Oracle EBS to SAP S/4HANA: Migration Economics That Survive Board Scrutiny

Why a credible Oracle EBS to SAP S/4HANA plan is the lever

An Oracle EBS to SAP S/4HANA migration is the second-most expensive contractual exit a CIO will sign during a career. The most expensive is the one that runs without a buyer-side economics model. We have defended both sides of this trade — clients who used the migration plan to negotiate a softer Oracle landing and stayed, and clients who exited Oracle E-Business Suite outright. The financial outcome of either path is set in the first ninety days, before the SAP RFP closes. It is set by the quality of the Oracle EBS to SAP S/4HANA migration TCO model the buyer brings to the table.

Oracle's account team is structured to defend the EBS install base because EBS pulls Oracle Database, WebLogic, Identity Management, BI, and on a large estate Exadata. Loss of an EBS workload triggers a cascade. Oracle's deal desk will authorise discount levels in defence of an EBS account that it will not authorise anywhere else in the portfolio. The defended discount only appears when the threat package is credible: a costed S/4HANA model, named system integrator, board-approved budget tranche, and a published cutover sequence. Without all four, the renewal conversation produces an 8 to 15 percent discount and no structural protection. With all four, we have negotiated 45 to 65 percent off renewal pricing, audit suspension language, support uplift caps, and right-to-reduce processor counts. The buyer-side model is the lever that unlocks the deal-desk tier of authority. The same cascade applies to other middleware-heavy Oracle estates: a parallel WebLogic to Tomcat or WildFly migration on the retained Oracle workloads tightens the negotiation position further by removing Oracle's argument that the customer is locked into the wider middleware stack.

The same model has to honestly tell the customer when staying on Oracle is the cheaper outcome and when displacement is. Most ERP migration business cases are written by the destination vendor and do not pass that honesty test. The buyer-side Oracle EBS to SAP S/4HANA migration economics framework forces the honesty. For organisations that have not yet decided whether to migrate at all, we maintain a separate side-by-side Oracle EBS vs SAP S/4HANA licensing and TCO comparison that quantifies the EBS Premier Support runway through 2034 against the S/4HANA RISE subscription stack. The continuation case wins more often than either vendor's sales material admits — and the alternative Microsoft path is covered in our Oracle EBS vs Dynamics 365 F&O comparison.

The forensic entitlement baseline is the foundation of every part of the model. We pull it from the OMA and the underlying Order Forms, reconcile it against USMM output and the GLAS portal, and then check it against actual user counts in the EBS responsibility tables. Skip the baseline and every later line in the model is approximate. See the Oracle database licensing guide for the entitlement framework that has to be cleaned up before the migration economics can be built.

The Oracle cost stack that goes to zero (and the part that does not)

The Oracle-side cost saving on an EBS to S/4HANA migration is wider than the EBS application licence line. The full Oracle stack that depends on EBS contains application licences, the Application-Specific Full Use (ASFU) Oracle Database entitlements that came bundled with EBS, the supporting middleware, the engineered system maintenance if Exadata is in scope, and the operational labour that exists because the estate runs on Oracle. Every one of those lines either goes to zero or right-sizes when EBS is retired.

The Application-Specific Full Use line is the one most CFOs miss when they think about the EBS exit. ASFU Database entitlements were granted at a discount because they were restricted to running the bundled Oracle Applications. When EBS is retired, those entitlements expire — they cannot be redeployed to run a non-Oracle workload. Customers sometimes assume the ASFU sunk cost can be reused for the data warehouse or a separate Oracle workload; it cannot. The Order Form language locks the entitlement to the named application. Any reuse without re-licensing is an audit exposure that LMS will find. We push back forensically on this one because Oracle's account team will sometimes argue the ASFU is transferable when negotiating a renewal sweetener; the OMA language says it is not.

SAP S/4HANA cost the model has to load honestly

The S/4HANA side of the model is where SAP's business case overstates the saving. SAP's standard TCO calculator excludes the BTP consumption ramp, the SAP Activate methodology cost overruns, the third-party integration redevelopment, the data migration tooling, and the operational uplift for a HANA-first run-state. The buyer-side Oracle EBS to SAP S/4HANA migration economics model has to load every one of those lines at a defensible figure. We benchmark the SAP side against the destination contract and against three real engagements at comparable scale to keep SAP's commercial team honest.

