TSA accounts payable separation moves the vendor master, the approval workflow, and the payment run out of the seller's environment and into Newco without paying any invoice twice or dropping one entirely. It is a high volume process with low tolerance for error, which is why it belongs in a planned TSA financial operations exit rather than a last minute scramble.
An accounts payable TSA looks simple on the service catalog and is anything but. Under one line item sits invoice capture, coding to the right cost center, three way matching against the purchase order and receipt, the approval workflow, vendor master maintenance, and the payment run that moves cash out of a bank account. During the TSA the seller runs all of it on its own ERP, with its own controls and its own banking connections, and the buyer pays a service charge for the privilege.
That arrangement is convenient on Day One and expensive over time. The buyer is paying for a process it does not control, running on a system it cannot see into, with vendor data it does not own outright. Every month the carve-out stays on the seller's AP function is a month of dependency and a month of charges. The objective is not to keep the service running smoothly. It is to replace it cleanly and exit on schedule.
Buyers underestimate AP because the individual transactions are small. The risk is not in any one invoice, it is in the volume and the cutover. Thousands of invoices, hundreds of vendors, and live bank payments moving between two systems create exactly the conditions for duplicate payments and missed liabilities. Treating AP as a real workstream with a named owner, a cutover plan, and a reconciliation gate is what keeps the separation clean.
The vendor master is the foundation, and it is rarely clean. In a carve-out the seller's vendor file contains records that belong to the retained business, records shared by both, and records unique to the carved-out unit. The buyer needs the subset that supports Newco, with current banking details, tax identifiers, payment terms, and remittance addresses. Extracting that subset accurately is the single most important task in the AP separation, because every downstream payment depends on it.
The buyer should insist on a clean extract early, then validate it rather than trust it. Banking details in particular must be reconfirmed directly with vendors, because stale or incorrect bank data is both a payment failure and a fraud exposure. Duplicate vendor records, a common feature of large ERP files, should be merged before migration, not carried into the new system where they reproduce the same duplicate payment risk the buyer is trying to avoid.
Once the master is extracted and cleaned, the buyer freezes it for the cutover. New vendor setups during the final weeks of the TSA should route through a controlled process so the two systems do not drift apart. The vendor master also feeds the broader master data split across finance, which is why it should be coordinated with the wider TSA month end close coordination rather than handled in isolation by the AP team alone.
The payment run is where AP touches real cash, and the cutover is where most things go wrong. The buyer needs a dated cutoff that both parties honor. Invoices dated before the cutoff are settled by the seller under the TSA, invoices after are owned and paid by Newco, and a single reconciled open payables file records exactly what is outstanding at the handover. Without that dated line, the same invoice can sit in both systems and be paid twice, or fall between them and be paid by neither.
Before the buyer takes over live payments, it should run at least one parallel cycle. The new purchase to pay process generates a payment proposal, the team compares it against what the seller's system would have produced, and any difference is investigated before money moves. Parallel running is not a formality. It is the test that proves the new approval workflow, the bank connection, and the vendor data all work together under real volume.
The buyer should also confirm the banking mechanics are ready. New payment files must be accepted by the new bank, signatories must be in place, and any positive pay or fraud controls must be live before the first independent run. Coordinating the AP cutover with the wider banking and treasury separation prevents the awkward gap where the buyer owns the payables but cannot yet release the cash.
Duplicate payment is the defining risk of AP separation, and it is preventable with a small set of disciplined controls. The first is the frozen vendor master with validated banking details, so no payment leaves on stale or fraudulent data. The second is the dated cutoff that assigns every open invoice to exactly one party. The third is a duplicate check that runs across both the legacy and the new systems during the overlap, flagging any invoice number or amount that appears twice.
The buyer should also preserve the audit trail through the cutover. Approval limits, segregation of duties, and the three way match must carry into the new system from the first day it runs live, not be retrofitted later. A carve-out that loosens AP controls during the transition invites both error and fraud at the moment the process is least visible. The new controls should be at least as tight as the seller's, and documented so the first audit after exit has something to test against.
Finally, the buyer reconciles. After the cutover, the open payables file is reconciled to the general ledger and to vendor statements, and any difference is cleared before the TSA service is switched off. This reconciliation is the proof that the separation is complete, and it ties directly into the general ledger handover described in the TSA general ledger cutover.
AP is one of the faster financial services to exit because it is repeatable and well understood. A buyer that has stood up its purchase to pay system, migrated and cleaned the vendor master, and run a parallel cycle can usually take AP in house within a few months. The temptation to extend the service because it is running quietly should be resisted. The service charge accrues every month, and the dependency on the seller's controls quietly grows the longer it lasts.
The exit should be milestone driven, not date driven alone. The buyer defines the conditions for switching off the AP service, a clean vendor master, a successful parallel run, a reconciled open payables file, live banking, and tested controls, and exits when those are met. Tying the charge to milestones rather than an open ended term keeps the seller engaged and gives the buyer a clear target. Where the term or the charge is unreasonable, it should be addressed through renegotiation rather than absorbed.
AP separation rewards planning and punishes improvisation. The buyer that maps the vendor master, the payment run, and the controls before Day One exits on schedule and pays once for every invoice. The buyer that treats AP as routine discovers the volume and the cutover risk only when a vendor calls about a payment that went out twice. The difference is preparation, not luck.
It usually covers invoice receipt and coding, three way matching, approval workflow, vendor master maintenance, and the payment run itself, often using the seller's ERP and bank connections. The buyer pays a service charge while it stands up its own purchase to pay process. The goal is to exit before the charge outweighs the cost of running AP internally.
Duplicate or missed payments during the cutover. When invoices, vendor records, and open liabilities move between two systems, the same invoice can be paid twice or dropped entirely. A clean cutover date, a frozen vendor master, and a reconciled open payables file at handover are the controls that prevent it.
Long enough to migrate the vendor master, stand up the purchase to pay system, and run parallel for one or two cycles, but no longer. AP is a high volume, repeatable process that a buyer can usually replicate within a few months. A long AP TSA is expensive and keeps the buyer dependent on the seller's controls.
Define it in the TSA, not at handover. The cleanest approach is a dated cutoff: the seller settles invoices dated before the cutover, the buyer owns everything after, and both reconcile a single open payables file. Ambiguity here produces the disputes and double payments that make AP separation messy.
The collections side of the same cutover, and the cash that must not go missing.
Read the article →Where the reconciled payables file lands when the ledger moves to Newco.
Read the article →How AP accruals and the vendor master feed the close during the transition.
Read the article →The 90-day governance, IT, finance, HR and procurement separation plan we run on live carve-outs. Get the playbook plus the bi-weekly Day One Letter — short, signal-heavy, buyer-side.
No spam. Unsubscribe in one click. · Read the overview first →

Fixed-fee proposal in 48 hours. Senior team on day one. The first conversation is always free.
Seven buyer-side moves to exit a Transition Services Agreement on time and below budget. The mark-up, the extension-fee curve, exit sequencing, and the 11-month calendar.
Sequence to the first standalone close. The ledger, the bank and intercompany all have to land before month-end, or the TSA extends by default.
One tactic, one benchmark, or one pattern from a recent buyer-side engagement. Short. Signal heavy. Free.
Subscribe to The Day One Letter →