SAP estates are monolithic. The TSA exit is a sequenced engineering program.

The SAP estate inventory and scope boundary.

SAP separation starts with a complete estate inventory. SAP ECC or S/4HANA at the application layer with modules across FI, CO, MM, SD, PP, WM, and HR. HANA database or any AnyDB variant for transactional data. SAP middleware including PI, PO, or Cloud Integration. SAP BW or BW/4HANA for analytics. SAP GRC, Solution Manager, and Fiori as supporting layers. Each component carries Newco data, seller data, and shared master data inside a single production landscape that was never built to be split.

The inventory documents production, quality, and development environments, the integration points, the customizations including Z code and enhancements, and the licensed entitlements per module. Where the seller runs a global SAP instance, Newco data sits as one or more company codes, plants, and sales organizations inside that landscape. The separation work has to extract that footprint cleanly while preserving transactional integrity.

The scope boundary defines what Newco takes and what stays. Newco company code transactions move with Newco. Shared master data such as global customers, suppliers, and materials is cloned or split depending on use case. Z code that supports Newco moves with Newco. Enhancements that only serve seller processes stay. The boundary is finalized in the first 60 to 90 days of the runway and reflected in the data migration design.

A clean inventory and boundary unlocks every downstream SAP decision. The licensing strategy. The target landscape design. The migration approach. The integration redesign. Where the inventory is incomplete, downstream decisions are made on assumption and reopened later under cost pressure. The pattern aligns with the TSA exit IT separation sequence and the broader carve out IT plan.

Licensing strategy and the SAP commercial negotiation.

SAP licensing is the most consequential commercial decision in the program. Newco needs licensed SAP entitlements to operate. The options vary by product line. A new direct contract with SAP. Assignment of a portion of the seller entitlements where the contract permits. Time bound TSA usage under the seller license while Newco completes its own arrangement. RISE with SAP or GROW with SAP as a bundled subscription path. Each option has cost, timing, indirect access, and audit consequences.

SAP standard terms restrict assignment in most cases. Newco typically signs a new agreement covering named users, modules, digital access documents, and HANA runtime or full use entitlements. The negotiation is a one shot opportunity. SAP reads the situation as a captive buyer with carve out timing pressure. Without preparation, the resulting agreement carries higher list, weaker discounting, and tighter audit language than market.

The preparation covers usage metrics by module, the digital access document count, the user role design, the BATNA, and the credible target architecture. Where Newco can credibly migrate finance or supply chain to alternative platforms, the negotiation has leverage. Where Newco is committed to S/4HANA on a fixed timeline, leverage is thinner and has to come from the commercial structure rather than the alternative path.

RISE with SAP can be the right answer for Newco. The bundled subscription consolidates application, database, infrastructure, and basic services into one commercial. The bundle simplifies the commercial and accelerates the run state. The decision should be informed by long term operating economics, not by short term carve out convenience. The renegotiation discipline runs through the TSA cost-plus vs fixed-fee commercial framework.

Target landscape and the migration path.

The target landscape decision drives the rest of the program. Three patterns dominate. A system copy and prune approach where Newco takes a copy of the seller production and removes seller company codes and data. A greenfield S/4HANA implementation where Newco builds a new landscape with selective data migration. A brownfield conversion that lifts the existing ECC landscape into S/4HANA and then prunes seller content. Each path has different cost, risk, and timeline profiles.

System copy and prune is the fastest path and the most common for short TSA windows. The seller production is cloned. Pruning scripts delete non Newco company codes, sales organizations, plants, and the related transactions. Master data is filtered. The Newco landscape gets new endpoints, new client numbers, new identity setup, and new integrations. The cutover is a defined window of read only seller access followed by Newco production stand up.

Greenfield S/4HANA is the cleanest end state and the longest path. Newco implements a new S/4 landscape on a target chart of accounts, target organizational structure, and target business processes. The data migration is selective using SAP Migration Cockpit or third party tools. The cutover is preceded by parallel testing and dress rehearsals. The path is typically 14 to 24 months from kickoff to go live and is most appropriate when Newco operating model differs materially from the seller.