The two licensing models that matter on the S/4HANA side are S/4HANA Cloud Private Edition (formerly known as RISE) and S/4HANA on-premise under the traditional perpetual model. RISE is a subscription bundle that includes S/4HANA, HANA, BTP credits, and infrastructure at SAP's choice of hyperscaler. It removes the Oracle Database stack from the cost equation but adds annual price-rise clauses, consumption-pricing on BTP, and exit-cost language that mirrors Oracle's own renewal mechanics. The on-premise perpetual model preserves the customer's right to choose the database stack but reintroduces hardware and operational lines. We have seen both make sense at different scales — the RISE economics work best on estates with heavy custom code and moderate growth; the perpetual on-premise model works best on stable estates with a long horizon and an existing strong HANA operations team. See the Oracle Cloud licensing guide for how BYOL works in either direction when SAP and Oracle infrastructure coexist during cutover.

Indirect access is the SAP-side equivalent of Oracle's Custom Application Suite trap. Every system that reads or writes S/4HANA without a named SAP user — Salesforce, custom portals, integration middleware, RPA bots — can be deemed to consume licensed users. SAP's digital access licensing model (per-document pricing) is the new framework, but legacy named-user licensing is still being audited on existing customers. The buyer-side model has to reserve for an SAP audit just as it reserves for an Oracle audit. We have defended both. The buyer-side framework is structurally the same: forensic baseline, evidence, push back, negotiate the position with the deal desk.

CEMLI refactor — the line that breaks most models

CEMLI stands for Customisations, Extensions, Modifications, Localisations, Integrations. In an Oracle EBS estate of any meaningful size, the CEMLI tail is the single biggest driver of migration cost and the single most under-estimated line in vendor-built business cases. The marketing-grade S/4HANA business case assumes 80 percent functional fit out of the box and 20 percent custom code. Real EBS estates that we have measured forensically run 35 to 60 percent functional fit on day one of an S/4HANA implementation. The gap is the CEMLI tail.

The CEMLI tail has six pattern categories that drive the manual rewrite cost. We score the customer's actual estate against the patterns below using a combination of EBS responsibility audit, OAF / Forms / Reports / Workflow inventory, custom schema analysis, and integration interface count. The score produces a defensible refactor budget the buyer can carry to the SAP system integrator's RFP.

A typical mid-market EBS estate with 22 countries, 380 Forms customisations, 140 reports, 65 workflows and 95 interfaces produces a CEMLI refactor budget between $6.4M and $14M when scored properly. The vendor-built business case will quote $1.5M to $3M for the same scope. The gap is real. We have watched migrations fail user acceptance testing precisely on the CEMLI tail that was excluded from the original budget, and watched the system integrator come back at the eighteen-month mark asking for a $5M to $8M change request to close the gap. Build the gap into the model on day one or pay for it later under change-request economics that are roughly twice the planned price. The Compliance Review service framework also covers the EBS responsibility forensic audit we use as the CEMLI baseline.

Anonymised case: $42M EBS estate, 5-year displacement model

The case below is anonymised from a 2025 engagement. A European industrial manufacturer running Oracle E-Business Suite Financials, Supply Chain, Manufacturing, Procurement, and HR across 22 countries on Exadata, with $42M of net Oracle entitlement and $9.2M annual support spend. They engaged us to build the Oracle EBS to SAP S/4HANA migration economics that the board would approve as either a credible exit or a credible negotiation lever. The output is below.

The five-year picture looks unfavourable to migration. It usually does. Refactor and SI cost is front-loaded, while Oracle support saving is back-loaded. The model breaks even in year seven and produces structural saving from year eight onward, primarily because Oracle's 8 percent annual support uplift compounds out indefinitely on the stay-on-Oracle path while the S/4HANA run-state stabilises after year five. The point of the model, however, was not to migrate. We took the costed plan, the named SI letter of intent, and the board memo to Oracle's deal desk three months before the Oracle Master Agreement renewal. Oracle authorised a 58 percent discount off renewal list, capped annual support uplift at zero percent for the next five years, suspended LMS audit activity for the contract term, and granted right-to-reduce processor count by 25 percent at any anniversary. The saving against the original renewal trajectory was $14.6M over five years. The migration did not run. The model was the lever. See the Fortune 500 bank EA restructure case for a related pattern in the database stack.

