The Procure-to-Pay Process, Step by Step
Procure-to-pay (P2P) is the full journey from "we need this" to "the supplier is paid" — the eight steps, the controls at each, and where the process leaks when steps are skipped.
Procure-to-pay (P2P) is the end-to-end business process that starts when someone identifies a need and ends when the supplier is paid and the transaction is on the books. It spans two worlds that often don't talk — procurement (requesting, sourcing, ordering, receiving) and finance (matching, approving, paying) — and most of its failures live precisely in the handoffs between them.
The eight steps
- 1. Requisition. Someone states the need — item, quantity, purpose, budget line — and the request enters the system. Verbal needs are not requisitions; they are future disputes.
- 2. Approval. The request routes by threshold: small purchases to the budget holder, larger ones up the chain. The control is that commitment cannot precede approval.
- 3. Sourcing. For routine items, the contracted supplier; above thresholds, competitive quotations compared on the record — or for most organizations, an RFQ to several suppliers with a documented award.
- 4. Purchase order. The formal commitment: items, quantities, prices, terms. The PO is the contract the rest of the process measures against.
- 5. Receiving. Goods counted against the PO by someone other than the requester; services certified against the agreed scope. This step is where paper meets physical truth.
- 6. Matching. The three-way match — PO, receipt, invoice — runs before payment. Mismatches route to humans; matches flow through.
- 7. Payment. Released per terms, referenced to the invoice and PO, through a traceable channel.
- 8. Recording. The transaction lands in the ledger against its budget line — automatically if the system is connected, laboriously if not.
Where P2P leaks, step by step
| Skipped or weak step | The leak it opens |
|---|---|
| No requisition (purchases start with a phone call) | Maverick spend: commitments nobody approved against budgets nobody checked |
| Approval after commitment | The approval becomes theatre — the money is already spent |
| No real sourcing | Loyalty premiums and quotation theatre; prices drift years without challenge |
| No PO, or POs raised after the invoice | Nothing to match against; every invoice is negotiation |
| Receiving by signature instead of count | Short deliveries and substitutions paid in full |
| No match before payment | Paying for the undelivered, the unordered, and the duplicate |
| Payment outside the process | The audit trail breaks exactly where money moves |
| Manual recording | The ledger and the process drift apart; month-end becomes reconciliation archaeology |
Commitment accounting: the step most SMEs miss
An approved PO is money promised. Budgets that only track invoices received will happily approve spending against funds already committed — the classic double-commit overspend. Mature P2P counts open commitments against budget lines from step 4, not step 8.
What "good" looks like
A healthy P2P process in six tests
- Any purchase can be traced requisition → payment in minutes, with every document linked.
- No PO exists without a prior approved requisition; no payment without a match.
- Budget availability shows committed and actual, not just actual.
- Exceptions (emergencies, single-source, over-tolerance) are visible in a report, each with a reason.
- Supplier performance — delivery, quality, price stability — accumulates from the process itself.
- The team follows the process because it is the fastest path, not because a memo said so.
That last test is the design principle: P2P survives only when the compliant path is the convenient one — requisitions from a phone, approvals in a tap, matching automatic. For sector-flavored versions of the same chain, see NGO procurement, SACCO governance, and school term-cycle buying.
Run the whole chain in one system
Requisition to payment with approvals, matching, budget commitments, and a self-assembling audit trail.
See P2P in AWRAFrequently asked questions
What is the difference between procure-to-pay and source-to-pay?
Source-to-pay (S2P) starts earlier — strategic sourcing, supplier selection, and contracting — then contains P2P inside it. P2P assumes suppliers and terms broadly exist and governs the transaction flow. Most organizations build P2P discipline first; S2P maturity follows.
Do small organizations really need all eight steps?
They need all eight decisions — but scaled: below a petty threshold, requisition-approval-order can be one quick record and the match is a stapled receipt. What kills small organizations is not lightweight steps; it is absent ones, usually receiving and matching.
Where should P2P automation start?
At the two ends that hurt most: requisitions-with-approvals (kills maverick spend) and three-way matching (kills payment leaks). Sourcing automation and supplier portals come after the transaction backbone is solid.
How does P2P relate to order-to-cash?
Mirror images: P2P is you buying (money out); order-to-cash (O2C) is you selling (money in) — order, fulfil, invoice, collect. A connected system runs both against the same inventory and ledger, which is what makes the books agree with the warehouse.
