The invoice says £15,000. The purchase order says £12,500. The receiving document shows only 80% of the ordered quantity was delivered. Should this invoice be paid?

The answer requires matching information across multiple sources. If any one source is considered in isolation, the picture is incomplete. The invoice might look legitimate by itself. The purchase order might seem adequate by itself. Only when all three are compared does the problem—or the accuracy—become visible.

What Three-Way Matching Is

Three-way matching compares invoices against purchase orders and receiving documents before authorising payment. Each document captures different information, and the combination provides robust verification.

The purchase order confirms what was ordered, by whom, at what price, from which supplier. It represents authorised commitment to spend.

The receiving document confirms what was actually delivered. Did the goods arrive? In what quantity? In acceptable condition? It represents physical receipt.

The invoice claims payment for goods or services. It represents the supplier's assertion of what they're owed.

Matching these three documents answers critical questions. Did we order this? Did we receive it? Does the amount match what we agreed? Discrepancies at any point should trigger investigation before payment proceeds.

The Discrepancy Types

Different matching failures indicate different problems.

Invoice without purchase order might indicate maverick purchasing—someone buying without proper authorisation. Or it might be a legitimate purchase where the PO was never created, or was created but not linked. Either way, it deserves investigation.

Invoice exceeding purchase order value could indicate price changes, quantity variations, or attempt to overcharge. Small variances may be within tolerance. Large variances require explanation before payment.

Invoice without receipt confirmation suggests goods or services that haven't actually been delivered. This could be timing—the goods are in transit—or it could be attempted fraud, invoicing for deliveries that never happened.

Receipt not matching order might indicate shipping errors, partial deliveries, or quality issues. If only 80% of an order arrived, only 80% should be paid until the remainder is delivered or the shortfall resolved.

Pricing discrepancies between PO and invoice require resolution. Did the price change legitimately? Was there an error on the PO? Is the supplier attempting to overcharge? The answer determines the appropriate response.

Tolerance Thresholds

Perfect matching across all fields all the time isn't practical. Minor variations occur legitimately, and investigating every penny difference creates more cost than value.

Tolerance thresholds define acceptable variance. An invoice 2% higher than the PO might pass automatically. An invoice 10% higher might require manual review. An invoice 50% higher might be blocked entirely pending investigation.

Different tolerances may apply to different fields. Price variance might have tight tolerance—you agreed a price and expect it to be honoured. Quantity variance might have wider tolerance—shipping sometimes results in minor over- or under-delivery.

Tolerances should be calibrated to risk. High-value items might have tighter tolerances than low-value items. Strategic suppliers might have different treatment than occasional vendors. The goal is focusing attention where it matters most.

The Process Flow

In a typical three-way matching process, invoices are received and captured, automatically matched against purchase orders and receipts where possible, routed for exception handling when discrepancies exist, and queued for payment when matching is successful.

Automation handles the matching itself. Modern systems compare invoice data against PO and receipt data, identify discrepancies, apply tolerance rules, and route items appropriately. Manual matching at scale would be impossibly labour-intensive.

Exception handling remains substantially human. When the system flags a discrepancy, someone must investigate. Is this a genuine problem or a data quality issue? Who needs to be involved in resolution? What's the correct outcome?

Resolution paths vary. The invoice might be returned to the supplier as incorrect. The PO might be amended to reflect agreed changes. The receipt might be updated when goods arrive. Payment might proceed once the discrepancy is explained.

Common Challenges

Organisations implementing or operating three-way matching encounter predictable difficulties.

Purchase order compliance is often low. If people buy without raising POs, invoices arrive without documents to match against. The matching process then surfaces this compliance problem, which is valuable but creates exception workload. Improving PO compliance is often prerequisite to effective matching.

Service purchases match poorly. A consulting engagement doesn't have a receiving document in the traditional sense. How do you confirm delivery of advice? Matching models need adaptation for services—perhaps sign-off by the budget holder rather than goods receipt.

Data quality undermines matching. If invoice data is captured incorrectly, if PO data has errors, if receipt entries are inaccurate, the matching will fail—not because of genuine discrepancies but because of bad data. Investing in data quality improves matching outcomes.

Volume creates workload. If a high percentage of transactions generate exceptions, the exception handling queue becomes overwhelming. Reducing exception rates through process improvement, tolerance calibration, and supplier education is essential for sustainable operation.

The Business Case

Three-way matching isn't free. It requires systems investment, process discipline, and exception handling capacity. What justifies this investment?

Error prevention catches mistakes before they become payments. The supplier who accidentally sent an invoice for £50,000 instead of £5,000. The duplicate invoice that slipped through. The price that drifted from contract. These errors are easier to prevent than to recover.

Fraud deterrence is significant. Knowing that invoices will be matched against orders and receipts discourages attempts to submit fraudulent claims. The control's existence has value even when it doesn't catch specific frauds.

Financial control satisfaction matters for auditors and regulators. Three-way matching is a fundamental control that external and internal audit expect to see. Its absence raises questions; its presence provides assurance.

Accurate payment timing follows from matching. When you know goods have been received and invoices are correct, you can pay appropriately—capturing discounts when available, managing cash flow deliberately, satisfying suppliers reliably.

Beyond the Basics

Mature organisations extend matching beyond the basic three-way model.

Contract matching compares invoices not just against individual POs but against master agreements. Is this supplier charging the rates in our framework contract? This catches price creep and non-compliant pricing.

Quality documentation matching incorporates additional receipt information. Did the delivered goods pass inspection? Is the required certification present? These quality gates affect whether payment should proceed.

Tax and compliance matching verifies that invoices contain required information for tax recovery and regulatory compliance. Invalid tax numbers, incorrect addresses, missing details—all caught before payment.

The matching principle extends wherever verification is possible. The more data sources that can be compared, the more confidence you have that payment is appropriate. Three-way matching is the foundation; sophisticated organisations build further.