers; bundles serve that retention.
Support Rewards leverage. Support Rewards — the programme that gives 25¢ or 33¢ of on-premises support credit for every dollar of OCI consumption — only delivers value when the customer also has a large on-premises support stream. A pure OCI buyer with no support stream gets no Support Rewards benefit; Oracle sales push the on-premises licence bundle to "activate" the Rewards path. The framing is misleading: Support Rewards are not a discount, they are a discount on a cost the customer wouldn't otherwise have.
The OCI-only negotiation levers
Lever 1: Benchmark against AWS and Azure
OCI-only deals are evaluated against hyperscaler benchmarks. Compute, storage, and database pricing on OCI is competitive with AWS and Azure at list, and outperforms them after the OCI commit discount. The buyer-side framing: "AWS quoted $X for equivalent workload, Azure $Y, OCI must be at or below those benchmarks." The framing forces Oracle to price against the alternative rather than against its own list.
Lever 2: 35 – 60% discount on Universal Credits
At $500K+ annual commit, the achievable discount runs 35 – 55%. At $2M+, 50 – 65%. At $5M+ with 3-year term, 60 – 70%. Oracle's first quote typically opens at 25 – 35% off; the negotiated outcome lands meaningfully deeper with credible hyperscaler context.
Lever 3: Annual ramp instead of flat commit
A "ramp" structure — Year 1 commit $1M, Year 2 $1.5M, Year 3 $2M — matches the customer's likely consumption growth and reduces the risk of overcommit. Oracle prefers flat commits because they front-load revenue; the buyer prefers ramps because ramps align with reality. Both are negotiable.
Lever 4: Carry-forward of unused credits
Default OCI commits expire annually — unused credits are forfeit. Negotiated carry-forward (10 – 30% of unused credits roll to the next year) protects against consumption variance. Critical for customers in early OCI adoption phase.
Lever 5: BYOL carve-out
BYOL (Bring Your Own Licence) rules permit existing on-premises licences to be used on OCI without re-licensing. Default BYOL has restrictions; negotiated BYOL carve-outs clarify which on-premises licences may be used on which OCI services, with documented audit-defensible reconciliation. For deeper BYOL coverage, see BYOL clauses to fight for.
Customer evaluating OCI for Autonomous Database, Exadata Cloud at Customer, and OCI GenAI workloads. Oracle's first proposal: bundle OCI commit ($3M annual) with Java Universal Subscription (14,000 employees × $5.70 = $958K annual) and a Fusion HCM commitment ($1.6M annual) for "consolidated 55% discount." Buyer-side advisory engaged. Forensic review: OCI was the actual workload need; Java commitment was inflated (forensic Java footprint of 1,200 active users); Fusion HCM was not on the technology roadmap. Negotiated outcome: OCI-only commit ($3M annual, 3-year, 58% off list with annual ramp from $2M Y1 to $4M Y3). Java handled separately with Affiliate carve-out reducing employee count to 4,800 ($164K annual). Fusion HCM removed entirely from scope. Total cost: $3.16M annual Y1 vs Oracle's bundled $5.55M. Saving: $2.39M annual, $7.2M over 3 years. Optionality preserved.
OCI-only contract clauses that matter most
Annual commit and shortfall language
Default OCI Universal Credit contracts permit Oracle to invoice the full annual commit even if consumption is lower. Negotiated shortfall protection — e.g., 80% utilisation floor with the balance refunded or rolled forward — protects against over-commitment. Critical for customers with uncertain consumption profile.
Rate card lock
Default OCI contracts permit Oracle to change the published rate card during the term, with the customer's discount applied to the new rate card. Negotiated rate-card lock (typically 12 or 24 months) prevents Oracle from raising the underlying SKU price while the customer's commit is in flight.
Service substitution
If Oracle deprecates an OCI service or substantially changes its specification mid-term, the buyer needs the right to substitute equivalent services or terminate the commit pro rata. Default contracts do not address; negotiated substitution language closes the gap.