Brownfield conversion uses the SAP S/4HANA conversion path and then prunes. It preserves customizations and process continuity but inherits seller technical debt. The decision is informed by the workload economics, the integration complexity, and the operating team capability. The pattern aligns with the TSA exit ERP separation framework and is sized inside the carve out IT investment case.

Data migration, identity, and integration redesign.

Data migration is the most technically intricate workstream. Master data first. Customers, suppliers, materials, employees, chart of accounts, organizational hierarchies, cost centers, profit centers. Each master data domain has a defined extract, transform, and load process with data quality checks and reconciliation reports. The master data migrates before transactions because transactions reference it.

Transactional data follows. Open transactions move first because they have to continue processing. Open purchase orders, open sales orders, open production orders, open work in process, open invoices, and open receivables. Historical transactions migrate in waves based on business need and storage economics. Some historical data stays in a read only archive that Newco accesses through a reporting interface. The archive approach reduces migration scope while preserving the data for tax and audit.

Identity migration moves from the seller directory and SAP GRC into Newco identity and authorization design. The user accounts, role assignments, and authorization profiles are mapped and rebuilt. Where the seller identity continues under TSA for a transitional period, the federation arrangement is documented with explicit access scopes and termination triggers.

Integrations are the workstream that most carve outs underestimate. The SAP estate exposes hundreds of integration points through IDocs, BAPIs, RFCs, PI or PO routes, CPI flows, and direct database queries. Each integration is inventoried, tested, and rebuilt in the Newco environment. Where integrations cross the seller boundary, the boundary contract defines who maintains the connection and at what cost. The third-party connections to banks, EDI partners, and tax engines are reviewed for re contracting.

Testing, cutover, and hypercare.

Testing is sequenced through unit testing, integration testing, user acceptance testing, parallel testing, and dress rehearsals. Each test phase has defined entry criteria, exit criteria, and remediation cycles. The most consequential test is the parallel run where Newco processes a defined transaction set in both the legacy seller environment and the target Newco environment and reconciles the financial outputs to the cent.

Cutover is the window where production moves from the seller SAP estate to Newco. The window is sized to migration complexity and the business calendar. Most cutovers happen over a long weekend or month end close window with a defined freeze on transactions in the legacy and a Monday morning go live in Newco. The runbook is rehearsed twice before the actual cutover with timed steps, rollback triggers, and named owners per task.

Hypercare runs for 30 to 90 days after cutover. Production issues get triaged within defined service-level commitments. Data reconciliation runs against the legacy environment to confirm completeness. The first month end close in the new environment is the gate that confirms financial accuracy. Where issues persist into hypercare, the carve out PMO and the technology lead decide whether the issue blocks TSA exit certification per service.

The TSA exit certification confirms that Newco operates independently and the seller can shut down the SAP services that supported Newco. The certification is granular per service in the TSA service catalog. Each service has an exit date that triggers financial and operational decoupling. The discipline runs through the TSA exit milestones framework and is locked into the governance committee operating rhythm.

Cost discipline and the integrator commercial.

SAP separation programs run between $5M and $50M depending on scope and the chosen path. The cost dispersion is wide because the architecture choice and the data scope drive most of the spend. The discipline is to scope tightly, choose the lowest cost path that meets the business requirement, and hold the system integrator to the agreed scope and timeline.

The integrator commercial is the second largest line item after the SAP license commitment. Selection criteria include S/4HANA experience, carve out experience, fixed-fee willingness, and senior team continuity. The contract is fixed fee for defined deliverables with disciplined change control. Hourly time and materials is not the engagement model.

The most common cost overruns trace to scope creep, integration discovery, and Z code recreation. The fix is disciplined scope definition before contract signing, the early discovery phase that documents integrations and customizations, and a change control process that prices every scope addition in writing. Where these controls are in place, the program lands within the original budget plus or minus 10 percent.

A clean SAP separation produces a Newco that runs its own estate with licensed entitlements, supported architecture, and the optionality to evolve the stack on Newco timeline. The discipline runs through the TSA exit acceleration program and is delivered under a Fixed Fee + Portfolio Retainer engagement model.

More on platform separation.