Carve out licensing Oracle SAP traps.

Carve out licensing Oracle SAP exposure is where the most expensive surprises hide, because both publishers price on metrics that were sized for the combined business and both treat a change of ownership as a reason to reopen entitlement and terms. When a unit running Oracle databases or SAP modules is separated from its parent, the licenses do not simply travel with it. Oracle and SAP carve out licensing turns on processor counts, named user classifications, indirect access and agreement exit terms that all shift the moment the unit becomes a standalone business, and each shift is an exposure if it is not measured first.

Why carve out licensing Oracle SAP exposure is so dangerous

Oracle and SAP sit at the centre of most enterprise estates, and their contracts are written to follow usage closely. In a carve out, two forces collide. The first is that the entitlement was counted at parent scale, so processor counts, user bands and engine metrics reflect the combined business rather than the unit being sold. The second is that the contracts carry change of control and assignment clauses that let the publisher review, reprice or refuse to transfer when ownership changes. As of June 2026, both publishers remain among the most active in post deal audits, alongside Microsoft, IBM and increasingly Broadcom for the former VMware estate. The public record shows the scale at stake: SAP pursued Anheuser Busch InBev for a reported 600 million dollars and Diageo for a reported 60 million pounds over disputed and indirect access licensing, figures that are widely reported rather than firm confirmed and are cited here as of June 2026.

The Oracle traps in a carve out

Oracle exposure in a carve out concentrates in a handful of areas. Processor and core factor counting is the first, because Oracle licenses databases by processor using a core factor table, and a standalone footprint has to be recounted rather than inherited. Virtualisation and partitioning is the second, because Oracle policy on which cores must be licensed in a virtual environment is strict and frequently misread, leaving deployments under licensed. Options and management packs are the third, because features such as partitioning, diagnostics and tuning packs are often switched on and used without a matching entitlement. The fourth is an unlimited license agreement in flight at close, where the unit has to be certified out of the agreement at a usage level that has to be measured and agreed before it leaves. Each of these is a counting question, and counting on the wrong footprint is what creates the gap.

The SAP traps in a carve out

SAP exposure runs along different lines. Indirect or digital access is the headline risk, where third party systems that read or write SAP data create license demand that the named user count does not capture, the precise issue at the centre of the publicly reported AB InBev and Diageo disputes. Named user classification is the second, because SAP users are licensed by type and a carve out is the moment to reclassify them to the roles the new entity actually needs rather than carrying the parent classifications. Engine and metric counts are the third, because SAP licenses many components on business metrics such as orders or spend that change when the unit stands alone. The fourth is the conversion path, where a move to S/4HANA carries its own terms that interact with a carve out and can reset the licensing baseline. The table below pairs each trap with the response a buyer should take before close.

How to defuse the traps before signing

The defence is the same in shape for both publishers and different in detail. Measure actual deployment and consumption on the standalone footprint, not the combined one. Test that measurement against the entitlement that will actually transfer, separating what the unit can take with it from what stays with the parent. Document indirect access for SAP and option usage for Oracle so a later audit meets evidence rather than guesswork. Then use the carve out as the moment to right size, reclassify and re contract at the new entity scale, which is often cheaper than carrying inherited metrics. This work sits inside our carve out and TSA separation service and the carve out and TSA software playbook, and it connects to separating shared software licenses and re licensing the carved out business from scratch.

Position this as commercial and licensing advisory rather than legal advice. The interpretation of any specific clause, including assignment and change of control language, belongs with the buyer own counsel. What an independent advisor brings is the measurement and the negotiation strategy that turns an Oracle or SAP carve out from an audit waiting to happen into a priced, managed transition.

Sequencing the Oracle and SAP work across the deal

Timing matters as much as measurement. Both Oracle and SAP exposures should be assessed in diligence, not after signing, because the result feeds both the price and the deal documents. In diligence the buyer measures the standalone footprint and quantifies the gap between what is deployed and what will transfer. At signing that gap drives the warranties and indemnities the buyer seeks, and it shapes the consent strategy, since the platform publishers are usually the consents that gate the rest of the separation. Through the transition services agreement the buyer keeps the unit compliant on borrowed entitlement while the standalone contracts are negotiated. At exit the new entity moves onto its own right sized agreements. Skipping the early measurement collapses this sequence into a scramble at close, which is exactly when the publisher has the most leverage.

The negotiation itself benefits from treating the carve out as a reset rather than a transfer. A unit that has carried inherited Oracle options it never used, or SAP user classifications that no longer reflect its roles, can use the separation to drop them and contract only for what it needs. Publishers will often resist a simple assignment of the old terms while being open to a fresh agreement at the new entity scale, because a new contract is also a commercial opportunity for them. A buyer who arrives with a documented, standalone measurement can steer that conversation toward a right sized deal rather than accepting the parent metrics by default.

Two practical safeguards round out the approach. First, never disable or enable Oracle options or SAP components in the run up to a measurement without recording why, because usage history is itself evidence in a later review. Second, keep the measurement working papers, including the scripts, scans and assumptions, so the standalone position can be defended if a publisher challenges it after close. Measurement that cannot be reproduced is measurement that does not hold, and on Oracle and SAP the difference between a defended position and an assumed one is frequently an eight figure number.