Exit data assistance
At term end, the buyer must be able to extract data from OCI without penalty. Default contracts provide minimal extract assistance; negotiated termination assistance (typically 90 – 180 days of read-only access plus structured data extract at no incremental charge) is essential.
Renewal pricing protection
Default OCI renewals are at-the-time pricing — no protection against discount compression. Negotiated renewal pricing protection — a documented discount floor applied to the renewal quote — prevents the renewal from clearing 15 – 25% worse than the original commit. For broader cap coverage, see Renewal Price Cap Clauses.
Negotiating an OCI Universal Credit commit?
We benchmark OCI commit pricing against AWS and Azure, build the buyer-side BATNA, structure the contract for optionality, and negotiate the BYOL carve-outs. Send the proposal. We return a benchmark plot and a negotiation strategy in five business days.
Request an OCI deal review →OCI-only vs bundle — the decision framework
The OCI-only-vs-bundle decision turns on five factors.
Factor 1: Existing Oracle estate. Buyers with significant on-premises Oracle Database investment and active support stream can extract more bundle discount through Support Rewards activation. Buyers without that estate get less bundle value and should hold to OCI-only.
Factor 2: Java SE footprint. Buyers with confirmed Java SE usage at scale may benefit from bundle pricing on Java. Buyers without confirmed Java usage at scale should never accept a Java line item in an OCI bundle — Oracle's quote will over-count.
Factor 3: Cloud-vs-on-prem trajectory. Buyers actively shifting workloads to cloud should structure for that trajectory. Bundles that include on-premises perpetual licences may not match the cloud trajectory and create stranded cost.
Factor 4: Optionality value. Organisations that may walk from Oracle at end of term should prefer OCI-only. Organisations committed to Oracle long-term may accept bundle lock-in for additional discount.
Factor 5: Multi-cloud strategy. Buyers pursuing Oracle Database@AWS, @Azure, or @Google Cloud should structure to preserve multi-cloud BYOL flexibility — OCI-only with explicit BYOL carve-outs preserves this; bundles often restrict it.
"An OCI-only structure is not the discount-maximising structure. It is the optionality-maximising structure. The choice between bundle discount and contract flexibility is the central decision in OCI procurement."
The OCI-only negotiation timeline
- Week 0 – 4: Workload scoping. Identify the specific workloads going to OCI, the consumption profile, the BYOL eligibility, and the multi-cloud context.
- Week 4 – 6: Benchmark RFI. Issue informal AWS and Azure pricing inquiries for equivalent workloads. The benchmark becomes the negotiation anchor.
- Week 6 – 8: First OCI quote. Oracle's opening quote typically arrives as a bundle. Decline the bundle, request the OCI-only quote against published rate card.
- Week 8 – 12: Negotiation. Push discount depth, ramp structure, carry-forward, rate-card lock, exit assistance. Use AWS and Azure benchmarks throughout.
- Week 12 – 14: Deal Desk escalation. Final discount and structural clauses cleared by Oracle Deal Desk. For escalation paths, see escalation paths inside Oracle.
- Week 14 – 16: Signature. Final Order Form signed with all amendments.
The end-to-end cycle for a $1M+ OCI-only commit is typically 12 – 16 weeks. Shorter cycles produce materially worse outcomes; Oracle's bundle pressure intensifies under time compression.
The Oracle Database@Hyperscaler context
Oracle's Database@AWS, Database@Azure, and Database@Google Cloud offerings change the OCI-only economics. Buyers running Oracle Database workloads now have a credible "Oracle Database in the hyperscaler of choice" option that doesn't require OCI commitment. The OCI-only structure should explicitly preserve the right to deploy Oracle Database on AWS, Azure, or Google Cloud at customer election — and should not restrict BYOL movement between OCI and the hyperscalers. For deeper coverage, see the Database@AWS guide and multi-cloud BYOL rules.
Common OCI-only deal mistakes
Mistake 1: Accepting the bundle without quoting OCI-only. Oracle's first quote should be challenged: "Quote OCI-only against rate card. We will evaluate the bundle premium separately."