Renewal redlines a credible exit unlocks at Oracle

When the Oracle EBS to SAP S/4HANA migration economics model is placed on the renewal table, Oracle's deal desk authorises clauses the local account team is not allowed to discuss. The list below covers what we have signed in the last three years on programmes where the SAP threat package was credible. The clauses do not appear when the renewal conversation opens with a request for a better price. They appear when the conversation opens with a costed exit plan and a budget approval memo.

The redlines also matter when the customer does intend to migrate. A signed audit suspension during the S/4HANA cutover window removes the single largest source of unbudgeted cost on a real EBS displacement. A signed module retirement clause sequences the Oracle support termination cleanly. A signed support uplift cap holds the residual Oracle spend flat during the eighteen months that EBS is still in production alongside the new S/4HANA stack. Without those clauses the migration has to be funded against a moving Oracle annuity. With them, the migration runs against a fixed Oracle baseline. The Audit Defence service and the Contract Negotiation service work together to land both the redlines and the audit suspension during the same renewal cycle.

Frequently asked questions

How long does an Oracle EBS to SAP S/4HANA migration take in practice?

For a mid-market Oracle E-Business Suite estate with Financials, Supply Chain and HR, a defensible Oracle EBS to SAP S/4HANA migration runs 22 to 34 months from board approval to final cutover. Large multi-country estates with extensive CEMLI customisation run 36 to 60 months. Anyone selling a 12-month full-scope migration is selling a brownfield lift that will fail at user acceptance testing on the customisation tail. The dual-run cost in the buyer-side model has to assume the longer of the two estimates, not the SI's contracted timeline.

What happens to Oracle Database licensing when EBS is decommissioned?

The Application-Specific Full Use (ASFU) entitlements that came bundled with EBS expire on the date EBS production stops. Full Use Oracle Database entitlements you bought separately under the OMA do not. The Oracle support termination must be sequenced after the S/4HANA cutover is stable, the EBS instance is in read-only archive mode, and any reporting workloads that still query Oracle have been migrated. Premature support termination removes patch access on a still-live workload and opens audit exposure with no easy remediation path.

Does SAP RISE actually reduce TCO versus running Oracle EBS on-premise?

RISE bundles S/4HANA, HANA, BTP and infrastructure into a subscription. It removes Oracle Database licence and support from the cost stack but adds SAP's annual price-rise clauses and the consumption-pricing risk on BTP credits. For estates with heavy custom code the steady-state TCO lands 10 to 25 percent below the equivalent Oracle EBS plus Oracle Database stack. For vanilla estates the saving is larger but the migration cost is also lower. The buyer-side model has to compare against the customer's actual run-state, not a generic benchmark.

Will Oracle audit during an EBS to S/4HANA migration?

Yes, frequently. Oracle LMS opens a soft audit on roughly half the customers who issue an RFP for ERP displacement and a formal audit on a third. The audit reserve has to be priced into the migration TCO. We carry a $400K to $2.1M back-licence claim reserve on every Oracle EBS displacement of meaningful scale and have defended every one to date without paying the original claim. The Oracle audit guide covers the forensic defence framework we use.

Can the EBS to S/4HANA migration plan be used as a negotiation lever first?

Yes, and this is what we recommend before committing the migration capital. A costed, board-approved Oracle EBS to SAP S/4HANA migration plan placed on the table during the next Oracle renewal moves discount authority into the deal desk. We have seen 45 to 65 percent off renewal pricing, multi-year support uplift caps, and audit suspension language appear when the SAP threat package is credible. The migration does not have to run for the model to pay back. The Oracle EBS to SAP S/4HANA migration economics model is the most expensive piece of analysis the customer will buy in a decade, and the most valuable.

Related Oracle migration economics reading