Mistake 2: Over-committing. Universal Credit commits should match conservative consumption forecasts. Over-commit is invoiced even if unused. Under-commit can be expanded mid-term.
Mistake 3: Ignoring carry-forward. Without carry-forward, Q4 consumption variance triggers forfeit of unused commit. Negotiate carry-forward into every multi-year deal.
Mistake 4: No rate-card lock. Oracle's right to change underlying rate-card pricing mid-term can erase the negotiated discount. Lock the rate card or lock the SKU pricing explicitly.
Mistake 5: No exit assistance. Default OCI exit terms are minimal. Negotiate explicit data extract assistance, read-only access during transition, and runbook documentation.
Evaluating Oracle Cloud against AWS or Azure?
We benchmark OCI commits against hyperscaler equivalents, structure the deal for buyer optionality, and negotiate the BYOL and exit clauses that matter at end-of-term. End-to-end advisory engagement.
Explore the OCI advisory service →OCI-only structure and the broader Oracle estate
The OCI-only structure does not preclude future expansion. A customer who signs an OCI-only Year 1 commit can add Fusion Cloud, Java Universal Subscription, or additional on-premises licences in subsequent Ordering Documents — each negotiated on its merits, each price-discovered separately. The "OCI-only first, expand later" approach generally produces better economics than the "everything bundled at year 1" approach because each subsequent commit is negotiated against the prior baseline, not absorbed into a single bundle anchor. For the broader Oracle negotiation methodology, see the Oracle negotiation master guide and the Oracle cloud licensing guide.
Frequently asked questions
What is an OCI-only deal structure?
An OCI-only deal structure is an Oracle Cloud commit that stands alone — no bundled on-premises licences, no Fusion Cloud SaaS subscription, no Java SE Universal Subscription, no Support Rewards interlock. The customer commits to Annual Universal Credits and consumes OCI services against that commit. The structure is the cleanest way to procure Oracle Cloud without taking on Oracle's broader product portfolio.
Why would I want an OCI-only structure instead of a bundle?
Three reasons: (1) flexibility — no interlock with on-premises commitments that may not match the cloud roadmap; (2) negotiating leverage — pure OCI deals are evaluated against AWS and Azure benchmarks rather than against Oracle's bundle economics; (3) exit optionality — OCI-only contracts can be wound down at term end without unwinding entangled on-premises and SaaS commitments. The trade-off is the foregone bundle discount.
What discounts are achievable on an OCI-only Universal Credit commit?
OCI-only Universal Credit deals achieve 35–60% off published OCI rate card at $500K+ annual commit levels. Higher commits and longer terms push the discount toward 60–70%. Bundle structures (OCI + on-premises + Java + Fusion) can achieve nominally higher discounts but at the cost of optionality. The decision is between discount maximisation and contract flexibility.
How long should an OCI-only commit term be?
1–year commits offer minimal Oracle discount uplift and maximum optionality. 3–year commits unlock 10–15% incremental discount with manageable lock-in. 5–year commits unlock 20–30% incremental discount but lock in to OCI through full cloud cycle. For organisations with stable OCI strategy, 3–year is the sweet spot. For organisations evaluating multi-cloud or considering Oracle Database@Hyperscaler alternatives, hold to 1–year and re-negotiate.
Related reading
Want an OCI-only deal blueprint for your next commit?
Send the workload scope and the consumption profile. We return an OCI-only commit structure, AWS/Azure benchmark plot, and a buyer-side negotiation strategy. Confidential. Five business days.
Schedule an OCI deal review →Independent · Confidential · Not affiliated with Oracle Corporation
Free briefing every Friday.
Oracle audit alerts, Deal Desk intelligence, Java licensing updates, and negotiation tactics — written by former Oracle insiders. Read by 2,000+ enterprise buyers.
No spam. Unsubscribe anytime. Not affiliated with Oracle Corporation.